Hill 1111

312930010

‘V V__ «WW

1 111 z.

1.11- i’ .5!
’11! '11 3": 3.1.. 9.. :

1 I

3 -

Y. A ’5 l; ". ig‘c -.
9!.

 

 

 

 

 

 

 

,4.-.

This is to certify that the

thesis entitled

PRODUCT INNOVATION AND DIFFUSION IN THE
WOOD-BASED PANEL INDUSTRY

presented by

Larry Alan Leefers

has been accepted towards fulfillment
of the requirements for

Ph. D. degreein Forestry

 

Z2163: 2f Qfiawfig‘yr‘

Major professor

0-7639

 

OVERDUE FINES:
25¢ per du per item

RETURNING LIBRARY MATERIALS:
Place in book return to remove

charge from circulation records

Ifgvglyz golm

 

 

 

 

INNOVATION AND PRODUCT DIFFUSION
IN THE WOOD—BASED PANEL INDUSTRY

By

Larry Alan Leefers

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Forestry

1981

ABSTRACT

PRODUCT INNOVATION AND DIFFUSION IN THE
WOOD-BASED PANEL INDUSTRY

By

Larry A. Leefers

The diffusion of particleboard and southern pine plywood is exam-
ined in this study. The effects of diffusion on wood requirements
are analyzed along with the roles of various factors affecting the
diffusion of panel product innovations.

Particleboard wood requirements are projected to increase from
5.4 million tons in 1979 to between 9.3 and 10.5 million tons in 2000.
Over the same period, southern pine plywood wood requirements are pro-
jected to increase from 4.5 million tons to between 6.8 and 7.7 million
tons. The projections are based on (1) an aggregate panel consumption
submodel, (2) a logistic function submodel, and (3) average wood require-
ments per unit of output.

Though aggregate panel consumption is a function of expected econ-
cnnic activity, diffusion of specific panels is based on many interacting
:factors, many of which are not amenable to quantitative analysis. Among
true more important factors are: panel characteristics, building codes
arui standards, process innovations, economic variables, raw material
pxnice and availability, and competition from other wood—based panels.

Two diffusion submodels, the logistic function and the Gompertz

(nirve, are used to estimate potential market shares and growth rates

for the panels. However, these models are not useful for explaining
the diffusion process. Multivariate models were utilized to provide
a partial explanation for this process.

Southern pine plywood's rapid diffusion was enhanced by (1) its
similarity to existing low quality softwood plywood, (2) its accepted
standardization coinciding with initial production, (3) its low price
relative to western softwood plywood, and (4) its regional advantage.
The large amount of raw material available for processing promoted
the establishment of the southern pine plywood industry. Factors which
will eventually slow its diffusion include: (1) almost total reliance
on one end-use, sheathing, (2) rising stumpage costs, and (3) competi-
tion from new structural panels.

Particleboard's diffusion was aided by its many end-use applica-
tions, its declining relative price, and the large amounts of residues
available for processing. The slow evolution of standards and particle-
board's atypical characteristics initially hindered its diffusion.

As new specialized products capture portions of particleboard's end—
use markets and as competition for raw materials increases, its mar-

ket share growth will continue to slow.

ACKNOWLEDGEMENTS

This study was prepared for the U.S.D.A. Forest Products Labora—
tory. Without the funding they provided, this undertaking would not
have been possible. I am especially indebted to Dr. Christopher D.
Risbrudt who acted as Principal Scientist for FPL and whose disserta-
tion on technological change provided the impetus for this study.

I am grateful to Dr. Robert S. Manthy who provided insight and
advice over the past few years. I also thank Dr. John Burde for whet-
ting my interest in forest economics and guiding me to Michigan State
University.

I thank Judie Shepherd for typing the rough draft and Jill Manthy
for typing the final copy.

To my wife, Becky, Je t'aime et merci. To Karin, our daughter,

goes my thanks for being so entertaining.

ii

TABLE OF CONTENTS

LIST OF TABLES ............. . ..... ..... .......... . ............... vi

LIST OF FIGURES .. ...... . ................ . ....................... vii

SUWARYANDCONCLUSIONS 0......0.0.0.0.00.0.........OOOOOOOOOOOOO 1

INTRODUCTION .. .............. ..... ............ . .................. 17
The problem ..... ....... . ..... .............. ........ ........ 17
Objective of the study .. ........ ... ........................ 19
Scope of the study ......................................... 19
Definitions and concepts ................................... 22

METHODS OF EXAMINING INNOVATION, PRODUCT DIFFUSION,

AND ITS EFFECTS ....... . ....................................... 23b
Interacting factors ........................................ 23b
Characteristics of goods ............................... 27
Process innovations ................................... 28
Building codes and standards .......................... 29
Economic and marketing factors ........................ 30

Raw material situations .............. ........ ... ..... . 32
Models .. .......... ..... .................................... 33
Aggregate panel market submodel ....................... 34
Market share submodels ................................ 40
Estimation procedures .......... ... ................ .... 45
Methods of estimation ..... ...................... . 45
Completeness and identification .................. 48
Autocorrelation ................. ....... . ......... 49

iii

Multicollinearity .............. ..... . ........ ..... 50

Significance tests ................................ 51
Specification errors .............................. 51

PAST TRENDS IN INNOVATION AND PRODUCT DIFFUSION .................. 53
Product characteristics ..................................... 53
Particleboard characteristics .......................... 54
Southern pine plywood characteristics ....... . .......... 55
Process and product trends .................................. 56
Particleboard process and product trends ............... 56
Southern pine plywood process and product trends .. ..... 63
Product standards and building codes ........... ... .......... 68
Particleboard standards .................. . ............. 69
Plywood standards ...................................... 70
Building codes ..... .................................... 71
Product consumption, price and end-use trends ............... 72
Particleboard consumption, price and end-use trends .... 72

Southern pine plywood consumption, price and end-

use trends ........................................... 76

Raw material situation ................................... ... 79
Particleboard raw material situation ................... 79

Southern pine plywood raw material situation ..... . ..... 85

MODELLING PRODUCT DIFFUSION ...................................... 92
Aggregate panel market submodel ............................. 92
Market share submodels ...................................... 103
Logistic function and Gompertz curve submodels ......... 103
Multivariate diffusion submodels ....................... 108
PROJECTIONS AND NEWER PRODUCT INNOVATIONS ........................ 118

iv

Comparison of logistic function and Gompertz curve

for projecting market shares ....... . ...................... 119
Projections of panel consumption and wood requirements ...... 127
Future particleboard trends ....... . ..................... 134

Future southern pine plywood trends .................... 136

New product innovations ..................................... 138
Waferboard or structural flakeboard .................... 140

Medium density fiberboard .............................. 146

Composite panels ....... . ........ . ...................... 149

Other factors influencing panel diffusion ...... ... .......... 151
APPENDIX: DATA USED IN ECONOMETRIC AND DIFFUSION MODELS ......... 155
REFERENCES CITED ................................................. 202

Table 1.

Table 2.

Table 3.

Table 4.

Table 5

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 11.

LIST OF TABLES

--Projected total wood requirements for particleboard
and southern pine plywood at high, medium, and low
levels of disposable personal income, 1990 and 2000 ...... 2

—-WOod-based panel product market share (diffusion)
and price trends, 1969 and 1979................. ......... 8

--Estimated parameters of the multivariate, temporal
diffusion submodel for particleboard and southern
pine plywood using seemingly unrelated regression
analysis ..... . ..... . ....... . ..... ............... ........ 10

-—Summary of factors influencing particleboard and
southern pine plywood consumption and diffusion.........13

.--New capital expenditures for plant and equipment in

the particleboard (SIC 2492) industry, 1972 and
1977 (in millions of dollars)... ............. . .......... 60

-—U.S. resin consumption by type used in the manufac—
ture of other platen type particleboard, selected
years (in millions of pounds)... ..... ...................62

--New capital expenditures for plant and equipment in
the softwood veneer and plywood (SIC 2436) industry
1972 and 1977 (in millions of dollars) ........ . ......... 65

--Southern pine plywood producers, capacity and plant
startup years by city and state, 1963 to l968.........66-7

--U.S. particleboard production, imports, exports,
apparent consumption, and price index, 1955-1979
(in millions of square feet, 3/4-inch basis).... ...... 73-4

-uPercentage of total particleboard industry produc-
tion by end—use, 1973 (in percent) ....... ....... ........ 75

-Particleboard consumption in manufacturing in the

United States by product group, 1960, 1965, 1970,
and 1976 (in millions of square feet, 3/4-inch basis)...77

vi

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

—-U.S. southern pine plywood production and price
index, 1964-1979 (in millions of square feet,

3/8-inch basis)... .78

--Type and percentage of raw material used by the
particleboard industry, l973........................81
--Delivered costs of resins and total materials,
containers, and supplies cost to the particle-
board (SIC 2492) industry, 1972 and 1977............83
—-Production workers, hours worked, and average
hourly earnings in the particleboard (SIC 2492)
industry, 1972 through 1977.........................84
--Softwood sawtimber net annual growth to annual
removal ratio on forest industry and all come
mercial timberland in the South 1962, 1970 and
1976.00.00.00. ..... 00.... 0.0.0.0... ...... 86

——Comparison between southern yellow pine plywood
production and log consumption, l964-l978.........88-9

--Wood-based panels and selected end-use markets ...... 93

--Estimated linear and log-linear structural
demand equations based on product prices and
income........

--Estimated parameters of linear time trend for
southern pine plywood and particleboard............104

--Estimated paramters of the logistic function
and Gompertz curve for particleboard and

southern pine pIYWOOdOOOOOOOOOOOOO0.0.0.00000000000105
--Estimated structural demand and supply equa-

tions for particleboard............ ................ 109
--Estimated structural demand and supply equa-

tions for southern pine plywood....................110

--Estimated
diffusion

parameters of the multivariate, temporal
submodel for particleboard...............ll4

--Estimated
diffusion

parameters of the multivariate, temporal
submodel for southern pine plywood.......115

--Market share projections based on nonlinear
diffusion functions, 1990 and 2000 (in percent)....120

--Estimated parameters of the logistic function
and Gompertz curve for particleboard, selected
years-0.00.000... ..... cocoon 00.00.123-4

vii

viii

Table

Table

Table

Table

Table

Table

Table

Table

Table

28.

29.

30.

31.

32.

33.

34.

35.

36.

--Estimated parameters of the logistic function
and Gompertz curve for southern pine plywood selected
yearSOOOOOOOOOOO......OOOOOOOOOO0.0......00.000.00.000125-6

--Comparison of particleboard's market share poten-
tial to the last market share achieved in selected
samples..... ..... .....IOOCOOOOOOCO ........ 00.0.000000000128

--Comparison of southern pine plywood's market share
potential to the last market share achieved in
selected samles.........O0..........0.0.0.0000000000000129

--Projections of panel consumption and total wood
requirements for high, medium, and low levels of
disposable personal income, 1990 and 2000...............131

--Comparison of particleboard consumption projections
made in this study with projections by the U.S. Forest
Service (in millions of square feet, 3/4-inch basis.....l32

—-Comparison of southern pine plywood production pro-
jections made in this study with projections by the
U.S. Forest Service (in millions of square feet,
3/8-inch basis).........................................133

--New mills and modernization projects scheduled for
completion in 1979 and announced specific projects

requiring capital expenditures........................141-2

--Medium density fiberboard production, 1975-1979
(in millions of square feel, 3/4-inch basis)............l48

—-Examp1es of products substituting for wood-based
panels in various end-uses.. .............. ..............154

ix

LIST OF FIGURES

Page

Figure l.--Roles of qualitative factors in promoting (+) or
hindering (-) particleboard and southern pine
plywood diffusion over time. . . . . . . . . . . . . .12

Figure 2.--Product life cycle . . . . . . . . . . . . . . . . . .25

Figure 3.-—Submodels requiring exonometric analysis and the
purpose for using the submodels. . . . . . . . . . . .35

Figure 4.--Total panel, non-southern softwood plywood, southern
softwood plywood, and particleboard consumption,
1955-1979. . . . . . . . . . . . . . . . . . . . . .97-8

Figure 5.--Total panel, hardboard, hardwood plywood, insula-
tion board, and medium density fiberboard consump-
tion, 1955-19790 0 o o o o o o o o o o o o o o 0 099-100

Figure 6.--Market share of particleboard as a percent of total
panel consumption, 19.55-1979 o o o o o o o o o o o o 102

Figure 7.-—Market share of southern pine plywood as a percent
of total panel consumption, 1964-1979. . . . . . . . 102

Figure 8.--Logistic function fitted to particleboard market
share data, 1955—1979. . . . . . . . . . . . . . . . 106

Figure 9.--Gompertz curve fitted to particleboard market
share data, 1955—1979. . . . . . . . . . . . . . . . 106

Figure.flL-—Logistic function fitted to southern pine ply-
wood market share data, 1964-1979. . . . . . . . . . 107

Figure llu—43ompertz curve fitted to southern pine plywood
market share data, 1964-1979 . . . . . . . . . . . . 107

Figure 12.--Partic1eboard market share projections based on
the logistic function, 1980-2000 . . . . . . . . . . 121

Figure 13ouParticleboard market projections based on the
Gompertz curve, 1980—2000. . . . . . . . . . . . . . 121

Page
Figure l4.--Southern pine plywood market share projections
based on the logistic function, 1980-2000. . . . . .122

Figure 15.—-Southern pine plywood market share projections
based on the Gompertz curve, 1980-2000 . . . . . . .122

xi

SUMMARY AND CONCLUSIONS

The primary objective of this study is to analyze the effects
of product innovations and diffusion on raw material requirements in
the wood-based panel products industry. In addition, the roles of
economic and other factors in determining the direction and dif-
fusion rate of panel product innovations are examined.

Particleboard and southern pine plywood are the primary panels
analyzed in this study. Total wood requirements for these panels are
projected to increase substantially by the end of this century (see
Table 1). When compared to 1979 levels, particleboard wood require-
ments and the corresponding consumption levels are projected to
increase between 49 and 61 percent by 1990 and between 71 and 93
percent by 2000. Similarly, southern pine plywood wood requirements
and consumption are projected to increase between 33 and 44 percent
by 1990 and between 53 and 72 percent by 2000.

Particleboard and southern pine plywood consumption and wood
requirements were projected based on: (1) an aggregate panel con-
sumption submodel, (2) a logistic function submodel, and (3) average
wood requirements per unit of output. U. S. disposable personal in-
come and time were used as exogenous variables in the submodels. The
projections were based on extrapolation of past diffusion trends and

on expected economic activity.

2

Table 1. Projected total wood requirements for particleboard and
southern pine plywood at high, medium, and low levels of
disposable personal income, 1990 and 2000.

 

 

 

 

Year Income Wood requirements
level Particleboard Southern pine
plywood

1/

(thousand tons)—
1990 High 8710 6452
Medium 8388 6213
Low 8045 5960
2000 High 10464 7703
Medium 9844 7245
Low 9270 6823

1/

-In thousands of tons, dry weight (ovendry)basis for particle-
board and in thousands of tons, based on ovendry mass and volume at
12 percent moisture content for southern pine plywood.

Source: Table 31.

3

The aggregate panel consumption submodel was estimated for
projecting total panel consumption in 1990 and 2000. The model
included all wood-based panels for which consumption and production
data were available and covered all panel end-uses. Secondary time
series data were not available for examining year-by-year consumption
of panels in specific end-uses such as new housing, residential
upkeep and improvement, manufacturing, and new nonresidential con—
struction. This is a limitation of the submodel since penetration
into different end-use markets is likely to vary for a given panel.

Total panel consumption was estimated to be a positive function
of U. S. disposable personal income, an important determinant of wood-
based panel products, demand household formation, and furniture con-
sumption. The submodel fits the long-term trend very well, but
performs poorly during periods of great market fluctuations experi-
enced during the 1972 to 1979 period. The strong reliance of wood—
based panels on the construction industry led to this cyclic pattern
of panel consumption.

The submodel is not structured to project business cycles;
instead, it is used to estimate long-term trends in panel consumption
based on different levels of disposable personal income. The impacts
of changes in family structure, family formation, and possible sub—
stitution of other products for panels are not incorporated in this
univariate submodel.

The diffusion of new products can most readily be seen in terms
of market shares. Market share, that is, a given panel's percent of
the total panel market, was used in this study as a measure of panel

diffusion. Both univariate and multivariate diffusion models were

estimated to provide some quantitative information about panel
diffusion.

Though no two panels will follow the same diffusion path, the
diffusion paths of particleboard and southern pine plywood evinced
S-shaped patterns. This pattern is supported by diffusion theory
which hypothesizes that diffusion follows a sigmoid pattern. That
is, panel diffusion starts slowly, accelerates through an inflection
point, and slows as the process nears completion.

Two nonlinear models, the logistic function and Gompertz curve,
were used to estimate this pattern, but yielded different results
regarding potential market shares and growth rates. Higher potential
market shares and longer periods to grow from 20 to 80 percent of
the potential resulted when the Gompertz curve was estimated. Results
comparing the two models were more consistent for the southern pine
plywood data than for the particleboard data. The growth rate of
southern pine plywood was approximately twice as rapid as the particle—
board rate. For example, the period of time required to grow from
20 to 80 percent of the potential was 7 years for southern pine ply-
wood and 14 years for particleboard based on the logistic function.

Several assumptions are implied when the logistic function and
the Gompertz curve are used to estimate past diffusion trends and
project future trends. First, the functions imply that new products
will gain in market share up to some maximum or potential level.
Second, the proportion of market share already gained is an important
determinant of future levels. Finally, the increase in market share
is a function of supply, demand, and other factors.

These assumptions highlight some of the major limitations of

of simple nonlinear models. Though intuitively a panel will increase
in market share only up to a maximum level, there is no independent
means for estimating that level. Therefore, market share projections
are based on a nonlinear extrapolation of past trends. A number of
factors that will influence the potential are examined in this study;
however, it is not possible to satisfactorily estimate the extent to
which a panel will diffuse in a complex market. Similarly, the models
provide no explanation of factors affecting diffusion.

In order to assess the usefulness of these functions for pro-
jecting trends and potential market shares under different conditions
of data availability, five time periods of varying length for particle-
board and southern pine plywood were fitted with the logistic function
and the Gompertz curve. For most data sets tested, the logistic
function yielded more conservative estimates of market share potential
than did the Gompertz curve. Thus, model selection affects projections.
In all cases where the standard error of the potential was small, over
85 percent of market potential had been reached by the last year of
the data series. Therefore, these diffusion functions should be used
for projections only when the diffusion process is substantially
underway as was the case with particleboard and southern pine plywood.

Despite the shortcomings of the nonlinear functions, they are
useful for studying diffusion. For example, the S-shaped models
represent the diffusion pattern better than simple linear trend models.
In addition, they provide an estimate of potential market share and
a measure of the rate of panel diffusion. As a result, they can be
used to estimate the relative magnitude of future particleboard and

southern pine plywood diffusion.

While the univariate models are useful in depicting temporal
diffusion trends, they are not very helpful in understanding the role
supply and demand factors play in determining panel consumption and
diffusion. The supply and demand factors influencing particleboard
and southern pine plywood diffusion were screened by means of
structural supply and demand equations.

0n the demand side, product price and multifamily (3 or more
units) housing starts were inversely related to quantity demanded.

One and two unit housing starts, mobile home shipments, and furniture
manufacturing exhibited a positive relationship in the particleboard
case. In the southern pine plywood case, western softwood plywood
price and one and two unit housing starts were significant variables
with positive signs; increases in these variables are expected to
lead to increases in southern pine plywood demanded, all things being
equal.

In the particleboard supply equation, adhesives price and wages
had negative signs while lumber production and energy price had posi-
tive signs. The negative signs were expected since increases in input
prices lead to shifts in the supply curve, ceteris paribus. Lumber
production was used as a proxy for raw material (wood furnish) avail-
ability and had the expected positive relationship to quantity supplied.
For the time period analyzed, the net effect of energy prices on the
quantity of particleboard supplied was positive indicating that com-
peting products were adversely affected by price increases to a greater
degree. The most significant supply factor in the southern pine ply-
wood supply equation was mill productivity. As productivity increased,

quantity supplied increased holding other variables constant.

7

The same factors which influence supply and demand for a panel
also influence its diffusion. Table 2 presents data relating market
share and price trend data for major wood-based panels since 1969.
Apparent domestic consumption of western softwood plywood, hardwood
plywood, and insulation board remained relatively stable over 1969
through 1979 period indicating they are mature products. Hardboard,
particleboard, and southern pine plywood consumption more than doubled
over the same period; they are substantially into market share
growth stage and are approaching maturity.

The increased price of western softwood plywood and competition
from newer panel products contributed to its decrease in market share.
Lack of suitable domestic wood supplies, competition from other panels,
and fast growth of newer panel segments led to a decrease in hardwood
plywood's market share. The relative flexibility and greater diverstiy
of end-uses for competing panels accounts for the decreased market
share for insulation board.

Particleboard and hardboard increased in market share and decreased
in relative price over the 1969 through 1979 period. This inverse
relationship corresponds to the expected demand price-quantity rela-
tionship and is further corroboration of the role demand factors can
play in panel diffusion. Southern pine plywood's market share-price
relationship is the same when viewed relative to western softwood
plywood, a substitute in the sheathing market.

Lagged market share variables along with the supply and demand
variables discussed above were tested as explanatory variables in a
temporal multivariate market share model. The lagged market share

variables were included to show the accumulated effects of supply,

xfiwaoma< "mousom

.ooHumomH .xmccfi mowua umoavoua moHunoSEoo Ham ou o>wumaom.l

 

 

\N
.Ucouu mmmfi nwsounu momfi\fl
mcfimmmuoom wcfimmouoma campy
mm m.m mead
OOH N.mH acma memos coaumasmcH
wcflmmouoon wcammouoon wcopH
an q.m memfi
ooH q.¢fi mead ooo3maa woo3wumm
wcflmmmpoma mcfimmouocH ocouH
mm N.mN mnmfi
OOH m.m~ moofi @umononmm
wcflmmouooc wcfimmouocH ozone
me o.¢ mnmfi
OCH N.o momfi wumonoaofiuumm
wcwumSuosam wcfimmouocH wcouh
cm m.o~ mmmfi
col m.os moms ceasefia mafia anmnosom
mcwmmmuocH mcfimmouoon vcmuH
omH m.m~ memH
oofi ~.mm :moma woozkaa poozumom :uoumoz
Apnoea mam xmvcfiv chouu paw unmouoav
\Moofium oumcm noxumz \Mficouu\ummw maxu Hoamm

 

.mmofi cam mcmfi .mwcouu mowua mam AcowmdMMfiov mumsm Dozens uoswoua Hocma ommmnloooz

.N magma

9

demand, and qualitative factors on product diffusion. The seemingly
unrelated regression results presented in Table 3 were not signifi-
cantly different from the ordinary least squares results. The analysis
indicates that past market share was a significant variable in the
diffusion of product innovations. For southern pine plywood and
particleboard, end-use market demand factors were also significant
explanatory variables. Particleboard market share growth was a
positive function of furniture and mobile home market growth. Inter-
estingly, southern pine plywood market share growth was greatest
during periods when housing starts declined.

The spatial aspects of diffusion were examined for southern
pine plywood in relation to western softwood plywood. A model relat-
ing the percent of softwood plywood shipments to 29 Rand-McNally Major
Trading Areas was formulated. As the distance from a Pacific North-
west supply point (Portland, Oregon) to the Major Trading Areas
increased and as the number of southern plywood plants near the Major
Trading Areas increased, the southern pine plywood market share in-
creased. Transportation cost is one of the main factors leading to
these results and to southern pine plywood's regional advantage in
the eastern United States.

A number of qualitative factors that are not amenable to sta-
tistical modelling play an important part in promoting and hindering
product diffusion. Among the more important fattors are: character-
istics of the good, building codes and standards, process innovations,
and raw material price and availability. Other factors such as non-
wood product competition, product promotion, research and development

activities, and government policies, also influence product diffusion,

.mN “Sum QN mMHflmn—w "mQUHDOm

 

10

 

.Ho.m u .3.o .om. u mm .mmmH nwsou£u «omfi Baum mumv co momma moumafiumM\M
.om.~ n .3.9 .mm. H mm. .mmmfi nwsounu «omfi Scum mumv do momma moumafiumM\fl
Hmmwoo. Nomfimo. ucmumcoo
mooooo. mmoooo. I N w H .mupmum wcfimsom loumnm
uoxuma moshaa
Omnooo. oqqqoo.H oumnm uwxuma wowwmg mafia aposusom
oeeooo. NNQNHO. I ucmumcoo
mooooo. mmoooo. mucoaawnm was: oHHnoz
«moooo. «mmooo. wafiuauommscma mouSuwa
cam opsueausm mo woocH
\Ioumnm poxums
ammoko. ommmsk. spasm umxpma umwwmo mumonmauwucmm
a mo
pouuo Aav uaofiofimmooo Axv moanmfium> va oHanum>
oumvamum coamwouwmm uuoumcmaaxm unopaoaoa

 

.mfimhamam scammouwou omumaoua: hawafiaoom wch: woosham mafia aumzunom
one vumonmaoeuuma now Houseman scamsmwwo Hmuoaaou .oumaum>fiuasa osu mo muouoamuma woumEaumm .m manna

11

but are beyond the scope of this study.

There are similarities and differences in these factors and their
roles in the diffusion process. The qualitative and quantitative
factors presented in Table 4 provide a basis for explaining differences
in particleboard and southern pine plywood diffusion. In addition,
the factors provide a framework for examining other new products.

Figure 1 summarizes the effects of qualitative factors on the
diffusion of particleboard and southern pine plywood over the product
life cycle. Notable similarities between the panels were the testing
and modifying of production processes and the large amount of raw
materials potentially available for processing before and during the
product introduction stage. These factors will undoubtedly be
important to future product successes.

Southern pine plywood's rapid diffusion was enhanced by (1) its
similarity to other existing low quality softwood plywood, (2) its
accepted standardization coinciding with initial production, (3) its
low price relative to western softwood plywood, and (4) its regional
advantage. To a large degree, these factors led to southern pine
plywood's broad acceptance in the marketplace. Factors which may
eventually slow its diffusion include (1) almost total reliance on
one end-use, sheathing, (2) rising sawlog costs, and (3) competition
from new structural panels.

Particleboard diffusion was somewhat slower than southern pine
plywood's market share growth. Particleboard's unique properties
have led to wide acceptance over time, but during its introductory
stage the uniqueness was a hindrance. A multitude of end-use appli-

cations and declining relative prices have aided particleboard's

12

mafia cuonusom mam wumonoaofiuumm AIV mcfiumocwz no A+V wcfiuosoua ca muouomm o>aumufiamow mo moHom||.H ouswflm

 

.oEHu uo>o comeMMHw woosmmm

 

 

 

wHHMDH<Z
mem<z

 

A+v
Ai

EH3OMU
mem<z

 

 

ZOHfiobnomHZH
HUDQOMA

 

BHVHS lBMHVW

wooahaa mafia anonusom
wumonmaoauumm
coeumsuflm Hmfiuouma 3mm

oooBhaa mafia auousom
opmonoaowuumm
mwumvamum mam mowoo waflvafiom

voo3haa mafia auonusom
vumonoaowuumm
mcowum>oaca mmoooum

wooskam mafia cuonusom
wumonoaoauumm
mwoom mo mofiumwuouomumao

 

 

13

 

eweuae>ve :oHu
leooH HeSOfiweu .hufi>fiuoe
uexuea emnlwae .mmeoodm

hufi>fluoe uexuea eealoae

muouoem Homes

 

nexuea veueasasoo< .mmeoosm nexuea veueaaaboo< Imam GOHmSMMHQ va
hufi>fiuo=woua
.meofiua undue .hufi>fiuoe >uwafiaeafie>e movemep .meofiua
nexuea emulwce .mooflun uaacfi .hua>fiuoe uexuea omslvso muouoew Howoa
euaufiumnsm .eofiua nonvoum .meoHua ousuaumnnm .eoeum uoswoum waeaew wee haaa5m Amy
waaeafiwfiuo eHQeHHe>e meOum haaecwwfiuo cowuesuam
nonawuaem mo meanao> owned maneawe>e mesvfimeu e>Hmmez Heeueuez Bem on
wafiuesuonam waflafiaoev kaaepecoo eofium Amv
uaoa%eauewas
wcwfiueegm .xooumeuoo eunufiausm mew: >ueawum Aqv
fioueemeu :oueeeou
useaaue>ow one huumsocfi uceacue>ow ode >uumsoafi
”onMHwoa one veuuanH “wefimacoa one weuuomaH meemeooum Amv
coaueofimwmmeao euefiweeaH maowueoflMfivoE Heaveuu mvuewaeum ANV
voo3maa hufiaenw haaecfiwfiuo Heuoepeso
30H pesuo ou uceHe>Hsvm meeuuoaoum enced: nonvoum AHV
wooahaa mafia cuesusom vueoneaofiuuem uouoem

 

.aOHmDMMfiv wce coaumasmcoo woo3kaa mafia cuesusom one vueoneaofiuuea waaocesawcfi euouoem mo hueaesm

.s wanes

14

diffusion.

In general, if a panel has product characteristics similar to
those of an existing product, diffusion occurs more readily. This
phenomenon is due to decreased consumer resistance to a familiar pro-
duct. For example, waferboard, medium density fiberboard and com-
posite panels are more acceptable because particleboard and hardboard
had gained widespread approval prior to their introduction. Eventu-
ally, the physical characteristic limitation of a panel will slow its
diffusion in the total market. That is, it will diffuse only through
end-use markets for which it is suited.

The development of process innovations is important in all stages
of a product's life. Major innovations may lead to the formation of
new industries, and "nuts and bolts" innovations during the market
growth and maturity stages may keep a panel competitive with its
substitutes. Overall, process innovations promote the diffusion of
products.

Formal standards promoted the acceptance of particleboard and
southern pine plywood. They were especially helpful in the case of
southern pine plywood for which standards existed when the first
panels were introduced. More rapid acceptance occurred as a result.
For particleboard, the slow evolution standards initially hindered
its diffusion. In general, early development of standards and code
approval will aid in the diffusion of newer panel products.

The availability and price of wood inputs are also important
supply factors influencing the past direction of the panel products
industry. Large quantities of relatively inexpensive manufacturing

residues spurred the establishment and growth of the particleboard

15
industry. Likewise, the high price and reduced availability of Douglas-
fir peelers contributed to southern pine plywood's establishment and
growth as a panel product. The volume of southern pine sawtimber grow-
ing stock was equally important. Over time, increased cost and decreased
availability of wood inputs will hinder the diffusion of a given panel.
However, if the capital is available for investment, the existence
of large quantities of raw materials can promote the establishment
of panel industries and lead to the introduction of new products.
In conclusion, projections are difficult to make when large amounts
of pertinent information are available and even more difficult when
little is available. One example of a projection by an industry expert
based on few trend data is the following:
...talk about dozens of new plywood plants in the South during
the next decade does not seem justified. Land ownership prob—
lems, availability of raw material on lands dedicated princi-
pally to other products, and the as yet unfelt price competi-
tion from the massive western industry all seem to militate
against over-rapid expansion. A total of perhaps 8 or 10 mills
during the coming decade may be a more realistic estimate
(Fassnacht, 1964).

This projection was surpassed by the end of 1965 when the twelfth plant

started production. Due to the complexity of the diffusion process

and lack of data on the newest products, estimates on the magnitude

of their diffusion would only be speculative. However, some general

statements can be made regarding the direction of product innovation

and its effects on wood requirements based on the review of qualitative

factors.

First, new products will utilize more diverse raw materials and
in greater quantities than has been the case in the past. Waferboard

and composite board are ample evidence of this trend. Roundwood en-

gineered into flakes, particles, and strands by equipment, such as

16
the chipping lathe headrig, is becoming more widespread. Increased
production, particularly of waferboard, indicates new panels can be
competitive even during periods of low market activity.

Second, many products that will play a significant role in satis-
fying panel product demand at the turn of the century are just becom-
ing known or may still be on the drawing board. The gains by particle-
board and southern pine plywood are cases in point. Raw material avail-
ability was important to the success of those products, and presently
underutilized resources may be the key to future successes.

Third, the evolution of standards and the awareness of their im-
portance has opened the door for future product innovations. As stand—
ards become based on performance criteria rather than composition,
new panels will be engineered to compete with existing panels. Stand-
ards being developed by the American Plywood Association are examples
of this type of evolutionary and innovation-promoting standard.

Finally, new product and process innovations expand our resource
base as they diffuse through the market. The new products, in turn,

slow the diffusion of existing, traditional products.

INTRODUCTION

Technological change is a major factor influencing economic growth
and productivity. Technological change and its subsequent diffusion
through an industry can shift supply functions through process innova-
tions and demand functions through product innovations (Bentley, 1970).
The initial effect of a process innovation is a new relationship be-
tween inputs and outputs, whereas, the initial impact of a product
innovation is the substitution of one product for another. Over time,
innovations lead to significant changes in raw material and other input
requirements and extensive substitution of new products in various
end-uses.

The interaction of social, political and economic factors influ-
ences the rate and magnitude of diffusion. In order to get more insight
into the complex process of the diffusion of product innovations, the
relationships between these factors must be analyzed in their historic

context .

The Problem

 

In 1974, Congress enacted the Forest and Rangeland Renewable Re-
sources Planning Act. This act requires the United States Forest Ser-
vice to periodically assess the present and anticipated uses of our
Nation's renewable forest resources. Research Work Unit 4151 at the

United States Forest Service Forest Products Laboratory is responsible

17

18
for projections of timber requirements and trends in harvesting, pro-
cessing, product design and end—use technology of wood products. As
such, they are interested in trends of advancing technology in the
wood-based panel products industries and their effects on the relation-
ships between wood product consumption and timber use. The impact
of product innovation on raw material requirements has not been ade-
quately assessed in the past. The purpose of this research study is
to investigate factors influencing product innovation and diffusion
in the wood-based panel products industries and the effects of these
changes on raw material requirements.

Presently, the Resources Planning and Assessment Staff Unit uses
historical data as a base for resource demand and supply projections.
This implicitly assumes a continuing stream of technological changes
and innovations for the future based on little specified knowledge
of factors affecting the diffusion of innovations in the past (U.S.
Department of Agriculture, Forest Service, 1977). By identifying
forces that have historically influenced the diffusion of innovations,
the forces can be incorporated in future models.

Two major forest product innovations requiring wood inputs are
particleboard and southern softwood plywood. Particleboard was an
obscure panel product in the early 1950's, but over 6.3 billion square
feet (3/8-inch basis) were consumed in 1976 (U.S. Department of Agri-
culture, Forest Service, 1980). Southern pine plywood comprised zero
percent of United States softwood plywood production in 1964; yet,
in 1976 it had captured over 36% of U.S. output (Anderson, 1979). Fac-
tors influencing these product innovations and their diffusion will

be useful in assessing newer panel products such as medium density

19

fiberboard, composite board and waferboard.

Objective of the Study

 

The primary objective of this study is to analyze the effects
of product innovations and diffusion on raw material requirements in
the wood-based panel products industry. The secondary objective is
to hypothesize the future roles economic and other factors will play
in determining the direction and diffusion rates of panel product in-

novations.

Framework for Analysis

 

Product innovation and diffusion are complex phenomena involving
both economic and noneconomic factors (Warner, 1974). Economic factors
include product prices, input prices, substitute prices and end-use
market activity. Product characteristics, standards and building
codes are examples of noneconomic factors.

Due to this complexity, several approaches for examining innova—
tion, diffusion and their effects are useful. First, process innova-
tions, product innovations and end-uses (existing and potential) can
be described. Second, economic, social and physical input trends that
may influence the rate and direction of innovation and diffusion can
be identified and evaluated. Finally, mathematical models can be used
to relate trends in these factors to innovation, diffusion and raw
material requirements.

The first two approaches are qualitative in nature. Major process
innovations can be described and classified as to (1) year of adoption,

(2) adopting company, (3) location, and (4) origin of process. By

20
examining these traits, a greater understanding of innovation and dif-
fusion can be gained. This information is enhanced when coupled with
trend data on inputs, production, prices, product standards and other
important factors.

Though there is presently insufficient knowledge of important
factors to adequately forecast future technologies and their effects
in the wood-based panel products industry, mathematical models can
be used to quantify some effects of existing product innovations. Pro-
duct consumption, price and raw material requirements are some major
effects of concern to forest economists.

The analytical approach most commonly used by economists is to
model supply and demand, separately or simultaneously (McKillop, 1967;
McKillop, Stuart, and Geissler, 1980; Mills and Manthy, 1974). These
models provide estimates relating product consumption and production
to various factors including price, income and population, but they
do not directly yield estimates on the rate of adoption and potential
market share of new products. Econometric models for particleboard
and southern pine plywood supply and demand are used in this study
to corroborate the role of economic variables in product diffusion.

Mathematical diffusion functions, on the other hand, have been
widely used to model the diffusion process. The modified exponential
function, the Gompertz curve, and the logistic function are examples
of diffusion functions used by researchers (Lekvall and Wahlbin, 1973).
The functions are used to fit empirical observations and to make pre-
dictions concerning developments in diffusion processes already underway.

Selection of the appropriate diffusion model or functional form

poses a problem and is decided on the basis of observed data (Maddala,

21
1977). A considerable amount of empirical research supports the use
of the logistic function to describe the diffusion process for new
products. Since observed market share data for particleboard and
southern pine plywood follow an S-shaped pattern, the logistic model
will be utilized in this study because it provides information on the
rate and magnitude of product diffusion. Results based on the Gompertz
Curve are also presented. The models assume that diffusion follows
a sigmoid pattern starting slowly, accelerating through an inflection
point, and slowing as the diffusion process nears completion.

The diffusion models are used to compare the diffusion of particle—
board and southern pine plywood. The purpose in studying these products
is to compare their rates of adoption, potential market shares, and
wood requirements along with differences and similarities of factors
affecting the diffusion process. This information will be used to
hypothesize on the roles these factors will play on other existing

and future wood-based panel products.

Scope of the Study

 

A number of important product innovations have occurred in the
wood-based panel products industry since Werld War II. This study
focuses on two product innovations so that the study will be manageable
and because time series data are available only for a limited number
of innovations. Rather than studying the products of individual firms,
aggregate product forms and their related wood requirements are ana-
lyzed.

Data utilized in this study are primarily from various Census

of Manufactures, Current Industrial Reports, and forest products

22

publications. The product forms studied are particleboard which corre-
sponds with Standard Industrial Classification (SIC) 2492, Particle-
board, and southern softwood plywood which is a component of SIC 2436,
Softwood Veneer and Plywood. The time frames are years 1955 to 1979
for particleboard and 1964 to 1979 for southern pine plywood.

These product forms have shown significant growth in recent years.
The number of plants producing particleboard in North America increased
from 23 in 1955 to 72 in 1979. Particleboard production increased
from an estimated 140 million square feet (3/8-inch basis) to over
7.2 billion square feet during this period. Three plants were responsi—
ble for the production of 80 million square feet (3/8-inch basis) of
southern pine plywood in 1964. Seventy-two southern plywood plants
produced approximately 7.6 billion square feet in 1979. Southern pine
plywood accounted for 41.2 percent of United States softwood plywood

production in 1979.

Definition and Concepts

 

Economic and management literature are replete with definitions
of technology and related concepts. For this study, the following
commonly used definitions will apply. Technology is defined as the
social pool of knowledge of the industrial arts (Schmookler, 1966).
Technological change is regarded as the advance of technology whose
rate is determined by new technology produced in any given period.
Technique is a specific method of producing a good or service; change

of technique is ...the employment of existing but unused technology"
(Manthy, 1974).

An innovation is defined by the National Research Council (1978)

23
as the introduction of new things or methods. Implementation by indus—
try or some other group is required by this definition. In a given
industry, the first enterprise to adopt a new technique is considered
an innovator. Subsequent adopters are imitators. This study focuses
on innovation and subsequent adoption or diffusion.

Technological innovations, which are implemented technological
changes, are classified in three categories: "nuts and bolts" innova-
tions, major technological advances and creation of large new systems
(National Research Council, 1978). "Nuts and bolts" innovations are
exemplified by product differentiation and minor changes in production
processes. Major technological advances, which are examined in this
study, provide opportunities for new processes or products. The cre—
ation of a large new system may or may not be based on a new technology,
but it does require a new combination of techniques in a complex man-
ner.

Another often cited classification distinguishes product innova-
tions from process innovations (Schmookler, 1966). New commercial
products are product innovations, whereas process innovations are new
methods used for creating products. Major process innovations in the
particleboard and southern plywood industries are identified in this
study. Products can be viewed as (1) general product classes (e.g.
wood-based panels), (2) product forms (e.g. particleboard and south-
ern pine plywood), and (3) brands (e.g. GP Particleboard and Plum Creek
Fiberboard) (Kotler, 1976). This study analyzes the diffusion of pro-
duct forms in the wood-based panel products industry. Market share
is analyzed as the indicator of panel diffusion. Hereafter, product

forms will be referred to as products.

23a
Innovation and product diffusion have a variety of effects. For
example, they may cause changes in the nature and quantity of wood
inputs utilized. Other effects include factor and product substitu-
tion, the creation of new resources, and changes in industry struc-

ture.

METHODS OF EXAMINING INNOVATION,

PRODUCT DIFFUSION, AND ITS EFFECTS

In order to assess the effects of product innovations on raw mater—
ial requirements in the wood-based panel products industry, an understand-
ing of the diffusion process and factors influencing successful diffusion
is paramount. Historical examination of interacting factors such as
process innovations, products characteristics, building codes and stand-
ards, raw material situations, and economic data series are useful
in analyzing product innovation and diffusion. In addition, econometric
models can be used to supplement our knowledge and to measure some
effects of product diffusion. Specifically, the models can be used
to estimate future panel consumption, rates of diffusion, market shares,
and wood requirements.

The rationale for examining specific interacting factors is pre-
sented in the following section. Since this is an aggregate analysis,
these factors were selected on the basis of economic theory and a liter-
ature review (Leefers, 1979). The final section of this chapter provides
the basis for the models used and the estimation procedures employed

in this study.

Interacting Factors

 

The interaction of social, political, and economic factors influ—

ences the rate and direction of technological change (Rogers and

23b

24
Shoemaker, 1971). Following the first implementation is the diffusion
of the innovation. In this case, the primary focus is on the diffusion
of particleboard and southern pine plywood, factors influencing their
diffusion, and the effects of diffusion in terms of wood requirements.
The magnitude of diffusion is the primary determinant of wood require-
ments for the new product.

A concept frequently used to describe the path a product follows
over time is the product life cycle (McCarthy, 1975). Figure 2 illus-
trates this concept in terms of sales. There are four major stages
in the life of a product: (1) product introduction, (2) market growth,
(3) market maturity, and (4) sales decline. Introduction is a period
of slow growth and is followed by a stage of more rapid growth. Matur—
ity is a period when sales slow due to widespread acceptance by most
buyers. The final stage marks the decline of product sales (Kotler,
1976). This decline is often caused by new products replacing the
old.

Support for the life cycle concept is based on the hypothesis
that new products must overcome customer reluctance or resistance to
established purchasing patterns (Buzzell, 1966). The idealized S-
shaped product life cycle curve, of course, does not hold for all pro-
ducts, and the period of time for each stage may vary dramatically.

For example, product classes such as wood-based panels may continue
in a growth or mature stage for long periods of time.

Since product innovation and diffusion are closely related to
existing products and markets, it is important to view the introduction
and growth of products in the context of the whole market. New products

that will have a significant long-term effect can be expected to grow

25

 

.eHuNw.eMHH posooum

.N euawfim

 

 

 

 

 

mirh
mzHeomm weHmseaz mezome oneoanomezH
mmaem Hemmez Hemmez economm

 

SBWVS

 

 

 

26
at a faster rate than the market. The diffusion of new products in
the market can most readily be seen in terms of market shares.

The product life cycle can be redefined using market share of
consumption rather than sales as the variable of interest. Any demar—
cation between the various stages is subjective.

For this study, the product introduction stage is the period which
begins with the product innovation and ends when the panel has reached
twenty percent of its potential market share. Though the distinction
between the introductory stage and the market share growth stage is
somewhat arbitrary, one characteristic of the growth stage is the point
of maximum marginal market share growth (inflection point). After
this point, market share growth slows. The mature stage is a period
when market share initially stabilizes indicating that the growth rate
of product form consumption is equal to the growth rate of product
class consumption. The mature stage can also exhibit a decline in
market share with growing or stable product form consumption. Existing
products are expected to exhibit this condition as new products are
introduced. The decline stage is identified as a long period during
which product consumption and market share decline.

The market share associated with a particular product innovation
is determined by the interaction of supply, demand and other forces
along a path to maximum market share (Griliches, 1957). Since the
data available on individual brand diffusion and intrafirm process
adoption are limited, this study analyzes the impact of aggregate vari-

ables on product form diffusion.

27

Characteristics of Goods

 

Traditional demand theory permits economists to analyze the effects
of different preferences on demand, but provides no mechanism for tracing
the effect on demand of changes in physical properties of goods (Lancas-
ter, 1971). Traditional microeconomic analysis begins with a preference
map. At that stage, the characteristics of the goods, other than price,
have already been absorbed into the analytical framework. Though the
characteristics (e.g., weight, color, size, etc.) may change, this
information is rarely used at a later stage in demand theory analysis.

Product variants, model changes, and new goods are difficult to
incorporate into traditional economic models. Either (1) the changes
are ignored or (2) the variant is treated as an entirely new good. In
practice, our ability to investigate physical changes in goods is lim-
ited by the data available on the altered goods.

When changes are ignored, the characteristics of the variant are
implicitly absorbed into the aggregate product grouping. For instance,
graded density particleboard may be viewed simply as particleboard.

In the second case, the variant is treated as a new good which acts

as a substitute for the original product. As a substitute, we explicit-
ly acknowledge that certain comparable attributes or characteristics
link the new with the old. Southern pine plywood, for example, can

be treated as a new good, as a substitute for western softwood plywood,
and not simply as more softwood plywood.

Ultimately, the utility a good will provide establishes the extent
to which it will be used vis-a-vis other goods. This utility, in part,
is closely related to the characteristics and properties of the good.

Theoretically, the demand for new goods could be predicted based on

28
observed behavior of consumers toward existing goods with similar charac-
teristics (Lancaster, 1971). In principle technical data concerning
product attributes will determine the closeness of substitution. Thus,
the characteristics of a particular wood-based panel determine the
range of uses to which it is suited and the extent to which it can
substitute for other products.

Though any good possesses a number of physical properties, not
all properties are relevant to choice between goods. From a practical
standpoint this situation is further exacerbated by the fact that indi-
viduals react differently to specific attributes depending upon their
own preferences. Therefore, in an aggregate analysis it is difficult
to objectively ascertain which characteristics increase the probability
of an innovation's successful diffusion. This is especially true for
products that have many properties and a multitude of potential applica-
tions.

A useful method of incorporating characteristics into traditional
economic analysis is to identify relevant physical properties emphasizing
those properties that are most pertinent to specific end-uses. In
some instances, a product's characteristics may remain unchanged over
time. However, knowledge regarding the product and its characteristics
will increase in a free market as time passes. This increased knowledge

may be due to marketing efforts, word—of—mouth, and other factors.

Process Innovations

 

Within a product class, process innovations can chdnge the produc—
tion function for existing products or serve as the basis for new product

forms and products. The new processes can be developed within or outside

29

the adopting industry and can be imported from other countries. Regard-
less of the origin, process innovations have played a major role in
defining the characterisitics of panels as they exist today.

Many technological innovations have been adopted in the wood-
based panel industry. These innovations range from ”nuts and bolts"
innovations such as using forklifts for loading plywood into railroad
cars (Cour, 1955) to the creation of large new systems such as the
Georgia Pacific's Fordyce, Arkansas, plywood plant (Baldwin, 1977).

The effects of adopting new processes are reflected in prices
of finished goods, raw material utilization, and products available
in the market. As is the case with characteristics of goods, process
innovations are difficult to incorporate in an aggregate analysis.
However, many important process innovations can be identified and their
roles described.

The speed of adoption of new techniques has been analyzed by Mans-
field (1963). His study supports the hypothesis that speed of adoption
is a positive function of firm size and expected returns. In more
concentrated industries, process diffusion generally is more rapid
due to the trialability and observability of the innovation and the
high degree of intrafirm communication, ceteris paribus (Rogers and

Shoemaker, 1971).

Building Codes and Standards

 

"'evolutionary documents'

Building codes and product standards are
that have been amended to accept new products and test methods as they
have been developed (Pease, 1980)." The nature and flexibility of

the codes and standards are factors determining how rapidly new products

30

will gain acceptance. If panels do not have code approval, they cannot
be used for many potential end-uses (Brown, 1979). If standards are
nonexistent, poor quality products will sometimes be produced. The
result is bad publicity and fewer sales (Anonymous, 1966a).

New standards are created to encompass new products, to recognize
more wood species, and to account for changing log quality (Anonymous,
1966b). The standards, in effect, change to accommodate raw material
situations, process and product innovations, and the producers' capabil—

ities.

Economic and Marketing Factors

 

As mentioned previously, supply, demand and other factors influence
the diffusion of new products. Product prices and consumption in various
end-uses represent the interaction of supply and demand over time (Ris-
brudt, 1979). Consumption data can be transformed readily into market
share data once the total market and appropriate measurement units
are defined.

Marketing professionals view the period of growth in product consump-
tion and in knowledge about the product as the initial stages of the
product's life cycle (Kotler, 1976). The marketing mix of product,
price, place and promotion changes as the product moves from the intro-
ductory stage through the growth stage.

The initial message for new products is a description of what
they are and what they do. The characteristics, unknown to potential
users, must be promoted. As knowledge concerning the physical attri-
butes of a product increases, a marketer's activities must be shifted

to other facets of the product mix. As time passes, most potential

31
consumers become aware of a product's characteristics and act according—
ly. The eventual result is a slowing of market share growth.

The availability of distribution channels determines whether or
not products reach potential consumers. Since most new panel products
are developed by companies with existing distribution channels, the
"place" marketing factor is not included explicitly in this analysis.

Panel products can be classified as either industrial goods or
consumer goods depending on their end-use. Unfortunately, most end—
use information is derived from production rather than consumption sta-
tistics (Food and Agriculture Organization of the United Nations, 1976).
The major end-uses in terms of volume for panels are new housing, resi—
dential upkeep and improvement, manufacturing, and new nonresidential
construction (U.S. Department of Agriculture, Forest Service, 1979).
Since panel demand generally is derived from demand for final consumer
goods (e.g., houses, furniture, etc.), changes in end-use consumption
will influence the diffusion of new products. That is, if end-use con-
sumption is growing, market penetration is more likely.

All other things being equal, increases in product prices are ex-
pected to hinder the diffusion of new products. Skimming price poli—
cies (i.e., higher prices) will lead to slower gains in market share
than penetration price politices (i.e., lower prices) (McCarthy, 1975).
Changes in price may be caused by shifts in supply, demand, or both.

The prices of competing products are important factors, too. As
competing products become more expensive relative to new products, mar-
ket penetration is enhanced for new products, ceteris paribus.

Income and population are expected to influence market share growth

since they also affect long—term demand. Disposable personal income

32
is an important determinant of demand for wood-based products and influ—
ences household formation and furniture consumption (U.S. Department
of Agriculture, Forest Service, 1979).

Finally, the availability of products in the market is influenced
by transportation costs. As the distance from the producing region
increases, inplace cost increases. As a result, product innovations
produced in a given region have a competitive advantage over imported

products, ceteris paribus.

Raw Material Situations

 

The raw material situation has always played a key role in the
development of the forest products industry. Timber availability and
depletion historically caused the movement of the logging industry
into the Lake States on to the South and the Pacific Northwest and
now, back to the South.

The panel products industry is also influenced by wood supplies.
In the west, DouglaS"fir is important for plywood production; loblolly,
shortleaf and longleaf pine are the main plywood species in the South;
and an assortment of species in the form of roundwood and residues
is utilized nationwide in the other panels. The availability and price
of these materials will determine, in part, the competitiveness of
the various panels and the direction of future panel developments.

Labor, capital, energy and adhesives are other input factors
affecting product innovations in the wood-based panel industry. The
plywood industry has evolved from labor-intensive operations using
animal glue and wooden presses into a capital-intensive industry with

multi-platen presses and a variety of glues (Cour, 1955). Likewise,

33
particleboard plants have grown in size and complexity requiring new
combinations of inputs. Energy inputs are becoming more expensive,
and particleboard mills are now competing for residues with plants which
may burn the residues for energy production. In total, these factors

will lead to changes in processing and input utilization.

Models
There are many approaches to modelling the demand for an existing
product and its substitution for other products. Some approaches used
are:
(1) Methods based primarily on the autocorrelation properties
of the time series to be forecast,
(2) Methods conceived on the basis of a Markov process which
hypothesizes that demand is a random variable, and
(3) Structural econometric models which attempt to provide causal
explanations for the phenomenon studied (Quandt, 1964).
The first approach includes time series decomposition in which trend,
cyclic and seasonal components are identified. The second approach
involves a brand switching matrix which can be used to calculate market
shares. The last approach is based on economic hypotheses and can be
subjected to statistical testing. This study utilizes the first and
third approaches for analyzing the diffusion or product innovations,
factors influencing diffusion, and the effect of diffusion on wood re-
quirements.
The model hypothesized for this study is composed of three sub-
models: (1) the aggregate panel market submodel, (2) the market share

submodel, and (3) the wood utilization submodel. The aggregate panel

34
market submodel is an econometric model based on relatively gross econ-
omic variables (Mills and Manthy, 1974). The market share submodel
is an extensive examination of trends in new product diffusion. Both
simplistic trend models and multivariate econometric models are used
and compared. The wood utilization submodel is based solely on published
data relating wood inputs to panel output. The two submodels requiring
econometric analysis and the purpose for the models are presented in
Figure 3.
Mathematically, the three submodels can be related to total wood
requirements as follows:
WRi(t) = C(t) x Mi(t) x Ui(t) (1)

where: Wri(t) = wood requirements for panel products i at time t,

C(t) = aggregate panel consumption at time t,

Mi(t) = market share of panel product i at time t, and

Ui(t) = average wood requirement per unit of panel 1 output

at time t.

Theoretically, this model can be expanded by considering panel consump-
tion and market share by end-use category. However, in practice it
is difficult to expand the model because published data are generally
based on production rather than consumption statistics. To the extent

possible, end-use trends are incorporated in the discussion.

Aggregate Panel Market Submodel

The aggregate panel market is equivalent to the general product
class, wood-based panels. Product classes have long life histories
because they are highly population related (Kotler, 1976). Three steps

are required in modelling aggregate panel consumption. First, the

35

Submodel Purposes
Aggregate panel market (1) Estimate past consumption and

test model parameters

(2) Project future consumption

Market share (1) Estimate past market shares and
test model parameters
(2) Estimate past market share
trends
(3) Estimate diffusion rates

(4) Project future market shares

Figure 3. Submodels requiring econometric analysis and the purpose
for using the submodels.

36
extent of the market is determined, that is, which products will be
included in the analysis? Second, a plausible method of aggregating
dissimilar products is developed. Third, an aggregate model based
on the first two steps is hypothesized.

Since the emphasis of this study is on intra-industry competition
between wood—based panels, substitutes such as aluminum, concrete,
plastic, and so on are not explicitly considered. Empirical results
from a study by McKillop, Stuart, and Geissler (1980) support not in—
cluding non-wood structural product prices as explanatory variables
for wood-based panel products.

Two complimentary approaches are used to determine which products
are included in the analysis. One approach is to examine panels in
terms of end-uses; those panels having similar end—uses can be included
in the aggregate model. The second approach is based on economic theory.
A demand function can be estimated in linear or log-linear form, and
the coefficients associated with other panel prices will indicate which
panels are economic substitutes and which are complements.

The demand for a good is a function of tastes and preferences,
the price of the good, income, and the price of the other goods (Green-
wald and Associates, 1973). Since there is no prior knowledge regard—
ing the appropriate functional form for the demand curve, two simple
models were estimated for this study: the linear demand curve and
the constant elasticity demand curve (Hirshleifer, 1976). Estimation
procedures are discussed in the final section of this chapter.

The linear demand function has the following form:

n

P +b21+ZbiPi+e (2)

Y=b +131
0 y i=3

where:

37

Y = quantity of the panel demanded

P&= price of panel demanded

I = income,

Pi = prices of substitute and complement panels

bO = constant

bi = coefficients estimating the change in quantity demanded

associated with a one unit change of the variable, ceteris
paribus

e = unexplained variation of Y.

The coefficients associated with Pi have the same signs as the cross-
elasticities. The coefficient is positive if the good is a substitute
for Y and negative if the good is a complement (Hirshleifer, 1976).

The log-linear demand function is similar to the linear form,
except (1) the variables are the logarithms of variables used in the
linear model and (2) the coefficients, bi’ become estimated elastici—
ties. The elasticities associated with the logarithms of prices of
other goods are cross-elasticities of demand. As such, if the estimated
cross elasticity is positive, the goods are substitutes. If the cross-
elasticity is negative, the goods are complements.

An important factor that must be considered in determining which
panels will comprise the total market is the panel with which comparisons
and statistical analyses are made. If a panel with relatively few
end-uses is selected, the total panel market will be smaller and market
shares for individual product forms will be larger. By selecting a
panel with many end-uses, the opposite results are achieved. As long

as the panels under the study, particleboard and southern pine plywood,

are included in the analysis, the panel used as the basis for comparisons

38
is not critical. This condition exists because the total market can
be disaggregated to individual panels.

Markets can be defined in terms of dollar sales or physical units.
Physical units are used in this study to allow estimation of wood re-
quirements. As mentioned previously, physical consumption data for
all panels in various end-uses are not available over time. Therefore,
the mathematical models used in this study utilize data aggregated
over all end—uses. Supplemental end-use data are analyzed where pos-
sible.

In order to determine the total physical consumption of panel
products, a method of aggregation must be developed. Ideally, the
total market would be defined as the total surface area of substitute
panels consumed in a period of time. However, available data do not
permit this approach.

Two alternative approaches can be used. One approach is to select
a standard thickness for all panels, convert all panel consumption
data to that thickness, and add the new panel data to determine the
total market. This approach was used by the U.S. Forest Service in
determining panel consumption on a 3/8-inch basis (U.S. Department
of Agriculture, Forest Service, 1979). A different approach is used
in this study. Panel production, consumption, and wood utilization
factors are generally reported in standard measures unique to each
panel. These standard measures are closely related to average thick-
nesses used in various end-uses. Therefore, the surface measures with
various standard thicknesses have been aggregated to define the total
market. In using this approach, disaggregation by market share yields

the surface area in terms of conventional standard measures.

39
Having developed a method for selecting and aggregating substitute
panels, an aggregate consumption model was developed. This model was
used to project long-term future panel consumption based on projections
of exogenous variables in the model (Stone and Marcin, 1978). The

general form of the model is as follows:

n
C(t) = a + :a,x, + v (3)
o . 1 1
i=1
where:
C(t) = aggregate panel consumption at time t

xi = exogenous variables that shift demand and supply functions

a = constant

a. = coefficients estimating the change in total panel consumption

associated with a one unit change in the respective exogen-
ous variables, ceteris paribus

v = unexplained variation in C(t).

The most commonly used and available exogenous variables are gross
national product (GNP), disposable personal income (DPI), population
and time. These variables are extremely collinear (average r = .989
for the 1955 to 1979 period) so each one was tested individually as
an explanatory variable. Based on statistical results and its role
in housing and furniture purchase, disposable personal income was se-
lected as the exogenous variable for projecting future panel consump-
tion. Low, medium, and high values for disposable personal income
were used (U.S. Department of Agriculture, Forest ServiCe, 1979).

Though the appropriate functional form for the model is unknown,
Hair (1967) infers that the semi-log form with logarithms of indepen—
dent variables has substantial empirical support. Both the linear

model above and the semi—log model were tested in this study. The

4O
semi-log model was selected for future consumption projections due
to statistical performance and empirical support.
There are several problems with this type of model: (1) the esti-
mated relationship may not be stable over the period of estimation
and the period of projection, and (2) the explanatory variables have
to be projected, too (Maddala, 1977). In addition, there are statis—

tical problems discussed in the last section of this chapter.

Market Share Submodels

 

Four market share submodels were used to estimate the diffusion
of particleboard and southern pine plywood: two were univariate non-
linear models and two were multivariate linear models. The univariate
models, the logistic function and the Gompertz curve, were used for
analyzing the diffusion trend and rate over time. One multivariate
model was used to analyze factors influencing the market share of par-
ticleboard and southern pine plywood over time. The other multivari-
ate model was used to analyze factors affecting the market share of
southern pine plywood over space.

Supply and derived demand structural equations for particleboard
and southern pine plywood were estimated to screen variables used in
the multivariate diffusion models and to isolate factors associated
with the interaction of supply and demand. Since projections of these
variables are not available, only the nonlinear univariate models were
used to project future diffusion of the panel products.

The diffusion of successful product innovations, including most
grades of paper and board, evinces an S-shaped pattern (Warner, 1974;

Hair, 1967). The logistic function and the Gompertz curve have been

41
used to model this diffusion pattern (Griliches, 1957; Buongiorno and
Oliveira, 1977; Mansfield, 1961). The general form for these univari-
ate functions is M = f(t) where M is the market share for a product
innovation and f(t) is a simple nonlinear function of the time that
has elapsed since the introduction of the innovation.
The specific form for the logistic function is:

where:

M = market share

P = potential or maximum market share constant

a = location constant (determines horizontal movement of curve)

b = proportionality constant (determines rate of increase of

market share)

t = time (1955 = 1, 1956 = 2, ..., 1979 = 25)

e = 2.71828, the exponential constant
The function is monotonically increasing and lies between two asymp-
totes, M = 0 and M = P. The function is symmetric with a point of
inflection at P/2 (Oliver, 1964).

The Gompertz curve has the form:

M = P a(bt) (5)
where:

M = market share

P = potential or maximum market share constant

a and b = constants determining the growth rate of market share
t = time (for particleboard: 1955 = 1, 1956 = 2, ..., 1979 = 16)
This curve is asymmetric and approaches P asymptotically. The point

of inflection is at P/e (Lekvall and Wahlbin, 1973).

42

The logistic function is one of the most widely used diffusion
functions in empirical research. Mansfield (1961) developed a strong
theoretical basis for using the logistic function to study the diffu-
sion of innovations. His work has been modified here to show its appli—
cability to market share diffusion.

Mansfield's basic premise is that the proportion of firms that
have not adopted an innovation at time t that will introduce the inno-
vation by time t+1 is a function of (1) the proportion of firms that
already adopted it by time t, (2) various economic variables, and (3)
other unspecified variables. By making a number of restrictive assump-
tions, he argues that the number of firms that have adopted an innova-
tion over time can be approximated by the logistic function.

A number of Mansfield's assumptions can be modified to show their
relevance to diffusion in terms of market shares. First, the function
implies that new products will gain in market share up to some maximum
level. This level can range from zero to one hundred percent of the
market. Statistical estimation procedures, however, do not impose
any restriction on the maximum level.

Second, the proportion of market share already gained is an impor-
tant determinant of future levels. This assumption is based in part
on the role knowledge and learning play in economic growth (Arrow,
1962). As use of the innovation becomes more widespread, uncertainty
concerning the innovation is reduced, thereby increasing adoption of
successful innovations.

Finally, the proportionality constant, which determines the rate
of increase of market share, is a function of supply, demand, and other

factors. The constant also has been defined as the rate of imitation

43
(Mansfield, 1961), the rate of acceptance (Griliches, 1957), and the
proportionate rate of growth (Maddala, 1977). In essence, this assump—
tion acknowledges that a multitude of factors influence acceptance
of new products.

In order to gain insight into the usefulness of the univariate
models under different conditions of data availability, the model para-
meters were estimated and compared for twenty different time periods.
In addition to the model parameters, two additional statistics were
estimated: (1) the inflection point and (2) the number of years required
for the product to increase from 20 percent to 80 percent of potential
market share. The latter statistic was calculated as an indicator
of the diffusion rate, and is equivalent to the market share growth
stage discussed previously.

The multivariate model utilized to estimate the market share trend
over time consists of one identity and two equations. By aggregating
over all end-use markets at a point in time, the sum of the market

shares of substitute products is equal to one, i.e.,

n
t

Z Mi(t) = 1 (6)
i=1
Mi(t) is the percentage of the aggregate wood-based panel market (in
decimal form) of the 1th product at time t and nt is the number of
products in the market at time t. The market share functions each

contain one endogenous variable which is influenced by a group of exo-

genous variables. The functional form was:

n

M1= a0 +§:‘_aixi +111 (7)
i=1
m

M2 = bo + Egibjxj + v2 (8)

44

M1 and M2 = market share of particleboard and southern pine
plywood, respectively
x1 and xj = exogenous variables affecting particleboard and

southern pine plywood, respectively

constants

a and b
o

a. and b structural coefficients estimating the change in
market share for particleboard and southern pine
plywood associated with a one unit change of the vari-
able, ceteris paribus

u1 and v2 = unexplained variation of M1 and M2

The two equation model was estimated as a system of seemingly
unrelated regression equations and by ordinary least squares. Annual
market share data from 1955 through 1979 for particleboard and from

1964 through 1979 for southern pine plywood were used in the model.

Portions of the study period were relatively stable while other portions

were turbulent in nature.

The influence of transportation costs and regional availability

of a product innovation, southern pine plywood, was examined in the

following model:

n
M = C0 + clN + c D + c x + e (9)

2 1;3 i i
where:
M = the percent of all softwood plywood shipments attributed to
southern pine plywood in a Rand-McNally Major Trading Area
(MTA) in 1978
N = the number of southern pine plywood plants in the state con-

taining the MTA and in adjacent states in 1978

D = the distance from Portland, Oregon in miles

45

x additional exogenous variables

1
co = constant
c1, c2 and c1 = structural coefficients

e = unexplained variation of M
Data for twenty-nine Major Trading Areas were included in the analysis.
The sample was comprised of the top 20 softwood plywood markets and
the top 20 southern pine plywood markets. The model provides a method

for examining the spatial diffusion of southern pine plywood in 1978.

Estimation Procedures

 

Several statistical problems must be considered when applying
econometric analysis for estimation, parameter testing and projection.
The appropriate method of estimation, completeness and identification,
autocorrelation, multicollinarity, and specification errors are of

particular concern. These factors must be considered for each submodel.

MethOds of Estimation

 

Estimation methods were determined for five functional relation-
ships: (1) the supply-demand models for determining substitutes and
screening variables, (2) the aggregate panel market submodel, (3) the
univariate nonlinear market share submodels, (4) the multivariate mar-
ket share submodels, and (5) the southern pine plywood spatial diffusion
submodel.

The supply—demand models contain endogenous price and quantity
variables that are simultaneously determined in the market. If ordinary
least squares (OLS) is applied to a single equation in a model and

more than one endogenous variable is included in the equation, OLS

46
estimates will be biased and inconsistent due to correlation between
endogenous variables and the disturbance term. The OLS estimating
technique, however, provides an optimal solution in the case of recur-
sive models (Johnston, 1972).

For more general simultaneous equation models, several estimating
techniques have been developed which yield consistent estimates for
large samples. The techniques are indirect least squares (ILS), two-
stage least-squares (ZSLS), limited-information maximum likelihood
(LIML), three-stage least-squares (3SLS), and full information maximum
likelihood (FIML). The first three techniques are classified as single—
equation methods and the 3SLS and FIML techniques are systems methods.

The systems methods were not considered for this study because
they require large samples to be useful. The sample sizes for the
simultaneous equation models were n = 13 and n = 11 for particle board
and southern pine plywood, respectively. The sample size was restricted
by the availability of comparable product price data.

Both the ILS and the LIML methods require the use of all exogenous
variables in the model. In addition, ILS requires exact identification
(by the order condition) for unique solutions. Due to sample size
and the number of exogenous variables considered in the supply-demand
models, ILS and LIML were not used in this study.

Two-stage least squares (ZSLS) was selected as the method of esti-
mation for the simultaneous equation model because (1) it is a single-
equation method, (2) it provides consistent estimates even with over-
identification, and (3) it is characterized by low bias and small vari—
ance. In any case, the selection of a given estimation technique for

small sample simultaneous equation models is based on results of Monte

47
Carlo studies and personal preference.

The instrumental variable method of ZSLS estimation was used in
this study. When all exogenous variables in the complete model are
used as instruments, the instrumental method is identical to ZSLS.
Valid estimation can be based on fewer than all instruments when a
complete model involves a large number of exogenous variables (Hall
and Hall, 1978). That is, a demand function can be estimated even
when the supply equation is not fully specified.

A number of supply and demand variables were used as instru-
ments in developing the final model specification (see Appendix
for a list of variable names and their values). In order to prevent
the first stage regression from being completely deterministic, a
set of instruments not exceeding the sample size was used in each
equation.

The aggregate market and the southern pine plywood spatial dif-
fusion submodels were estimated by means of ordinary least squares.
This aggregate model included one endogenous variable and one endog-
enous variable and several exogenous variables. Provided the OLS
assumptions were not violated, the estimators will be unbiased and
have minimum.variance for the class of linear estimators.

The Gompertz curve and the logistic function parameters were
estimated by nonlinear least squares. The statistical technique in-
volves the minimization of the sum of squared residuals when the
regression equation is nonlinear in its parameters. Gauss's method
of estimation was used as the procedure for minimization (Hall and
Hall, 1978). An alternative technique was used by Griliches (1957)

to estimate model parameters. However, this technique was not used

48

because it was based on an independent and somewhat arbitrary estimate
of potential market share.

The multivariate market share submodels contain the appropriate
endogenous market share variable and a series of exogenous supply
and demand variables in each equation. Two techniques were used for
estimation: ordinary least squares and seemingly unrelated regression.

The OLS technique assumes implicitly the regression disturbance
in one regression model is not correlated with the disturbance in
another regression model. Since this condition is not explicit, most
market share analyses do not account for the relationship between
error terms (Stern, 1964; Weiss, 1968). The parameter estimates based
on OLS are included in this study for comparison with seemingly unre—
lated regression estimates.

Seemingly unrelated regression, on the other hand, is used in
models that have mutually correlated disturbance terms. This condi-
tion is expected in equations (7) and (8) due to the market share
identity (equation (6)). This subtle link allows estimation to pro-

ceed using the maximum amount of information regarding the model.

Completeness and Identification

 

These concerns apply not only to the supply and demand models esti-
mated by ZSLS. The particleboard and southern pine plywood models
contain two equations (a supply and demand function) and two unknowns
(panel price and quantity). Therefore, the models are statistically
complete.

Apparent consumption data for particleboard and production data

for southern pine plywood were used as the quantity supplied and

49

quantity demanded variables in the functions. The wholesale price
index for the particleboard industry (19678100) and the wholesale
price index for southern softwood plywood (1969=100) were used as
both supply and demand price. Price data was deflated by the whole-
sale price index for all commodities.

The equations in this study are all identified by the rank con-
dition and overidentified by the order condition (Maddala, 1977).
Therefore, the equations are estimable. The instrumental variable
method of estimation requires that at least one instrument (predeter-
mined variable) from another equation be used for each endogenous
variable other than the "dependent" variable. Overidentification

simply makes additional instruments available.

Autocorrelation

 

Autocorrelation among error terms causes several statistical
problems. First, the estimated coefficients are still unbiased, but
no longer have the minimum variance property. Second, the standard
error estimates of the coefficients may underestimate the true standard
error. Finally, tests using the t distribution are no longer strictly
applicable (Neter and Wasserman, 1974).

The Durbin—Watson statistic was used to test for the presence of
autocorrelation. For sample sizes smaller than 15, the test ranges
were extrapolated (Durbin and Watson, 1951). The test was inconclusive
or showed no autocorrelation for all multivariate linear models.

The Durbin-Watson test indicated autocorrelation in the univar—
iate aggregate panel submodel and in the univariate nonlinear models.

This was expected as the omission of one or more key variables in

50
business and economic regression equations often leads to positively
autocorrelated error terms. These equations, when used for projections,
will produce inefficient estimates (Johnson, 1971). However, this
is not a problem in this study since projection confidence intervals

are not used.

Multicollinearity

 

Though predetermined variables may be highly correlated, this
does not inhibit obtaining a good fit to the data. Multicollinearity
has several effects on statistical estimation. There is a loss in
the precision of estimation, that is, it is difficult to disentangle
the effects of the collinear variables. Also, coefficient estimates
are very sensitive to particular sets of sample data.

Multicollinearity always exists in economic variables. The degree
of effects is the major concern. Two methods were used to reduce the
apparent effects of multicollinearity in the models. First, if two
variables were extremely collinear (simple correlation between the
two of .99 or greater), one was dropped from the study and the other
was retained to represent the pooled effect of the collinear variables.
In the case of somewhat equivalent variables such as housing starts
and residential construction expenditures, only one variable was re—
tained in the final model though both were retained for test runs.
Second, extensive test runs were used to examine the addition and dele-
tion of variables for different sample sizes.

Even with this two step process, the selection of the final equa-
tion variables was based on judgment. This was especially true in

the case of the supply and demand equations.

51

Significance Tests

 

Variables were tested for significance using a one-tailed t-test.
The test indicates the probability level at which the estimated coeffi-
cient sign is its true sign. For the aggregate panel market submodel,
the Cochrane-Orcutt procedure (Maddala, 1977) was used for re-estimating
and testing parameters after the presence of autocorrelation was indi-
cated by the Durbin—Watson statistics. This procedure was used to
statistically correct for first—order autocorrelation. After correction,

all parameters were significant at the 1% alpha level.

Specification Errors

 

Specification errors can result from: (1) omission of relevant
explanatory variables, (2) disregard for qualitative changes in vari-
ables, (3) inclusion of irrelevant variables, (4) inappropriate mathe-
matical form used in estimation, and (5) incorrect assumptions concern-
ing the disturbance terms in the regression (Kmenta, 1971). Initially,
a large number of variables were selected based on economic theory
and a literature survey. Many variables were later dropped due to
high collinearity and the results of test runs.

The unavailability of data on important variables, such as, wood
residue price for particleboard, would have some effects on final model
parameters. When possible proxy variables were used to account for
the hypothesized relationship between the dependent and excluded vari-
ables.

In an aggregate analysis, qualitative changes are ignored. In
the case of wood-based panels, the lack of available data on panel

quality changes inhibits incorporating changes into the analysis as

52
an independent factor.

A number of functional forms are used in the analysis. Justifi—
cation of particular mathematical forms was presented on a case by
case basis in previous sections of this chapter. Standard assumptions
regarding the disturbance term were made with the exception of the
seemingly unrelated regression. In that case, correlation between

the disturbance terms in the individual equations was assumed.

PAST TRENDS IN INNOVATION

AND PRODUCT DIFFUSION

Qualitative and quantitative factors interact as new products
diffuse through the market. By examining past trends in these factors,
some insight into their relationship to successful diffusion is gained.
Many of the factors are intertwined; however, they are presented sep-
arately to emphasize their importance.

Particleboard and southern pine plywood are successful product
innovations. They have gained widespread acceptance in end-use markets.
Their product characteristics play a large role in this acceptance.

In addition, product standards, building codes, process innovations,
raw material price and availability, and various economic factors have

been instrumental in this diffusion process.

Product Characteristics

 

Panel products possess three general characteristics: physical
properties, strength properties, and working properties. Physical
properties include density, moisture content, water absorption, and
fire resistance. Examples of strength properties are bending strength
perpendicular to the plane, deflection under load, impact resistance,
and screw holding. Accuracy to size, machining properties and surface
quality are the primary working properties (Akers, 1966).

Property comparisons of different panels are made with consideration

53

54

of end-use requirements. Due to its properties, particleboard is used
extensively in furniture and construction, whereas, southern pine ply—

wood is used only in construction.

Particleboard Characteristics

 

Particleboard is:
A generic term used to describe panel products made from
discrete particles of wood or other ligno-cellulose material.
Other materials can be added to the production process to
improve the board. Thermosetting resins are added to the
particles to serve as a binder. The particles are bound into
a solid board when the particles and resins are placed under
heat and pressure (Dickerhoof and McKeever, 1979).
Particleboard is produced in a variety of densities and thicknesses.
For brevity, only the general panel characteristics are presented here.

Particleboard is a popular core panel in furniture because it
has excellent gluing qualities on all planes and superior working prop-
erties. The surface quality is especially important due to the increas-
ing use of thin veneers as laminates. The panel surface is uniform
and the panels stay flat making it a good product for furniture manufac-
turers (U.S. Department of Agriculture, Forest Products Laboratory,
1974).

Particleboard characteristics are well suited for construction,
too. After manufacture, the board is constant and uniform in its prop—
erties. Its use as floor underlayment is attributed, in part, to its
impact resistance and flatness. However, particleboard is more suscep-
tible to damage by excessively damp and wet conditions than plywood
panels. The ability of manufacturers to create large panels enhances

particleboard's acceptance as mobile home and prefabricated house deck-

ing.

55

In general, the flexibility of manufacturing processes have per—
mitted particleboard to be tailored to specific end-use characteristics.
A key factor in particleboard's future success will be the continued

modification of its characteristics.

Southern Pine Plywood Characteristics

 

Plywood is defined as:
...a flat panel built up of sheets of veneer called plys,
united under pressure by a bonding agent to create a panel
with an adhesive bond between plys as strong as or stronger
than the wood. Plywood is constructed of an odd number of
layers with grain of adjacent layers perpendicular. Layers
may consist of a single ply or two or more plys laminated
with grain direction oriented parallel to the long dimension
of the panel. The odd number of layers with alternating grain
direction equalizes strains, prevents splitting, and minimizes
dimensional change and working of the panel (U.S. Department
of Commerce, National Bureau of Stands, 1974).
The primary species utilized in southern pine plywood are loblolly,
longleaf, and shortleaf pine. Slash and spruce pine are used to a
lesser extent.
Plywood characteristics vary considerably. The veneer grades
used as plys determine the quality of the final product. CD sheathing
is the mainstay of the southern pine plywood industry. However, plyform,
underlayment, 303 Siding, and AC/Ad panel production is increasing
(Baldwin, 1977).
Structural plywood panels such as CD sheathing are engineered
for end—uses where tension, compression, shear, cross-panel flexural
properties and nail bearing are important (U.S. Department of Commerce,
National Bureau of Standards, 1974). Grade C face veneer restrictions

are well suited to southern pine species since small knots, discolora-

tion, and synthetic repairs are permitted. Grade D back veneers have

56

fewer restrictions.

Process and Product Trends
Processes adopted by the U.S. panel industry often are imported
from other countries. As the number of processes increases, products
become more varied and specialized. The result is greater acceptance

of wood-based panels in end-use applications.

Particleboard Process and Product Trends

Particleboard was developed in this country to utilize waste wood
and to compete with established wood products (Reid, 1958). The indus-
try has grown as new processes and products have emerged. The major
processes used in the manufacture of particleboard are the extrusion
process and the platen press process. Product characteristics vary
according to the process used.

Early attempts to produce particleboard in the United States during
the 1920's failed primarily due to the lack of suitable adhesives.
During the 1930's and early 1940's, several breakthroughs in resin
production and application enabled particleboard production to proceed.
The resin shortages during World War II gave way to reindustrializa-
tion following the war; the particleboard industry has grown drama-
tically since that time (Moslemi, 1974). A number of types of particle-
board and production techniques have been developed over the post-
war years. The alternative production methods often were developed
to avoid infringement on Swiss and German patents (Akers, 1966).

Akers (1966) identified four general types of particleboard: (1)

single layer, platen pressed, (2) three layer platen pressed,

57

(3) graded density, platen pressed, and (4) extruded board. These
boards can be produced in a variety of thicknesses and in a range of
densities. The final product is determined primarily by market require-
ments.

The single—layer, platen pressed board was the first type developed.
The boards were homogeneous and utilized planer shavings, industrial
wood residues, and flax shives as raw materials. The particles are
in a random criss-cross pattern parallel to the surface of the board.
Early single layer boards were produced in batches. The first domestic
continuous press process was designed by Ralph Chapman and installed
in the Chapwood plant at Philomath, Oregon in 1957 (Anonymous, 1957a).

The three-layer board is similar to the single—layer type, but
has a sandwich construction with courser materials placed between flakes
or other material for smoother surfaces. This process allows for greater
bending strength and was first used in furniture as coreboard.

Fred Fahrni, a Swiss engineer, developed the first 3—layer type
of board which became known as Novoply in the United States. The United
States Plywood Corporation began producing Novoply in 1951 at their
Anderson, California, plant. Edwin Behr developed a similar board
in West Germany for use in furniture. The Behr process was first intro-
duced to the United States in 1957 by Roddiscraft, Inc. at Arcata,
California (Anonymous, 1957c). Though it was the first of its type
in the U.S., twenty-one similar plants had been built previously world-
wide.

The graded density board, originally manufactured in 1958 by Baehre
Metallwerke in West Germany, was designed with a high density surface

and a low density core. Finer particles are near the surface, and

58
unlike three-layer boards there are no abrupt changes between particle
sizes. The boards are used in furniture and building construction.
The Bahre—Bison process is used widely in the United States. Edwin
Behr also developed a graded density board.

Other platen press particleboard manufacturing systems adopted
in the formative years of the industry were designed by major companies,
such as Industrial Development, Miller Hofft, and Columbia Engineering
and by various smaller companies (Mottet, 1962). The processes were
usually tailored to the raw material available and to specific end-
use needs.

Extruded boards are made by forcing the wood and resin mix through
two parallel metal platens. The particles are perpendicular to the
surface of the board. As a result, the board has low bending strength
and is used as a corestock for furniture.

The three extrusion processes most used in this country are: (1)
the Kreibaum process, (2) the Lanewood process, and (3) the Chipcraft
process (Mottet, 1962). The Kreibaum vertical extrusion process was
developed by Otto Kreibaum in West Germany and first used in the United
States by the Jasper American Manufacturing Company at Henderson, Ken—
tucky in 1954. The Lanewood process for horizontal extrusion was in-
stalled initially by the Lane Company at Alta Vista, Virginia in 1953
(Anonymous, 1957b). One of the earliest Chipcraft horizontal extrusion
process was installed by Berkline Corporation, Mbrristown, Tennessee,
in 1953 (Anonymous, 1957b).

Three recent developments in product or process design are wafer-
board, oriented strand particleboard, and Mende process particleboard.

These newer panels are discussed in final chapter along with newer

59
process innovations. The production of new panels is another indica-
tion of continuing innovation in this industry.

The raw materials for particleboard are industrial wood residues,
forest residues, roundwood, resins and additives. The wood materials
are prepared by hammermill or flaker, dryed and screened, mixed with
adhesives and additives, formed into boards, and pressed. In addition
to major process innovation and diffusion, evidenced by new production
systems, improvements in many aspects of the production process have
occurred. Recent capital expenditures presented in Table 5 indicate
that refinement in processes rather than construction of new plants
is occurring in the particleboard industry.

The traditional method for preparing particles is to use an assort-
ment of attrition mills, hammermills, chippers, ring flakers and drum
flakers. Refinements in this equipment and in screening continues
to improve the uniformity of materials used in manufacturing (Food
and Agriculture Organization of the United Nations, 1976). The shaping
lathe headrig and "fingerling" production are among the newest develop—
ments in flake preparation (McSwain, 1977).

Particle dryers have increased in size and efficiency. In some
cases, wood residues are being utilized as an energy source for drying.
Particle moisture, once measured by hand sampling, is now monitored
and controlled by infrared spectrometers and devices measuring electri-
cal resistance. After drying, the particles are stored or blended
with resins and additives. Short—retention-time blenders have replaced
many long-retention-time blenders over the past 15 years (McSwain,
1977).

Urea-formaldehyde and phenol-formaldehyde are the primary resins

60

Table 5. New capital expenditures for plant and equipment in the
particleboard (SIC 2492) industry, 1972 and 1977 (in millions
of dollars).

 

 

 

Year Capital Expenditures
New buildings and New machinery
other structures and equipment
1972 $2.7 $32.1
1977 3.2 29.3

 

Sources: U.S. Department of Commerce, Census of Manufactures, 1972,

1975b; U.S. Department of Commerce, Census of Manufactures,
1977, 1980b.

61
blended with the wood particles. Interior particleboard is made with
urea-formaldehyde resins and exterior particleboard is bonded primarily
with phenol-formaldehyde. Both are water soluble and thermosetting.
Urea-formaldehyde is used more extensively in the industry according
to available data (see Table 6). Improvements in adhesive cure time
and tack control have occurred over the years. Experimentation is
increasing with alternative binders such as isocyanates, cement, mala—
mine based resins, and tannin formaldehyde adhesives as industry concern
for resin cost and availability continues (Coppens, Santana, and Pastore,
1980; McSwain, 1977).

Additives are also blended in with the adhesive and wood particles.
Common additives are rosin and/or paraffin emulsion as water repellents,
phenol formaldehyde for added strength, penta-chlorophenol and arsenical
compounds for fungus and insect protection, and borate for fire retar-
dancy (Akers, 1966). By changing the additive mix, new board character-
istics can be attained.

The next stage in the production process, board forming and trans-
port, has undergone several changes as the industry has evolved. The
improvements are concentrated in platen press systems which account
for the majority of particleboard produced in the United States. Some
of the most notable advances in this area are the increased size and
sophistication of matforming equipment, the development of fiber allign-
ment systems, and the introduction of caulless transport systems. These
innovations allow for a broader range of wood inputs and create product
variants for end-use markets (Food and Agriculture Organization of
the United Nations, 1976; McSwain, 1977).

Platen-type and continuous prepresses have been developed for

‘ 62

Table 6. U.S. resin consumption by type used in the manufacture of
other platen type particleboard, selected years (in millions
of pounds).

 

 

 

Year Quantity of resin consumedl/
Urea Phenol formaldehyde
formaldehyde and other types
1966 262.0 6.2
1969 631.7 25.0
1972 1215.2 22.6
1975 1025.6 11.6

 

-l/Dry basis prior to 1969; wet basis beginning in 1969.

Source: U.S. Department of Commerce, Bureau of the Census, selected
years(a).

63
caulless transport systems. The prepresses reduce the maximum press
opening needed which leads to faster press closing. The result is
a reduction in surface precuring, increased core density, and better
internal bonds. Particleboard presses are either continuous types
or discontinuous single-opening or multi—opening types. Modifica-
tions in single-opening presses have improved their competitive posi-
tion in recent years, but multi-opening presses are still dominant.
Shorter press cycles, larger board sizes, and better thickness control
are the main advantages of single-opening presses. The development
of continuous presses for thin (Mende) particleboard production is
another example of recent improvements in board pressing (MCSwain,
1977).

Board finishing and secondary processing is the last step in the
production process. Notable advances in this processing stage include
the extensive adoption of wide-belt sanders and improvements in lami~
nating and curing. These advances lead to broader acceptance of par—

ticleboard due to increased quality.

Southern Pine Plywood Process and Product Trends

 

The southern pine plywood success story is based on adopting tech—
niques for peeling small diameter logs, developing suitable resins,
and utilizing substantial southern timber resources. The small sized
low quality peeler logs are used primarily for construction grade ply-
wood.

A substantial number of process innovations have been incorpor-
ated in the rapidly growing southern pine plywood industry. Many of

these have been installed during plant construction. Recent trends

64
in nationwide capital expenditures are presented in Table 7. Approxi—
mately $26 million of the 1972 total expenditures and $45 million of
the 1977 total occurred in the South (U.S. Department of Commerce,
Census of Manufactures, 1972, 1975a; U.S. Department of Commerce, Cen-
sus of Manufactures, 1977, 1980a).

Before the first southern pine plywood plant was opened in Decem—
ber, 1963, considerable industrial research had taken place. Promis-
ing yield trials of southern pine logs in western mills provided an
impetus for interest in the South (Fassnacht, 1965). Though consider-
able government and industry experimentation in an attempt to find
suitable adhesives was ongoing, Georgia Pacific Corporation's develop—
ment of the first adhesive capable of satisfactorily laminating south—
ern pine was a major breakthrough (Koch, 1972 and Anonymous, 1980d).
With the adoption of high-speed automatic lathes, the southern plywood
boom was underway; thirty-four plants began operation in the first
five years of the industry's existence. These plant startups indicate
which companies were innovators and early imitators (see Table 8).

The production process begins in the log yards used to store and
sort logs in the South. Huge jib cranes often are used to stack and
move the logs (Baldwin, 1977). Once inside a plant, the logs are
peeled with automated, high-speed lathes. Only rotary-peeled veneer
is produced (Koch, 1972). Back-up rolls, high-speed log charging,
and geometric chuck centering increase the efficiency of the process
(Food and Agriculture Organization of the United Nations, 1976; McSwain,
1977).

Southern pine veneer is mostly sapwood. It is cut from logs

averaging 14 inches in diameter with 4 to 6 inches left as core. Jet

65

 

 

 

 

Table 7. New capital expenditures for plant and equipment in the
softwood veneer and plywood (SIC 2436) industry, 1972 and
1977 (in millions of dollars).
Year Capital Expenditures
New buildings and New machinery
other structures and equipment
1972 $ 9.0 $56.5
1977 10.6 95.1
Sources: U.S. Department of Commerce, Census of Manufactures, 1972,

19753; U.S. Department of Commerce, Census of Manufactures,
1977, 1980a.

66

 

coo4 om m2 .e444>e4:o4 ofiweoemlefiwuoeo
ocoH om m2 .ucoaseem .auoo voo3h4m eaHm eu4ea
oom4 on n2 .hufio exoaooom .ou woo3>4m hem oxeeaeeeso
oom4 004 <4 .aeHuOHM .oo voosk4m ue>sooce>
coma won <4 .cmeHOm coaaaeeoneoosa4m .m.:
oo¢4 co <4 .v4e4wca43 .ocH .umepx:440
oom4 ow <4 .comwoa .oo voo3z4m ecefimfino4
@004 cc <4 .sepcwz .auoo cumnunomlefinSDHOU
cooH om <0 .neae>em ofiwfioemlefiwuoeu
mom4 oo4 <> .eeuoaam oem4oemue4wuoeu
moa4 Nu x9 .cfist4 mHocHHHHImcm3O
mcm4 ow oz .nusoaz4m .ou nemaeesuexez
mess on we .eane4mee44nm .oo tomameetmsmz
mom4 004 <4 .m4emxeo .ou woo3h4m um>sooce>
mo¢4 mm <4 .COOmsm cuenusomaaewucem
moo4 om4 m< .Ne uuemeouo 04m4oemne4wuoeo
mc¢4 om4 m< .4e uuemmOHU onHoemlekuoew
mo¢4 co 4< .GOuadm .oo penas4 nouoom
ecoss4m .m.:

«oaH om x9 .440oen \.oo .un4 ween neeausom
«oo4 co x9 .ee4m44m .ou nonas4 hnufim
moo4 0N4 m< .eomeuom owmwoemlewwuoeu
new»

mauueum lkuwoemeo GOHueoo4 woodwoum

\a

 

ou mom4

.mueum wee hufio >4 mueo% anuueum uce4a use .huwoeaeo .mumoapoua

.wom4

mooBm4a mafia cuesusom .w eHneH

67

.wmo4 .esoakcoc< "meadow

.mHmen socelw\m .ueem eueavm 5044445 :4 weusmeea mueoeaeo.l

\4

 

wom4
wom4
wcm4
w©o4
wom4

noa4
mom4
n©a4
nom4
n©m4
n©m4
mom4

©©m4

oo
oo
oo
om
oo4

co
mo
om
oo
oo
00
0a

Om

Um .e444>4emmnm
<4 .wcHHmea aHmHm
<u .emouo%e3

4< .ceEaeno

4< .m444m eawm

oz .eusocoz

m2 .ueum04o

<4 .momom

<4 .meno0uficouez
<0 .mwaHHQm ueceo
4m .wce4me440

4m .ece4me4su

02 a MHHH>COm¥UMW

U4440emle4wuoeo
muosvoum ummuom hcosuc<
.auoo aEeo .m.:

.auou aEeo GO4GD
4evo04m ce4442oez

.auou mooB>4m e4wcewuH
ofiwfioemle4uuoeo

.oo ue4634 ucoeeua
.ocH .%4mlnunom
e003>4m auezuuoz ueeuo
ofimfioemlewwuoeu

.muoo 4e048e£u e4xu<

.oo Heeseenummez

A.e.ucoov w «Heme

68
dryers are the predominant method for drying the plys (Koch, 1972).

Experimentation with various dryer configurations (baffle arrangements,
temperature settings, drying times, and so on) continues in the South.
However, drying is still a major bottleneck in the production process
(Food and Agriculture Organization of the United Nations, 1976).

Curtain coaters, spray lines, and soft roll spreaders are used
for glue waste while cutting down on manpower requirements. Most
southern pine plywood is glued with phenol-formaldehyde resins (Food
and Agriculture Organization of the United Nations, 1976; McSwain,
1977).

The mechanization of the production process has increased in
recent years. One example of mechanization was the advent of auto-
matic layup systems in 1968 (Mast and Pease, 1969). These systems
provide an efficient means for continuous production (DeLess, 1975).
Large plant size is one result of extensive automation. Faster cure
time, automatic charging, and automatic discharging equipment are
examples of increased mechanization in pressing. Other improvements
include better blister detection, new patching and plugging materials,
and more extensive use of wide-belt sanders and improved abrasives

(Food and Agriculture Organization of the United Nations, 1976).

Product Standards and Building Codes

 

Product characteristics and product standards are tied to building
codes and panel end-uses. Panel standards generally fall into two
categories: general descriptive type standards and specific end—use
standards (Frashour, 1961). The former are developed to encompass

most producing companies, whereas the latter are more exacting and

69

restrictive. Government Commercial Standards and Voluntary Product
Standards are examples of general descriptive type standards. Typical
end—use standards are FHA (Federal Housing Administration) Interim
Standards, Industry Association Standards, and ASTM (American Society
for Testing and Materials) Standards.

The Voluntary Product Standard (previously called the Commercial
Standard) is voluntary, however, grademarking based on the standard
is believed to promote product acceptance in the market (Anonymous,
1971b). The purpose of the Commercial Standard is "...to establish
quality criteria, standard methods of testing, rating, certification,
and labeling of manufactured commodities (U.S. Department of Commerce,

National Bureau of Standards, 1966a)."

Particleboard Standards

 

Particleboard standards evolved from the 1956 FHA Interim Standard
for Floor Underlayment, Kitchen Countertops to be Overlaid with Plastics,
and Wardrobe Doors to the Commercial Standard 08236-66, Mat—Formed
Wood Particle Board, issued in 1966.

A number of standards for specific end-uses were developed between
1956 and 1966. Particleboard standards for specific end-uses have
been promoted by the National Particleboard Association (NPA). Examples
of these standards are: NPA-l, Standard for Particleboard for Mobile
Home Decking; NPA-2, Standard for Particleboard Decking for Factory
Built Housing; and NPA-3, Specification for Particleboard Floor Under-
layment Coated with Wax-Polymer Type Hot-Melt Coatings (U.S. Department
of Commerce, National Bureau of Standards, 1977).

The 1966 Standard contains information on interior and exterior

70

panels including panel form, materials, dimensions, dimensional
tolerances, and properties. Inspection, test methods, marking, and
certification are described in the commercial standard, too. The
property requirements for mat-formed particleboard are based on type

of use (interior or exterior) and panel density (U.S. Department of
Commerce, National Bureau of Standards, 1966a). C8236-66 was recently
updated as ANSI (American National Standards Institute) A208.l
(Anonymous, 1980c). The new standard consolidated existing particle-
board standards such as ANSI Al6l.l for kitchen cabinets and ANSI Al6l.2

for countertops.

Plywood Standards

 

Standards applicable to southern softwood plywood have existed
throughout the short life of the product. The first Commercial Standard
was C5259-63, Southern Softwood Plywood, which was in existence when
the first panels were produced. This undoubtedly paved the way for
more rapid acceptance of this new panel. 08259-63 was combined with
Commercial Standards for Douglas Fir Plywood and Western Softwood
Plywood in 1966. Product Standard PS 1-66, Softwood Plywood, Construc-
tion and Industrial, was the result (U.S. Department of Commerce, Nation-
al Bureau of Standards, 1966b). PS 1-66 was updated and re-issued
in 1974 as Voluntary Product Standard PS l—74, Construction and
Industrial Plywood.

PS 1-74 covers "...the wood species, veneer grading, glue bonds,
panel construction and workmanship, dimensions and tolerances, marking,
moisture content, and packing of plywood intended for construction

and industrial uses (U.S. Department of Commerce, National Bureau of

71

Standards, 1974). Test methods are also included. Plywoods are
classified by their exposure capability (interior or exterior) and
grade. In general, higher grades have more restrictions on the veneer
quality. The primary panel grade for southern pine plywood is interior
CD sheathing with exterior glue. Interior, intermediate and exterior
glues can be used with interior type plywood.

Since they provide consistency, standards are desired by pro-
ducers, distributors, specifiers, architects, home builders and con-
tractors. The standards often are adopted in model, state, and city

building codes.

BuildinggCodes

 

Building codes are the legal minimum criteria which must be met
in new construction, alteration and modification of buildings.
According to a 1976 study (Vogel, 1976), over 7000 building codes
existed in the United States. The study covered three Model Building
Codes, twenty State Building Codes, and thirty City Building Codes.
CSZ36-66 and PS 1-74 were included in the study. The particleboard
standard was referenced in two Model Building Codes (the 1975 Basic
Building Code and the 1973 Standard Building CodeL,nine State Build-
ing Codes, and six City Building Codes. The plywood standard was
referenced in flrasame Model Building Codes, eleven State Building Codes,
and seventeen City Building Codes. As expected, the more traditional
product, plywood, had gained wider acceptance.

Building codes can prevent or allow the utilization of new pro-
ducts in construction. Therefore, acceptance of product standards

by building regulatory agencies is necessary for the widespread

72
diffusion of panel innovations. Ventre (1979) studied the adoption
of innovations in residential construction. He found that the regula—
tory agencies acted as technological gatekeepers and in the aggregate
adopted innovations slowly at first, then more rapidly until adoption
finally leveled off. As more regulatory agencies accept innovations,
adoption by the construction industry is facilitated and product dif-

fusion is increased.

Product Consumption, Price and End-Use Trends

 

Total consumption and price trends reflect the interaction of
supply and demand over time. An important factor in determining total
panel product demand is economic activity in end-use markets. One
indication of a panel's competitiveness in a given market is its abil-

ity to penetrate the market successfully.

Particleboard Consumption, Price and End-Use Trends

 

Particleboard consumption in the United States increased from
70 million square feet (3/4—inch basis) in 1955 to 3586 million square
feet in 1979 (see Table 9). Available price data indicates that the
wholesale price of particleboard relative to other goods has declined
since 1967. Rising costs for transportation, energy, and adhesives
along with increased competition for particleboard furnish may reverse
this trend in the future.

The declining price is one impetus for the increased consumption
of particleboard; particleboard's diverse uses is another (see Table
10). Particleboard is used primarily as core stock, floor underlayment,

and mobile home decking. For 1979, 48.4 percent of the particleboard

73

 

 

q.c¢ quN cm 54 mmmN mnm4
o.mm mmmN N44 N «o4m qnm4
m.w© qmqm mm N4 «men mmm4
w.No owom me «4 544m Nnm4
m.No Nme cN w «mmN 4nm4
«.mo mmm4 04 m «054 ohm4
m.mo4 «454 «4 N4 @454 mc¢4
m.w¢ oNq4 o 4 qu4 wom4
c.004 «444 N 4 m444 mom4
I mom I 4 mom oom4
I now I q mow mom4
I oqo I N wmo wom4
I wmq I 4 omq mom4
I woe I I woe Noo4
I on I I on 4oo4
I wON I I woN com4
I omN I I cmN ama4
I OmN I I omN wmm4
I mw4 I I mw4 nmm4
I 444 I I 444 omm4
I ON I I On mmm4
Acosueeo4v
.Ixoec4
\MUHHQ
m4eme4os3 .Icowugasmcoo cowuonwoua
meueHMem \N uceueaa< wupoaxm muuoaaH o4umeaon new»
..lAwfimen noc4Iq\m .uoem whence mo mc044448 SHV mmm4
Immo4 .erC4 eoflpa wee .504uaanmcoo oceueamm .muuoaxe .eupOQEH .a04uosvoua vweone4owuuea .m.D .m o44ea

74

.mo4om4ueum none4 mo seeusm .pone4 mo Samsoueeeo .m.: mom?

.mom4 .mm4enm use unw443 ”enema peuoe4ee

.mUHBOGOUm 4e4uumswcH mo deeusm .euuee

I500 40 uaeEuueamQ .m.D ”va whee» veuoe4mm .mSmceo ecu mo seeuam .moueEEou mo useEuuemen .m.:
”AUVmuemz wmuue4mm .msmceo ezu mo deeusm .moumasou mo uaeaupeamn .m.D mnn¢4 .ma4e4m mow¢4 .uum48e4 “meousom

.msmceo emu mo deepen >4 3e4>eu peecD.l

\m

.u4m4e3 c4 weuuoaeu euev anomafi Eoum meueefiumm.l

\q

.kea4 o“ “04ta e4ae44e>e Soc Aoo4ueea4v mmaoaeoeeOu 44m toe xmece

e04pa e4eme4043 man an weue4mee Aoo4umom4v xemcfl eowua e4eee4043 usauso %pum5@¢4 vueone4owuuem.l

 

\m

.muuomxe I muuoaafi + CO4uosmoua oflumeEoe u :04uaasmcoo uaeuema<\m

.mueon emeooua mace: use wueonue444 muHecev Bowmee ev24054 uoc meoa\w
o.Nm nwmm oo4 \m¢w4 40mm mnm4
m.mo «wen m¢4 \m¢m4 040m wma4
N.qm Neon mm \wcm4 Nmmm nmm4
w.qm mm4m om \m4m Non omo4

4.e.uc00v a e4aee

75

Table 10. Percentage of total particleboard industry production by
end-use, 1973 (in percent).

 

 

Intended end-use Percent of total
industry
Production
Floor underlayment 27
Furniture core 27
Mobile home decking 15
General purpose 10

core stock

General purpose 8
stock, furniture
core, cabinets

Cabinets 4

Door core 4

Siding 1
1/

Factory-built -—
housing, decking

Other 4

 

1/

—-Less than 0.5 percent.

Source: Dickerhoof, 1976.

76
market was for industrial grades, 34.6 percent was for floor underlay-
ment, 8.6 percent was for mobile home decking, 3.6 percent was for
door core, and miscellaneous uses comprised the remaining 4.8 percent
(Lambert, 1980). Activity in these end-use markets determines the
particleboard product mix to a large extent.

Changing styles in the furniture industry have influenced particle-
board consumption. As the trend toward cleanline furniture continues,
particleboard's properties become more desirable (Wilson, 1975). The
furniture production index, which measures physical output, increased
185 percent from 1960 to 1976. Particleboard consumption in furniture
increased 1533 percent during the same period (see Table 11). Though
not as dramatic, increases in average per unit consumption of particle—
board has also occurred in the housing and mobile home markets (Dicker—

hoof, 1978; Wilson, 1975).

Southern Pine Plywood Consumption, Price and End-Use Trends

 

Southern pine plywood production increased from 80 million square
feet (3/8-inch basis) to 7650 million square feet over the 1964 to
1979 period (see Table 12). Most production is consumed domestically.
For example, less than seven-tenths of one percent of 1978 shipments
were exported (Anderson, 1979).

The deflated price of southern pine plywood has fluctuated in
response to residential construction activity in recent years. During
slack periods in construction, panel prices decline. As activity in-
creases, prices rebound. Though stumpage prices are rising in the
South, proximity to the major markets in the Eastern United States will

continue to enhance southern pine plywood's in-place cost advantage

77

.Ow©~ .w0fi>hmm ummhom .m.D umUHDOm

.wueon euwmoaaou use .mueoeexe4m .wueonuemeB .cueonuenflw zuemcemlaswmma mevs4ocH.l

 

 

 

\4
N.Nm4 o4m4 oo4 ooq 0mm omm4
4.wc4 ocm ow omN com Omm4
4.mm one we @44 N4m mom4
w.4n 004 N4 em om oom4
Aoo4nmcm4v eusu4cusm
xewcfi muoswoua 4ea04usu4umc4 eunufisusm
coauoswoua penuo vce 4e404eaaoo w4osemsom
ensuficusm 4eu0H msouu nonwoum Meow

 

.4m4mmn euaaIs\m .umme museum no maoe44aa see eem4 eae .okma

.moa4 .ocm4 .asouw acetone >4 meueum ueuHGD enu :4 NG4usuoemaceB c4 :OHumasecoo \flcneone4owuuem .44 e44eH

78

Table 12. U.S. southern pine plywood production and price index,
1964-1979 (in millions of square feet, 3/8-inch basis).

 

 

Year Production Deflated
wholesalel/
price index—
(1969=100)
1964 80 _
1965 402 -
1966 1140 -
1967 1779 -
1968 2373 -
1969 2875 100.0
1970 3315 81.1
1971 4410 89.8
1972 5319 102.8
1973 5560 105.1
1974 5130 84.0
1975 5676 80.4
1976 6814 101.0
1977 7447 116.0
1978 7898 115.2
1979 7650 94.0

 

-l/Southern pine plywood wholesale price index (1969=100) deflated
by the wholesale price index for all commodities (1969=100); not
available prior to December, 1968.

Source: Anderson, 1979; U.S. Department of Labor, Bureau of Labor
Statistics, selected years.

79
relative to Douglas fir plywood (Holley, 1969).

Unlike particleboard, southern pine plywood production is almost
exclusively for CD sheathing. The product mix for southern region
shipments in 1978 was: 91 percent sheathing, 5 percent sanded, and
4 percent specialties (Anderson, 1979). The majority of shipments
were to the southern region. According to one source (Baldwin, 1977),
the product mix will shift to higher value added products, such as
plyform, underlayment, and siding, after mill operations stabilize
with efficient production of sheathing. However, until that time,
consumption will be predominantly in the form of sheathing in the hous-

ing and mobile home markets.

Raw Material Situation

 

The price and availability of inputs into the productive process
influences the competitiveness of any product. The inputs can be in
the form of wood, adhesives, labor, energy, or capital. As price in-
creases or as availability becomes restricted, there will be a response

in the price and production level of the panel product.

Particleboard Raw Material Situation

 

As mentioned previously, one factor that spurred the growth of
the particleboard industry was the availability of residues from wood
processing plants. Examples of wood inputs used in particleboard are:
planer shavings, plywood residues, furniture and veneer manufacturing
residues, and millwork residues.

In 1956, approximately 50 percent of particleboard wood inputs

came from sawmills. Veneer plants and furniture plants accounted for

80
30 and 14 percent of the inputs, respectively, with other wood proces-
sors supplying the remaining 6 percent (Reid, 1958). The situation
has changed since 1956; the sources of material now include pulp-type
chips and roundwood (see Table 13). Planer shavings were the primary
residue used in the West and South in 1973, and roundwood was the most
important source in the North (Dickerhoof and McKeever, 1979). More
recent data indicates planer shavings are declining in importance rela-
tive to other residues (Food and Agriculture Organization of the United
Nations, 1976; U.S. Department of Commerce, Census of Manufactures,
1977, 1980).

The availability of wood furnish is dependent on productive acti-
vity in supply industries. Also, the trend toward non-residue sources
is due to: (1) new sawmill technology that improves sawlog utiliza-
tion, (2) new pulp and paper industry technology for utilizing resi-
dues in production, and (3) increased use of residues as an energy
source (Wilson, 1975). In addition, other composition board products
are competing for many of the same raw materials.

Though a large portion of wood inputs are now purchased, there
is no satisfactory measure for the price trend in wood furnish for
particleboard. However, as industry is forced to utilize more non-
residue furnish, the furnish cost will increase considerably (Wilson,
1975).

In order to estimate wood requirements, some measure of physical
input to product output is needed. According to data collected by
Dickerhoof (1976), 1.51 tons of wood raw material (dry weight basis)
were used to produce one thousand square feet (3/4-inch basis) of par-

ticleboard on average during 1973. Approximately 88 percent of the

81

Table 13. Type and percentage of raw material used by the particle—
board industry, 1973.

 

 

Type of Percent used
raw by
material total industry
Planer shavings 65
Plywood mill waste 10
Sawdust 9
Roundwood 7
Chips (pulp—type) 5
Slabs, edgings, and 3
trimmings
Veneer core 1]
Other 1

 

l/Less than 0.5 percent.

Source: Dickerhoof, 1976.

82

furnish was softwood. By assuming the input mix does not change appre-
ciably, this factor can be used to estimate wood requirements given
panel consumption. Several measures have been developed for converting
material requirements to cubic foot measures, but they are not used
in this study (Risbrudt, 1980; Wright and Phelps, 1967).

As mentioned earlier, the end of resin shortages after World War
II made more resins available to manufacturers. The price of adhesives
and resins relative to all commodities (1967=100) was 100.2 in 1979
(see Appendix). However, this does not reflect the trend that has
persisted since 1972 which shows adhesive prices have increased 34
percent more than all commodities. The delivered cost of resins in-
creased at the same rate as the cost of all materials used in the par-
ticleboard industry between 1972 and 1977 (see Table 14). Continued
materials cost increases will exert upward pressures on particleboard
prices.

Average hourly earnings of production workers did not increase
as rapidly as adhesive prices from 1972 through 1979 (see Table 15).
The number of production workers and the total hours they work is deter-
mined by economic activity in end-use markets and the degree of mechan-
ization in the industry. Particleboard output per production hour
worked (3/4-inch basis) was 226 square feet in 1972 and 330 square
feet in 1977; labor-saving process innovations being adopted by in-
dustry are responsible for part of this increase.

Energy costs have risen sharply in recent years. As the cost
of energy enters in at each stage of wood harvesting and processing,
final product prices are expected to rise. However, energy also influ-

ences the cost of competing materials. Therefore, the net effect of

83

.nowm4 .mnm4 .meueuuemeeez mo memeeo .moeefifioo mo uemEueeemQ .m.D “mousom

 

o.m4N m.o4 4.5m nmm4

q.wm4m N.m w 4.Nmm Nmm4

 

Amue44oe c0444eev

 

umoo mm44meem Amee44ow e044448v Amee44ow e04444EV
wee .muoefieueoo meHmeu wfioe eeu meemeu
.m4e4eoue8 eeeuo wee 044oeeem eefiae4ea wee ewe:
4euOH umoo weee>44en new»

 

.eeme use Name .ANmsN oemv enumaeae eemoe
Ie4o4ueee one On umoo me44eaem wee .memeHeueoo .m4e4eouea 4eu0u wee meemmu mo mumoo weum>44mn .q4 o4neH

.4owa4 .mnm4 .meeeuoeweeez mo memeeo .eoeeEEoo mo uemaueeeea .m.D “moeeom

 

84

 

mm.m m.o4 o.m Nmm4

mo.m m.m o.c emm4

oo.q ¢.w m.q m>o4

oo.q w.N4 4.0 cum4

Nm.N w.m4 m.w m>m4

mm.mm w.m4 «.0 NN¢4
Awee44owv meexeo3 Ameow44fiev Aooo4v
eowuoewoue meexeos meexeos

mo mwefieeee eOHuoswoea >4 eOHuoewoue
>4unoe eweeo>< wexuoa mueom mo penaez uee>

 

.Nmo4 eweousu Nma4 .ANoeN UHmv
>uumewe4 wue04e4oflupee mew e4 ewewewee >4eeoe emeee>e wee .wmxu03 mueoe .muoxuo3 eowuoswoum .m4 m44eH

85
the rising energy cost is difficult to determine (Wisdom, 1977).

In the short run, rising energy prices will have two effects on
the particleboard industry. First, the industry will attempt to reduce
consumption of purchased energy. This is borne out by available data
on purchased energy per unit of particleboard output and by industry
efforts to become more energy self-sufficient (Egan, 1980; TenWolde,
1979). Second, other wood processors will begin competing for residues
as a source of energy (Wilson, 1975). Though competition will cause
residue prices to rise, significant quantities of unused manufacturing
residues still exist. For example, there were 547 thousand cubic feet
(solid basis) of unused manufacturing residues in 1976 while only 330
thousand cubic feet of particleboard furnish were consumed (U.S. Depart-
ment of Agriculture, Forest Service, 1979).

As presented earlier, recent trends in capital expenditures have
been primarily in the area of new machinery and equipment. However,
considerable capital has been invested in the form of new plants over

the years; 72 plants were operating in 1979 as compared to 23 in 1955.

Southern Pine Plywood Raw Material Situation

 

As was the case with particleboard, raw material availability
has played an important role in the growth of the southern pine plywood
industry. In the South, the net volume of softwood sawtimber on commer-
cial timberlands increased 39 percent from 1962 to 1977 despite the
growth of the plywood industry (U.S. Department of Agriculture, Forest
Service, 1979). Volume increases were recorded in every diameter class.
Net softwood sawtimber growth to removal ratios, however, have

been declining, especially on forest industry lands (see Table 16).

86

Table 16. Softwood sawtimber net annual growth to annual removal
ratio on forest industry and all commercial timberland

in the South, 1962, 1970, and 1976.

 

 

 

Year Net growth to removals
Forest All southern
industry commercial
lands timberlands
1962 2.14 1.65
1970 1.29 1.42
1976 .91 1.28

 

Source: U.S. Department of Agriculture, Forest Service, 1979.

87
The result may be increased competition for softwood sawtimber on other
lands and increased stumpage costs. In fact, average 1978 stumpage
prices (in 1967 dollars) for southern pine sawtimber sold from National
Forests in the Southern Region were 33 percent higher than prices paid
in 1969 (U.S. Department of Agriculture, Forest Service, 1979). South-
ern softwood plywood prices in 1978 were only 15 percent higher than
the 1969 level.

Table 17 presents Bureau of the Census data on log consumption
for plywood in the South. The data are based on local log scales and
are not strictly comparable over a long period of timecl/ In general,
there is no discernable trend in the log requirements per unit of out-
put. By assuming an average specific gravity of .51 for southern soft-
wood species and an additional 5 percent of veneer volume that will
be compressed into a 3/8-inch panel, approximately 1,168 pounds of
wood (ovendry mass and volume at 12 percent moisture content) are re-
quired per thousand square feet (3/8-inch basis) of southern softwood
plywood (U.S. Department of Agriculture, Forest Products Laboratory,
1974).

Bolt diameter, core diameter, and processing efficiency will deter-
mine wood requirements in terms of softwood bolts. As bolt diameter
decreases and core diameter increases, more bolts are needed to pro-
duct 1,000 square feet of 3/8-inch plywood (Koch, 1972). By maintain-
ing a stable bolt diameter and reducing core diameter, greater yields

per bolt and improved lathe productivity are possible.

 

1/

—-Comparability problems resulted in discontinuing this data
series in 1979 (Ambler, 1980).

88

 

 

mom. 44m.mq>.N qwm.wmq.> mnm4
mom. mmN.an.N oww.¢w>.w o>m4
mmm. w>w.omm.4 an.wmq.m mmm4
4mm. emo.me.4 emm.m4m.q e>m4
mom. «M4.qoo.N mwo.>me.m mnm4
«mm. www.mqm.4 4¢m.ooN.m N504
mom. emq.mmm.4 owm.N4m.q 4mm4
Nmm. mow.qMN.4 oqw.m4m.m Oma4
mum. mw>.mmo.4 «mo.Now.N oom4
Nwm. mqo.mmw mw0.mqm.N wom4
4mm. moo.Nmo mmm.mo>.4 nom4
4mm. om4.omq 40>.mmo.4 wwm4
mum. 4mm.4q4 moo.mnm m©m4
04uee Ae4eom w04 AmHmen noeHIw\m .umem
eOHuoewoum On .ueew weemsoeuv eeeevm weemeoeuv
eOHDQEemeoo mo4 eowuaaemeoo wo4 coeuoewoea wooB>4m pee>
.wnm4Iewm4 .e04ueaemeou w04 wee e04uoewoua wooB>4e eefim 3044e> eeenueom eeeaumn eomfieeeaoo .m4 044eH

89

.Anvmeee> weuoe4em .msmemo men 40 eeeeem .eoeeEBoo wo ueeEuueeen .m.D "moeeom

 

4mm.

qu.w>m.N 0>4.mm>.n w>m4

4.e.ucoov 44 e4nee

90

The net effects of increases in phenol-formaldehyde resin prices,
labor costs, and energy prices on the consumption of plywood are diffi-
cult to determine a priori. While southern pine plywood utilizes more
expensive resins, resin content per panel is lower in plywood than
in composite panels. On the other hand, Douglas fir plywood requires
less expensive urea—formaldehyde resins to obtain comparable interior
grade qualities; therefore, southern pine plywood is at a cost disad-
vantage in that respect. Lower transportation costs lessen the effects
of the resin cost differential though.

Historically labor costs in the South have been below the national
average. The annual average rate of increase in weekly earnings, how-
ever is greater than the national average (Irland, 1972). Over time,
this will tend to reduce southern labor cost advantages. Labor wages
are influenced by productivity. In the veneer and plywood industry,
output per production hour worked has been above the average for all
manufacturing. Process innovations have contributed to this trend
(Farris, 1978). Hence, productivity gains may offset rising labor
costs.

The effect of increases in energy price on the southern pine ply-
wood industry differs from the effects on the particleboard industry.
For example, residues are now being used as fuel for dryers reducing
dependence on purchased energy as is the case with particleboard. Com—
petition for wood raw materials, though does not increase as a result.

By 1979, seventy-two softwood plywood plants were operating in
the South. This represents a huge influx of capital into an industry
that began in late 1963. This capital has created large, efficient

mills with modern processing equipment. Innovations continue to be

91

incorporated in new plants as they are built. As a result, the south-
ern plywood industry competes successfully with other panel products

industries.

MODELLING PRODUCT DIFFUSION

Prior to analyzing diffusion in terms of market shares, the pro-
ducts to be aggregated were selected. First, general wood—based panel
end-use markets were determined for panels with available consumption
or production data. Second, structural demand equations were estimated
for particleboard. Finally, panels were aggregated to form the total

panel market.

Aggregate Panel Market Submodel

 

Table 18 presents an overview of panels and their end-use markets.
In many cases panels can act as substitutes for one another, but they
can also be complements and inputs into other panel products. For
example, hardboard and particleboard are substitutes as floor underlay-
ment. Insulation board and hardwood plywood can be complements in
the sense that a wall system may have a finished interior wall hardwood
panel backed with insulation board sheathing. An example of a panel
as an input, is the use of particleboard as a core material in plywood
sheets. Thus, the relationships among panels are complex.

Since particleboard is one of the panels being examined in this
study and because it has numerous end-uses, it was selected initially
as the panel on which to base the total panel market. Table 19 presents
the estimated parameters of the linear and log-linear structural demand

equations. As expected, the positive signs associated with hardwood

92

.nmo4 .>oeeeo mowo4 .eomemwe< "meoeeom

 

93

.oeeuHeuew
.Ame4w4mv eoeuoeeumeoo 4e4ueewwmee 3oz weeonee44w >u4meew Eefiwez

.Amwooe .m44e3v eOHuoeeumeoo 4e4ueew4mee 3oz weeon eOHue4eee4

.menufieeew mmeOHunwe wee
me4eeee .eeexe: 4e4ueew4moe “Am44e3
.me004mV e04uoeuumeoo 4e4ueew4mme 3oz wueoeweem

.eeeuweeem mme04u4wwe wee eeHeeee
.eeexee 4e4ueew4mee unwe4>4eem
.me004mv e04uoeeumeoo 4e4ueew4mee 3oz weeone404ueem

.eeeu4eeem ”meowufiwwe
wee me4eaee .eeexee 4e4ueew4mee

"Am44e3v e04uoeuueeoo 4e4ueew4mee 3oz w003>4e woo3weem

.me04u

Iwae wee meeeeee .eeexe: 4e4ueew4mee
mAmMOOM .m44e3v e04uoeeueeou 4e4ueew4mee 3oz wooa>4e ween eeeeueom

.e04uoeeumeoo 4e4ueew4meeIeoe “mumxeee

4e4eueewe4 meeeu4eeem mmeowuewwe

wee meHeeee .eeexe: 4e4ueew4mee

”Awe4>4eem .m44e3 .mmooe .weHme
.me004wv eowuoeeumeoo 4e4ueew4mmu 3oz wooB>4e e44 ee4weoa

 

muexeea emeIwem 4eeem

 

.muexeea emelwee wouue4em wee m4eeea womenlwooz .w4 e44eH

94

 

 

e0w NN e44eH mom .mmm4 emeoenu mwo4 Eoem euew SD43 weuuem meowueevm . omen eo ANV e04ueeve eem\fl
co.n o~.mI ueeumeou
am. om. Aqxv e4
em. Ho.N Amxv :4
mm. mq.4 ANNV e4
a... N3- 25 a E 5 \mamaees as
04.4mMN ow.NqomI ueeumeou
Aexv esooefi
o4.N aw.m 4eeomeea e44emoemww .m.D
Amxv eo4ee
no.e am.em ecossea eoosoeom
ANxv e04ee
cw.q4 mm.mm wooS>4n4 woozweem
A>v >e4ueeev
em.w wN.emI Alec manna eeeone4oeneem eeeopm4ueueee \Meemane 41V
a Any Axe Awe
mo eoeee ueefiowmmeoo me44e4ee> e4ne4ee>
weeweeum eoemmeemem >eoueee4exm ueeweeeea eOHueevm

 

leeooew

wee meofiea uoewoee eo women meOHueeve weeEew 4eeeuoeeum eeee44lw04 wee eeee44 wmweawumm .m4 e44e9

95

.emem uoeeeoo eeu we4>ee mo >u4444e40ea peeoeee om ee umee4 ue e>e4 mueeHUHMMeoo wmueawuee 44< “euoz

BA .35 .8. MI

II
M

.309 aqao

aq.N

.mueeaeuumeH >4eaem

A.w.ueoov m4 e44eH

96

and softwood plywood price indicate that they are net substitutes for
particleboard. Hardboard and insulation board prices were included
in test runs of the model, but were deleted because of (1) their col—
linearity with each other, (2) hardboard's collinearity with hardwood
plywood, and (3) little additional variance being explained by their
inclusion.

In order to accomodate the entire range of end-uses and the com-
plexity of product substitution, all panels listed in Table 14 were
included in the total market. Hence, the substitutability of hardboard
and medium density fiberboard for particleboard in furniture as well as
the use of insulation board for sheathing can be enveloped in the model.

Apparent consumption data were developed for particleboard, hard-
board, insulation board, and plywood other than southern pine plywood.
For southern pine plywood and medium density fiberboard, production
data were used in the model. This poses few problems since southern
pine plywood is primarily a domestic product at this time and medium
density fiberboard comprises a minute portion of the total market.
Hereafter, both production and consumption data will be generalized
as consumption data.

Figures 4 and 5 present consumption data for the panels in the
total market. The total market is growing over time but is very cyclic
in nature. This is due primarily to the strong reliance of wood-based
panels on the construction industry. Non-southern softwood plywood
which is comprised mostly of Douglas-fir plywood is a mature product
maintaining a high level of the total consumption. It is also charac—
terized by a declining market share. Hardwood plywood and insulation

board are also mature products with declining market shares.

97

Figure 4.——Total panel, non-southern softwood plywood, southern
softwood plywood, and particleboard consumtion, 1955-1979.

 

 

Legend:

 

Totoal panel consumption (standard measures)

 

Non—southern softwood plywood consumption (3/8—inch basis)
Southern pine plywood consumption (3/8—inch basis) ——--—-——-——-—————.

Particle board consumption (3/4—inch basis) ——————————————————

Figure 4.

98

 

 

I
1976 1980

1970

.1965
YEHR

 

1960

1955

 

l
0000?

 

T

I 00008 ‘00003 I 00001
(311 N01111u1 NellaunSNoa

°Isso

 

 

99

Figure 5.-—Total panel, hardboard, hardwood plywood, insulation
board, and medium density fiberboard consumption, 1955—1979.

 

Legend:

 

Total panel consumption (standard measures)

 

Hardboard consumption (l/8-inch basis)
Hardwood plywood consumption (3/8-inch basis) ——————————————
Insulation board consumption (l/2—inch basis) ——-——-——————__.__.___

Medium density fiberboard consumption (3/4-inch basis) ——————————

100

Figure 5.

 

 

 

o
co
'2
L 1' i“
\ 1 E-
\ 1 1'2
I “m
/ i 7
\ J .
\l‘l :o
\‘.I —§
Xi: :
l _,, a:
1 ~15
1" - >-
h
\ .
I -D
I'l '3
r ...-a
1‘ .
'17
I1 '3
a:
[1.
L.
o
to
I ' I ' I ' I 4 ‘33
00007 00008 00003 00001 0

(31:1 NOI'I'IIH] NOIldelSNOIJ

 

 

 

101

Southern pine plywood, particleboard, hardboard, and medium den-
sity fiberboard are increasing in terms of consumption and market shares.
The first three panels are substantially into the growth stage while
medium density fiberboard is probably in its introductory stage. The
diffusion of particleboard and southern pine plywood through the market
is illustrated in Figure 6 and 7, respectively.

An aggregate panel consumption submodel was estimated with ordinary
least squares using total panel consumption as the independent variables.

The final form of the model was:

C = -184025 + 32220.1 x ln(DPI) R2 = .95, D.W. = 1.21
(9496.55) (1464.30)
where:
C = total panel consumption (standard measures)
DPI = U.S. disposable personal income in billions of 1972 dollars

Since the presence of autocorrelation is indicated by the Durbin—Watson
statistic at the 5 percent alpha level, the equation was re-estimated
using the Cochrane-Orcutt technique. Using the one—tailed t-test,
both coefficients were significant at the .0005 alpha level. Estimated
coefficients changed relatively little so the above model was retained
for projection purposes.

The model fits the long-term trend data very well, but performs
poorly during periods of great market fluctuations. This was evident
when estimates for total panel consumption were compared to actual

consumption during the 1972 to 1975 period.

102

 

 

HRRKET SHRRE iPERCENT)

l

i

Q
I B
a .1 .

 

 

1050

Y Y‘V’

1 ff 7 V ' V I I V I V V V V I V V V ‘ I
1980 1005 1070 [075 1000

YEHR

. .4r .
1955

 

Figure 6.-Market share of,particleboard as a percent of

total pgnel consumption, 1955-1979.

 

HHRKET SHRRE (PERCENT)

 

 

 

31
0
124 0"°
4}
0
9: °
0
0’
In- ‘0
6
‘ad-w-r-HF-‘L I . A. . v , v r .4 I r . arij
1900 1908 1070 1078 ' 1000
YERR

 

Figure 7.--Market share of southern pine plywood as a
percent of total panel consumption, 1964-1979.

 

 

103

Market Share Submodels

 

Four market share submodels were estimated to examine different
aspects of the diffusion. The logistic function and Gompertz curve
submodels were used to estimate the time trend in market share dif-
fusion, the growth rate of panel diffusion, and the inflection points.
These trend models were used to project future market shares in the
next chapter. Two multivariate models were estimated to examine

temporal and spatial aspects of diffusion.

Logistic Function and Gompertz Curve Submodels

 

The market share data in Figures 5 and 6 were fitted with a simple
linear trend model using ordinary least squares and with the nonlinear
submodels using nonlinear least squares. The regression results are
reported in Tables 20 and 21. The fitted nonlinear functions are
illustrated in Figures 8 through 11.

Several general conclusions can be derived from the regression
results. First, the S-shaped models represent the diffusion process
better than the linear models. Second, the logistic function is
slightly superior to the Gompertz curve for the particleboard data,
and the opposite is true in the case of the southern pine plywood.
Third, as espected given the model formulation, the potential market
share (P) and the period of time (growth rate) required for a panel
to grow from 20 percent to 80 percent of its potential are greatest
for the Gompertz curve. Finally, the growth rate of southern pine
plywood was approximately twice as rapid as the particleboard rate. In
fact, the logistic function proportionality constant (b) which determines
the diffusion rate is substantially greater for southern pine plywood

than for particleboard.

104

.N.>4 weueemeeeme m4 eeecm uexees e4 eoeuewee> wee4e4axeeb .mN n m>m4 ....
.N u cmm4 .4 u mmm4 me weeemeee m4 e849 .> + Aeefiuv m + < u oeeem uexeez ”weeeu mafia neme44

 

 

 

\N
.eemeeueeeea e4 meoeee weeweeum \fl
4Nooo.v Amoo.v
omo. es. «soc. moo.OI eemoae4unneme mean I mmoe
Ammo.v Ammo.v
«mm. as. «no. mee.I eooae4m meme I some
oenmeomom 4.3.av m a
Na eomuezleweeen meeueaeeem ee>u 4eeem meee>

 

.weeone4o4ueee wee wooB>4e eewe enmeueom pow .Iweeeu mafia eeee44 40

\N MmeeueEeeee weueEHDem .oN e44eH

\

105

.Amv 4e4ueeuoe mo ueeoeen om Cu ueeoeee ON Eoum zoom Ou wee4m nexeea ecu weefieuoe meee> wo eo4aez.l

 

 

 

 

\N
.o4nmnm4 .... .Numwm4 .4uewm4 ee43 e>ueo Noumeaoo mean eee4ueom
e4u eom neeoxe mNummm4 .... .anmm4 .4nmmm4 we weeemeee m4 e849 .memeeueeuen e4 eeoeee weeweeum\w
aemo.v 4N4o.v Aoeo.v
w wwm4 mm. mm. «mm. «mo. mNN. eueeeEoo
mmm4 I qwm4
Asso.v Amoe.v Amoo.v .eooseea
m @494 mm. mm. mum. oqw.ml 44N. o4um4mo4 ween eee4ueom
Aeeo.v Awoo.v Ammo.v
m4 nwm4 no. me. Now. ONo. 4q4. eueeano
m>o4 I mmm4
A4No.v Amm4.v Awoo.v .weeo4
an meme mm. mm. 4mm. Aeo.mI 4mm. geomewoe Ie4ueueee
Ameee>v Auee>v
leuee uefioe oflumeueum 4 e m meee>
mmzoeo e04uoe4we4 NM eomuezIe44ueo meeueaeeem eOHOUeem \uoewoem
.wooB>4n ween eee4ueom
wee .4N o44ea

weeo4e4owueea now m>eeo wueeeEoo wee eowuoeew o4umwm04 e40 mo \flmuoueEeeem weueafiumm

106

 

NRRKET SHRRE (PERCENT)
0

 

 

 

firY r‘rT'

1080 . t '1385
YEAR

0 I j I I l I I I I 1 I I V V I I' I
1050 1055 1070 1078 1000

 

 

 

Figure 8.--Logistic function fitted to particleboard
market share data, 1955-1979.

 

HRRKET SHRRE (PERCENT)
0

 

 

r

 

1 T f fi'j’ T I

I l I I I I V I I I l 7 I F" I T I I I r 1
1050 1055 1000 1005 1070 1075 1000

YERR

 

 

 

Figure 9.--Gompertz curve fitted to particleboard
market share data, 1955-1979.

107

 

 

 

 

31

A

h-

53‘

(J

M

u:

a.m.

u:

m:

C

:0-4

m—

.—

u:

g .1

mo

,1:
a rfirg' Y I I I '1 Y I I I Y rwfij
£980 £988 1970 1978 1900

YEHR

 

 

 

Figure 10.--Logistic function fitted to southern pine
,plywood market share data,,1964-1979.

 

 

 

 

IO-
N
i-
z O-
“J N
U
C
LIJ
e ..
LIJ
m
C
I O-I
a) .-
.—
DJ
x '1
$10
I:
a I —r A I T I I I ‘ V I ‘I‘ T r I V _f I I
1000 1005 1970 1070 1900
YERR

 

 

 

Figure 11.--Gompertz curve fitted to southern pine
plywood market share data, 1964-1979.

108
Based on the Durbin-Watson statistics at the 5 percent alpha level,
autocorrelation is present in all models. For projection purposes,
this does not pose a problem. It is interesting to note that the 1979
actual marekt share has surpassed more than 80 percent of the estimated
potential in all cases except the Gompertz curve for particleboard.
The implications of this result on projections are discussed in the

next chapter.

Multivariate Diffusion Submodels

 

While the univariate models are useful in depicting temporal diffu-
sion trends, they are not very helpful in understanding the role supply
and demand factors play in determining panel consumption and diffusion.
Two multivariate models were used to gain insight into the effects
of supply and demand factors. Time series data were used in the first
model and cross-sectional data were used in the second model. Estimated
structural supply and demand equations were used to screen variables
for the time series model.

Tables 22 and 23 present the structural demand and supply equations
for particleboard and southern pine plywood, respectively. Though
the results must be interpreted with the power of the t—test as a re-
striction, the signs of the coefficients were generally the expected
signs.

For particleboard demand, quantity demanded is a positive function
of one and two unit housing starts, mobile home shipments, and furniture
manufacturing. These relationships are expected since they are the
primary end-use markets for particleboard. Particleboard price and

housing starts with three or more units exhibited the opposite

109

.e04ueede >4eeem e40 e4 ueeumeoo e4u ueeoxe ewfie uoeeeoo

e40 we4>e4 mo >u4444e4oee ueeouee cw ee umee4 ue e>e4 eueeHUfimmooo woueawume 44< neuoz

mm N I 3 a me I NIexm
sm.~ I .3.9 .am. I

 

 

I NI \N
.aeae peaches aeoe some meme eons emuuee meannesumxw
mm.wmwm mm.4wmI ueeumeou
w¢.wo4 4m.mqm eowuoewoea ee4ee4
cm.e4m4 mn.m4wNI eewez
mw.m wm.e4 eofiea >weeem
mw.>4 mm.mNI mowue ee>4me4w< >uwueeev
N4.44 m4.NmI eofiee weeo4e4o4ueem weeo4e4o4ueem \m>4eeem ANV
mo.oom cq.wmmNI neeumeoo
wefieeuoemeeea
>>.4 «4.wN eeeufieeem mo xeweH
Nm. No.0 mueoae44m e804 e444oz
mm. ww.mI mowed +m .muueue weemeom
4N. ww.4 mDHee N44 .mueeum wefimeom >ufiueeec
mm.4 mm.mI eoeee weeo4e4oeueem weeo4e4040eem \Mweeemm 44v
4 43 cc CC
40 eoeue ueeeoHMMeoo me44e4ee> e44e4ue>
weeweeum e04meeewem >uOueee4exm ueeweeeea eowueevm

 

.weeo4e4owuuee e04 meOHueeve >4enem wee weeaew 4eeeuoeeum woueaeumm .NN e44eH

.I c C . ......L.

 

 

.em4e uoeeeoo e40
we4>e4 mo >04444e4oem ueeoeoe om ee ueee4 ue e>e4 eueefiofimweoo weueSHume 44< “euoz

 

llO

 

cm 4 I 3 4 .mm I NIm\m
OHNI SQ mm INI\M
.meme canoes“ some acne meme eons eeooee economasmxw
>¢.wmq4 om.NNmm4I ueeueeoo
me.oe Na.¢oe su4>nuoseoea 44es eooss4e
eoflea eweaeeum
4N.m4 mq.m4 w043em ween eee4ueom
e04en >ueueeev wooB>4n
w>.m4 em.mm woos>4m ween eee4ueom eefia eee4unom \m>4eeem ANV
4q.wmn mm.4omI neeumeou
we. 4m.NI muwee +m .euueum weemeom
ow. «4.e muHe: N44 .mueeum we4meom
eofiee
em.44 oq.wm wooB>4e wooaumom eeeumez
eofiee >u40eeec wooa>4e
oq.q4 nw.NwI woos>4e ween eee4ueom mean eee4ueom \MweeEeo 44v
a 43 co :0
mo eoeee ueeHUHMMeoo we44e4ee> e44e4ee>
weeweeum eOHeeeemem >e00eee4exm ueeweeema eowueecm

 

.l.wooB>4e eeHQ eee4ueoe pom me04ueeve >4enem wee weeEew 4eeeuoeeum weueEHDmm .MN o44eH

\e

lll

relationship. As price increased, ceteris paribus, quantity demanded
decreased. This is the expected demand price-quantity relationship.
Multifamily units utilize only a small portion of particleboard rela-
tive to other end-use markets (U.S. Department of Agriculture, Forest
Service, 1974). The direct competition of multifamily units with mobile
homes and one and two family homes leads to this statistical result.

As expected, particleboard quantity supplied declines as wages
and adhesives prices increase. Lumber production was used as a proxy
for planer shavings availability and has the expected positive relation-
ship. This result must be viewed with caution since residue price
would be a more appropriate supply variable and because solid wood
products production is positively correlated with residential construc—
tion. The net effect of increasing energy prices on particleboard
was positive over the test period. In supply functions, increases
in board price are associated with increases in quantity offered in
the market. The only unexpected sign was the negative sign on particle-
board price. This perverse relationship may be due to the general
negative trend in particleboard prices (see Table 9) and the predomin-
ance of demand in determining price—quantity relationships.

Sheathing is the primary market for southern pine plywood, and
as a result, one and two unit housing starts are positively correlated
with plywood quantity demanded. The role of western softwood plywood
as a substitute for southern pine plywood is indicated by the positive
sign on the former price. As was the case with particleboard, multi-
family housing starts have a negative effect on plywood quantity demanded.

The southern pine plywood supply equation had the expected signs

on price and productivity variables. As either price or productivity

t1
9C.
St
of
at
b1;
00
re
9X
St
in

dw

Va 1

112

increases, producers offer more plywood in the market. As the price
of inputs increases, a negative effect on quantity supplied is expected.
The sign on stumpage price was positive; however, it was only signifi-
cant at the 20 percent alpha level and the power of the test was very
small. Despite the statistical result, stumpage price was retained in
the equation because the cost of stumpage (l) is an important factor in
the total cost of producing the panel and (2) is expected to increase
in importance as competition for stumpage intensifies. The positive
correlation between product price and stumpage price and the small
sample size may have led to this result.

The final supply and demand equations were derived through an iter-
ative process in which many variables were examined for inclusion in
the regressions. Additional variables considered in the particleboard
equations were: weighted price of substitutes, total U.S. dwelling
starts, disposable personal income, index of all manufacturing, value
of residential construction, value of residential additions and alter-
ations, value of private nonresidential buildings, value of public
buildings, percent of all housing starts that were multifamily units,
output per plant, softwood plywood production and railroad freight
revenue per ton—mile. Other variables, in addition to those above,
examined for inclusion in the plywood equations were: one unit housing
starts in the South, percent of total U. 8. one unit housing starts
in the South, five or more unit dwelling starts in the South, total
dwelling starts in the South, personal income in the South, plywood
mill wages, adhesives prices, and energy prices.

The variables in Tables 19, 22, and 23 and lagged market share

variables were tested as explanatory variables for the models described

113

by equations (7) and (8). The lagged market share variables were
included to show the accumulated effect of supply, demand qualitative
factors on product diffusion. In short, they were used as proxies for
past market success and the knowledge gained by users as a result of
that success.

The regression results presented in Tables 24 and 25 corroborate
the importance of past success and accumulated knowledge in the diffu-
sion of the new products. In fact, the adjusted R2 for particleboard
and southern pine plywood were .94 and .98, respectively, without the
end-use market variables included in the model.

In the particleboard submodel, economic activity in the furniture
and fixtures industry and the mobile home industry are important vari-
ables for particleboard diffusion. This was an expected result since
particleboard made substantial inroads in both markets. However, the
combined effects of those two variables were small when compared to
the importance of the lagged market share variable.

The effect of using seemingly unrelated regression estimates was
a slight shift in the estimated coefficients and a reduction in the
magnitudes of the standard errors. Based on that result, the ordinary
least squares particleboard model was estimated for the 1956 through

1/

1979 period. The resulting regression—- was:

M= —.011552 + .719247 x M(—l) + .000216 x IMF + .000027 x MOBILE
(.005083) (.073876) (.000085) (.000007)
where:
M= particleboard market share (in decimal form)

M(-l) = particleboard market share lagged one year

 

1/

— Standard error in parentheses.

. 140.4:3334044L1
. - .h— Lnyh. F Av 14:5»
, ~44m flung—IT.
.. I~—WL~—L —.~.\Ln~n~:_.I~L IlsquL—.>.— a
- I. ,I. . -—E n..P—~ Kr. .J.
I r LAIVL.:=FNI~n.A~ .
., Tufluh ~2:.~. HJ..~ a
.. QN 0447.4

E

.eme uoeeeoo e4»
we4>e4 mo >04444e4oee ueeoeee ow ee umee4 ue e>e4 mueefio4mweoo weueawueo 44< "euoz

.me04ueeve eme4u e4 eoeoeemmew ueeUHMHeme oe e4 eee49.l

 

114

 

 

 

\N
.om.N u .3.4 .mm. u NIm .mnm4 4weoe4u eom4 aoew euew eo wome4 eeueafiumM\fl
Onnwoo. NNON4o.I .w4wmoo. nqu4o.I ueeueeoo
euee8e44m
mooooo. NNoooo. mooooo. NNoooo. e604 e444oz
wefieeuoemneea
«moooo. qNNooo. wo4ooo. qNNooo. eeeueeeem mo xeweH
Nwmmmo. ommN4n. nqNNmo. mNmN4N. eee4e nexeee wewwe4
4 mo 44V nee4o444eoo 4 mo 44V ueowowmmeoo
noeee weeweeum eoemeeewem eoeee weeweeum e04mmeewem
\Meeuea4ume .IeeueEHDme e044e4ee>
eowemeemee weue4eeee >4me48eem eeee%Wm ueee4 >eee4weo >e00eee4exm

 

. lweeo4m4ofiueea now 4eon4em eowmeww4w 4euomaeu .eueflee>4u4ea Q40 40 memumaeeea weueawumm .4N e44eH

I

115

.ewHe uooeeoo e4u
we4>e4 mo >04444e4oee ueeoeen ow ee ueee4 ue e>e4 mueewofimmeoo weueEHume 44< neuoz

.meOHueeve eme4u e4 eoeeuewmww ueeUHMHeme oe m4 euo4H.l

 

 

 

 

\N

.4o.N u .3.4 .am. u NIm .mmm4 4weoeeu ewm4 Boem euew eo wome4 eeueefiuem\fl
4newoo. Nwm4mo. 4mmmoo. mwm4mo. ueeumeoo

mu4ee
mooooo. mmoooo. I o4oooo. mmoooo. I Nw4 .mueeum we4eeom
moemmo. osqsoo.4 smeemo. oeseoo.e steam smegma eemwme
a mo 43v oceeoneeeou a no Ape nemeuneemou
eoeee weeweeum eOHemeewem eoeee weeweeum eoeemeewem Axv

Mmeueefiume \INImeueEHuee me44e4ee>
eOHmmeeuwe weue4eeen >4we48eem me4e54uee meeenvm ueee4 >eee4weo >eoueee4exm

 

.I.woo3
I>4e eefie eee4ueom e04 4eon4em eowm3444w 4eeoeaeu .euewee>404ea e4u mo meeueeeeee wemmafiumm .mN m44eH

116

IMF= index of furniture and fixtures production (1967=100)

MOBILE= manufacturers shipments of mobile homes (in thousands
of units)

The adjusted R2 was .99, and the Durbin—Watson statistic was 2.29.
The results were very comparable to those derived from the shorter
time period.

In the southern pine plywood market submodel, the lagged market
share variable was again the most significant variable. The negative
sign on one and two unit housing starts was contrary to expectations.
However, closer examination of housing starts and market share trends
shows that the greatest increases in market share were attained during
periods when housing starts were declining. This was due in large
part to southern pine plywood producers' ability to cut plywood prices
to a greater extent than western producers could.

The interaction of factors influencing diffusion over time, in
general, do not incorporate spatial aspects of diffusion. For example,
local products have limited potential because knowledge of products
is restricted to a given geographic region. By 1978, knowledge regard-
ing southern pine plywood was fairly extensive. Therefore, a spatial
diffusion model was examined to gain insight primarily into the effects
of transportation costs on spatial diffusion. Distance from Portland,
Oregon (DIST) in miles and the number of southern pine plywood plants
in adjacent states plus the number of plants in the state containing
the major portion of Rand McNally Major Trading Areas (PLANTS) were
used as explanatory variables. The sample was developed by crossing
the top 20 U. S. softwood plywood markets and the top 20 southern soft-

wood plywood markets. The resulting sample size was 29.

117

The a priori expectation was that the percent (PERCENT) of total
softwood plywood shipments attributed to southern pine going to a Major
Trading Area would increase as the distance from Portland increases
and as the number of nearby southern pine plywood plants increases.

This model provides a rough indication of the regional advantage

panel products can develop. The results of the regressionl/ were:
PERCENT= -10.60 + .023 x DIST + 1.37 x PLANTS R2 = .83

(6.78) (.003) (.21)
As expected, the regional proximity of the panel's producers and the
distance from the major competing panel's supply region were positively
related to the market penetration of the new product. Transportation
cost is the main factor leading to these results.

Two additional variables were tested for inclusion in the model.
Total plywood shipments to the Major Trading Area in 1978 and total
housing starts in the Trading Area's major metropolitan areas were
used as proxies for market activity. Though factor market penetration
may be expected in large, dynamic markets, neither variable was
significant at the 20 percent alpha level. Inclusion of the variables
did not improve the explanatory power of the model. Thus, southern
pine plywood attained spatial diffusion which was determined to a

large extent by its proximity to Eastern markets.

 

1/

-— Standard errors in parentheses.

PROJECTIONS AND NEWER PRODUCT INNOVATIONS

The projection model formulated for this study was that future
wood requirements for a given panel project would be a function of
future aggregate panel consumption, the paneksmarket share, and the
average wood requirements per unit of panel output. Disposable per-
sonal income was selected as the exogenous variable for the aggregate
panel consumption submodel time was the exogenous variable in the
nonlinear diffusion submodels. Wood inputs per unit of output, on a
weight basis, were assumed to remain constant over the projection
period. Bolt diameter, core diameter, and processing efficiency will
determine wood requirements in terms of roundwood volume; one study
assumes that wood requirements per unit of output, on a volume basis,
will fall ten percent during the 1976 to 2030 period for lumber and
plywood (Adams and Haynes, 1980). For this study, however, wood
requirements wane projected on a weight basis.

Rather than project consumption for particleboard and southern
pine plywood based on both nonlinear submodels, the models were examr
ined more closely to determine which best represented the diffusion
pattern for each panel. In this examination, insight was gained into
the usefulness of these models for projections under different
conditions of data availability

Following selection of the market share submodels is the pro—

jection of panel consumption and wood requirements. Finally,

118

119

interacting factors that will influence future diffusion are dis-

cussed, including newer product innovations.

Comparison of Logistic Function and Gompertz
Curve for Projecting Market Shares

 

 

Market share projections based on the logistic function and the
Gompertz curve are presented in Table 26 and in Figures 12 through
15. The Gompertz curve market share projections are always greater
than the logistic function projections for the complete data sets.

In the case of particleboard, the disparity between estimates
will lead to great differences in projected wood requirements. For
example, the particleboard projection for the year 2000 would be
23 percent greater if the Gompertz curve rather than the logistic
function were used for projections.

Since it is desirable to assess the reliability of these func-
tions for projecting future panel market shares under different
conditions of data availability, five time periods of varying length
for each panel were fitted with the logistic function and the Gompertz
curve. Regression results are presented in Table 27 for particle—
board and in Table 28 for southern pine plywood. Preliminary results
from these tables indicate that the logistic function yields a
better fit than the Gompertz curve for particleboard and that there
is little difference in the case of southern pine plywood. The adjusted
R2, Durbin-Watson statistic, and standard error associated with mar—
ket share potential (P) were the basis for the preliminary conclu-
sions.

By comparing the actual market share in the last year of a

120

Table 26. Market share projections based on nonlinear diffusion
functions, 1990 and 2000 (in percent).

 

 

 

Panel Function Year
product 1990 2000
Particle— Logistic 11.00 11.07
board
Gompertz 12.83 13.64
Southern Logistic 21.06 21.07
pine
plywood

Gompertz 22.19 22.28

 

121

 

 

 

 

.—
z 4
u:
U 1
K
M
L
U
(z
(I
J:
(D
I.—
U
K
K
(I
I:
o 1 I ' r ' l ' 1 ' '1
1950 1900 1970 1900 1900 2000
YERR

 

 

 

Figure 12.--Particleboard market share projections based
on the lggiStic function,_1980-2000.

 

 

 

 

h
2
U
U
g
U
0-
U I
z
I
I:
(D
h
U
x
z
m
z
o v T v I v If I I 1 I
1950 1900 1070 1900 1090 2000
YEHR

 

 

 

Figure 13.--Partic1eboard market share projections based
on the Gompertz curve, 1980-2000.

122

 

 

 

10
N
ha
Eiu
U
a:
m
52
In
a:
a:
to
ma
1'-
In
K
g.
I:
91 .' r ' I I I ' fl
1930 1970 1900 1990 2000
YERR

 

 

 

Figure 14.--Southern pine plxgood market share projections
based on the logistic function, 1980-2000.

 

I
l
I
I
I
I
I
IIIII
3|
'1'
3
I
I
!
!

 

 

I— .
z:° '
U “' O
u
a: .
In
L a . .
In
a:
a:
I: o
m c-
.— o
In
a:
gIo
I:
o I .* I I I ' I I I
1980 I970 1900 1990 2000
YERR

 

 

 

Figure 15.--Southern pine plywood market share projections
based on the Gompertz curve, 1980-2000.

123

Aoma.v Aomc.v Amoo.v

 

 

mm. N©.N mmm. moo. NNo. Noumano
Aomm.v Aooq.v Ammc.v

no. mo.m Nmo.a ooo.~- omo. onumfimoa ommHImmmH
Aomo.v Anqo.v Aaoo.v

mm. Ho.a Nam. men. «mo. nomeEOo
Ammn.v Amn~.v Aoao.v I

um. um. mmm. omm.nu mmo. unumfiuoa «coaummma
Ammo.v Aooo.v AoqN.v

no. ma.n ego. mac. Nam. NoquEOQ
Ammo.v Ammn.v Aqmo.v

mm. HN.H awn. mmo.m- men. umumfimon meannmmon
Aoao.v Awoo.v Ammo.v

us. No. Nos. omo. Hem. NoquEOU
Anmo.v Amma.v Aooo.v

ma. No. HNN. Noo.mu can. ofiumflwoa mmmaummmn

owumfiumum n m m
N m comum3ICwnu3a mumumamumm :oHuoasm mumm»

 

.mummm
wmuomamm .vumonmaofiuuma new m>uso uuumaaou mam cowuocsm oflumwmofl msu mo twpmumamuma commafiumm .NN manmy

\a

124

mm.

mm.

OH.H

om.~

Ammo.v
mow.

nomo.v
mum.

Awfioo.v
mooo.

Anmm.v
mow.MI

.wmmmSucmumm CH muouuo wumvamum I

\H

Amoo.v

mfifi. Nuumaaou

Aqoo.v

«ca. aflumfimoa mmmaumoaa

A.u.uaoov RN magma

Id

in. .1151:

r (46 H1
E

:

Iiiil‘t‘iaklltc

T<;r£~1:h

3m

Rik PK hurfi

125

 

 

ammo.v Awoo.v Amoo.v
ma. om.H new. mac. «on. unconsoo
ammo.v Ammm.nv Amoo.v
mo. co.a owe. qmq.w- mm“. odomfimoo Nuaanoaa
Aemo.v Aqno.v Aeao.v
ms. ma. man. nmo. Hmm. Noumaaoo
Ammo.v Ammn.v Amno.v
so. be. mum. wwn.m- NnN. odomfiwon Neanuqoofi
Amqo.v Anso.v Ammo.v
om. wn.n mam. mmo. mmm. Noumaaoo
Aemo.v Ammw.v Ammo.v
om. mN.H “mm. okq.qs mNN. unomflwou memaumomm
Ammo.v Ammo.v Aono.v
mm. mm. «an. «mo. mNN. NumeEOu
Aqqo.v Amoo.v Amoo.v
”a. me. mum. oem.mI Ham. unomfimon mnonlqomn
UHumfiumuw n m m
N M comumBIcwnusa mumumfimpmm coauocsm mummw

 

.wooB>Ha mafia sumnusom MOM m>u50 NuquEoo cam cowuucsm owuwfluoa mzu mo

.mpmmh vmuomamm
Imumuwamuma wmumafiumm .wm manmfi

\H

«.3. at:
U
c mm ninth

126

.mwmmfiucwhmm CH kaHHw UHQUCNUW I

 

\H
Ammo.v Aqoo.v ono.v
mm. mo.N mum. moo. Hmfi. Noumaaoo
Aqmn.v Aqu.~V Amoo.v
mm. ON.H gem. wmm.nnu man. usumflwoo okmanqoma

A.w.u:oov mm magma

111
PU'
pe

St

fi
di

fu

C0:
GOI
ti.
re;
the
10;
th.
He:
80:
to

801

3011

Squ

127

sample set with the estimated potential market share, the
limitations of the nonlinear functions become evident. The come
putations presented in Tables 29 and 30 show that: (1) over 85
percent of market potential was already achieved in cases when the
standard error of the potential was small and (2) the potential
market share was consistently underestimated in samples based on the
first half of the available data. Therefore, the deterministic
diffusion functions should be used for projecting only when the dif-
fusion process is substantially underway.

For most data sets tested, the logistic function yielded more
conservative estimates of market share potential than did the
Gompertz curve. After comparing the statistical results, the logis-
tic function was selected for particleboard projections. Additional
regression tests made on the southern pine plywood data indicated
that the estimated market share potential associated with the
logistic function was more stable than the potential associated with
the Gompertz curve as the sample size approached 16 (total sample).
Hence, the logistic function was selected for projecting future
southern pine plywood market shares. Projections and factors expected
to play a major role in slowing the diffusion of particleboard and

southern pine plywood are discussed in the remainded of the chapter.

Projections of Panel Consumption and Wood Requirements

 

Based on high, medium, and low estimates of U. S. disposable per-
sonal income, total panel consumption is projected to range from
48.5 to 52.5 billion square feet in 1990 and from 55.5 to 62.6 billion

square feet in 2000. Historically, the greatest annual consumption

ll‘ IN‘ ( ‘PC (‘0‘!!! t ’l“ (I [1 ‘11!!! “(I 111‘11‘1IFIE HMO (filfifls'b
Cw 1<:C~J<< (1:4: IIJIII JEIF <44 <4 rI~4 1 1 a 1 . .1 J p w

128

.Hmwucmuoa wmumafiumm seas vmumfioommm uouum upmvnmum mwumA.I

 

 

\m
.Hmfiuamuoa mum:m umxume mp wmww>ww mumnm umxuma Hmsuom amok ummA\M
.wOHuma mafia magamm mnmfilmmmfi mnu mo pmm% umma m£u ma mmmfi ..w.m .vowuma mafia mamamm mo pooh ummg\fl
mm. o.m m.- uuuwanu
mm. o.m q.oa uaumawoa mamaumoma
mm. m.H N.N Nuuoaaoo
mo. a.a o.N uaumawoa mmmaummma
mm. o.m q.m muummaou
as. o.m m.m usumawoa soaanmmaa
\M¢N. o.w ~.om Noumssoo
\mco. o.w N.qH UHumeoa manImmmH
\mwo. o.m H.qH uuumaaoo
om. o.m H.Ha uaumawoa mkoaummma
wofiumu
mum m umxuma Aucmoumav Auamoumav
Hmwuamuoa ou .Impmnm umxume Hmfiucmuoa mumcm
Hmauum ummm ummq Hmruom ummz ummq umxumz cowuuasm mummw

 

.mmaaamm omuumamm
cw wm>mfiaom mumnm umxume ummH onu ou Hmfiucwuoa mumsm umxums m.vpmonmaofiuuma mo somfiumgaoo .mm manna

tea/cwficr. atria ...C—LCE to"; die r... Fevecctcc Extra t..§.2...E 0:25:57. cric :L..:::C.:. mc :CvgrrdECC -Cm. 3:295

129

.Hmwucmuoa mumnm umxume zn vowfl>fin mumcm umxuma Hmsuom amok ummq.l

 

 

\N
.vOfiumm mafia oHaEmm mnmfilqomfi osu mo pooh umma mnu ma mmmfi ..w.o .moauoa mafia mHaEmm mo umm> ummq\w
mm. o.- H.m~ Nupanoo
Ho.a o.Na m.aa anomawoa ohmauqoma
mm. o.m~ «.0H Nuumasoo
No.a o.ma m.ma uaumswoa msoanoma
mm. m.oN H.mm Nuponeoo
mm. m.om N.HN oflumfiuoq nnoHIqomH
em. m.om m.m~ Nuumasoo
am. m.om m.NN oHumHuoq mmmHImomH
No. m.o~ m.~m nonmaaow
no. m.o~ H.HN oaumfiwoa meaauqoma
Wofiumu
mumbm umxuma Assoopmav Aucmoummv
Hmfiucotoa cu Imumnm umxume Hmfiucouon mumnm
Hmsuom pooh ummq ammuom ummh ummq umxumz cowuocsm mummy

 

.mmfiqamm wouomamm ca
wm>mflnom mumsm umxuma umma mcu cu Hmfiucmuoa mumsm umxume m.w003%HQ mafia cumsusom mo :omaummfiou .om mHan

an
ar.
of
je

pa

pr

83

Se

f(

of

130

was in 1978 when 39.3 billion square feet of panels were consumed.

Projected particleboard and southern pine plywood consumption
and wood requirements are presented in Table 31. The projections
are based on the logistic function diffusion trends and estimates
of disposable personal income. The 2000 medium income level pro—
jections are 82 and 62 percent greater than the 1979 volumes for
particleboard and southern pine plywood, respectively. The same
percentages apply to wood requirements on a weight basis. The
projected average annual growth rates are well below those
experienced in the industry during the 1960's and 1970's.

The U.S.D.A Forest Service (1979) projected higher levels of
particleboard consumption than were estimated in this study (see
Table 32). However, panels other than particleboard were included
in the Forest Service estimates. The Forest Service projections
were based on price trends that have prevailed during the last two
decades.

The southern pine plywood production projections estimated by
the U.S.D.A. Forest Service (1979) are similar to those developed
in this study for 1990 (see Table 33), but all 11 percent lower in
2000. The Forest Service projections for plywood are based on
stumpage prices that will equate stumpage quantity demanded to
stumpage quantity supplied (U. S. Department of Agriculture, Forest
Service, 1979). Other estimates of southern pine plywood production
for 2000 are 9.9 and 10.2 billion square feet, 3/8-inch basis
(Anonymous, 1981c; Haynes and Adams, 1979).

The trend model estimates are derived from simple projections

of consumption and an extrapolation of market share trends given

m~o>v~ 30H tam

. E: v 108

. ._Ma _ ”—

hCL

I -mu. UCLFCUL w 3 COL

\N

~X9. v3

—G.LCL Tc:

.1Cf~* Is foE-avhrhlkfx

\a

b 11.15%:

I... II

1111.»! Ida]

u

1.1 R—

131

. m m mm .I
HN< an H m\m

.ooos%aa mafia cumnusom so“ acoucoo mudumwoe unmoumm NH um manao> ocm moms huoam>o so

comma .mGOu mo mommmSOSu ca ocm oumonmaoauuma MOM mammnAhuoco>ov uanmB >uo .mcou mo moammsonu :H.l

 

 

 

 

\N
.ooo3kaa
mafia Ghmfiuflom How mammn LUCHIw\m 65m wumoanUHuqu How mflmmfl LUCHI¢\M .ummw mumdvm we mUGmeOSU CH\IHI
mmmo comm amass comes 304 woossam
meme mama magma mmooa answmz mafia
mess Nose memH seems swam oaumfimoa comausom
oemm meow ammo wmmm son
sews wwmm mane mmmm assess women
«seem oasw ammo woam awn: ouumawoa -maoauumm
oooN Oman oooN oaoa \MAm>ma gaseous
mufimfimhfidwmm @003 COHUQEDmGOU Hmcmm mEOUSH SOHUUGDm Hmfimm

 

mHm>mH 30H mam .Esfioma .Smfiz “em

a.

.oooN mam omafi .maoocH Hmaomuoa mfinmmoamfio mo
.Icofiumanmcoo Hmong mo mcoauomnoum .Hm manoH

ucmEmufisumu ooo3 Hmuou oam

\a

132

Table 32. Comparison of particleboard consumption projections made
in this study with projections by the U.S. Forest Service
(in millions of square feet, 3/4-inch basis).

 

 

 

Year Logistic function trend Base level price
. and medium income and medium incom
level projection level projection—
1990 5555 7625
2000 6519 9020
l/

— Includes waferboard, flakeboard, composite board, medium
density fiberboard, and similar products.

Sources: Table 31; U.S. Department of Agriculture, Forest Service,
1979.

133

Table 33. Comparison of southern pine plywood production projections
made in this study with projections by the U.S. Forest
Service (in billions of square feet, 3/8—inch basis).

 

 

Year Logistic function trend Equilibrium level
and medium income level price projections
projection
1990 10.6 10.5
2000 12.4 11.1

 

Sources: Table 28; U.S. Department of Agriculture, Forest Service,
1979.

134

market potential. The other projections involve estimates of future
price and quantity relationships in the stumpage and final product
sectors. The trend in the price-quantity relationships is implicit

in the former and explicit in the latter. The panel consumption and
wood requirement projections vary considerably depending on which model
is used and what assumptions are made regarding model variables.

Though the trend model is simple in design, it does indicate the
direction and relative magnitude of future particleboard and southern
pine plywood consumption. As was shown in the preceding chapters,
many factors have influenced the diffusion of these panels. The ex-
pected effects of these factors on future consumption and wood require-
ments are discussed in the remainder of this section. Newer substitute

panels are discussed in the final section of this chapter.

Future Particleboard Trends

 

Particleboard's unique properties and the flexibility of its manu-
facturing processes will continue to make it an attractive panel for
various end-uses. One example of particleboard's diversity is Mende
Process board or thin panelboard. The Mende Process board was developed
by Wilhelm Mende and Company in West Germany (Anonymous, 1971a). The
process has been modified over the last 10 years, but the primary com-
ponent is a continuous press. Like batch produced particleboard, Mende
Process board is used as furniture core stock and in construction
(Food and Agriculture Organization of the United Nations, 1976). It
also is used for door skins and wall paneling (Bryan, 1978). There
were eight domestic thin panelboard plants in production during 1980

(Anonymous, 1981a).

135

Improvements in manufacturing processes and the diffusion of exist-
ing process innovations will lead to more efficient production in the
future. Computerized process control systems, electron beam curing,
and new binder developments are examples of process innovations that
can make significant inroads in the particleboard industry (Blackman,
1978; Pease, 1977). Process developments will continue to play and
important role in maintaining particleboard's competitive position.

Declining relative price has been an important factor in particle—
board's diffusion. Particleboard price and consumption are correlated
with the prices of substitutes, disposable personal income, end-use
activity, input prices, and raw material availability. These factors
will determine future consumption levels.

Traditionally, substitute prices have risen faster than particle-
board prices. This trend is expected to continue in the future for
softwood and hardwood plywood which rely on higher quality wood inputs.
As income and end-use activity increase, particleboard consumption
will also increase.

Supply factors may constrain future particleboard consumption.
Adhesive prices and availability are major concerns for producers.
Though the net effect of energy prices was estimated to promote particle-
board consumption, long-term effects may be negative. For example,
multiple unit dwellings requiring less particleboard per unit could
be the new housing norm in another 20 years. Increases in other input
prices are expected to hinder future consumption.

The availability and price of furnish is the problem most widely
discussed by industry researchers. New sawmill production techniques
are reducing planer shavings and sawdust per unit of output (Dicker-

hoof and McKeever, 1979). Also, lumber production is increasing more

136

slowly than panel production and the available residues are more in
demand as an energy source and for papermaking. Unused manufacturing
residues, bark, logging residues, and roundwood are potential sources
of future wood materials (Dickerhoof, 1976; Lehmann and Wahlgren, 1978;
Lewis, 1965). Regardless of the source, greater handling and furnish
preparation costs are expected (Wilson, 1975). The net effect is a
slowing of the diffusion process as has been the case in recent years.

Thus, the traditional particleboard product is nearing the apex
of its market share. Both supply and demand factors have been instru-
mental in its growth and will retard its future diffusion. However,
newer panel products, such as, waferboard, medium density fiberboard,
and composite panels are just beginning the diffusion process and will
be the growth sector in the panel industry in the foreseeable future.
They will play a major role in closing the projected consumption gap

identified in Table 29.

Future Southern Pine Plywood Trends

 

Southern pine plywood manufacturers are diversifying their product
mix. The characteristics required in CD sheathing were met by southern
pine plys, and now higher quality plys are being marketed. Examples
of higher value added products made from southern pine are: plyform,
underlayment, 303 siding, and AC/AD panels. The addition of these
products helps manufacturers offset rising log costs (Baldwin, 1977).
Higher grade face plys also will be utilized in composiuapanels. Expan-
sion into other grades is expected to help southern pine plywood cap-
ture a greater market share in the future.

Process changes will continue in the South so that producers can

137
utilize more of the softwood resources. Productivity gains reflect
the modern, efficient processes incorporated in southern plants(Farris,
1978). In fact, capacity is still expanding in the South despite econ-
omic conditions; Federal Paper Board recently announced plans for a
new plant and International Paper has begun construction on another
plant (Lambert, 1981). These plants are being built in anticipation
of future demand.

Substitution of other products for softwood plywood may dampen
future diffusion of the product. In a recent survey (Carney, 1977),
builders mentioned hardboard, aluminum, and phenolic waferboard as
products replacing plywood in new residential construction. Shifts
to other products in nonresidential construction may be due to econom-
ics, fire codes, and other factors (Anderson, 1978).

Siding and All-Weather' Wood Foundations are end-uses in which
plywood may make significant progress. However, the concrete slab
is the primary floor system in single-family homes in the South where
southern pine plywood has its greatest acceptance (Carney, 1977). There-
fore, North Central and Northeast markets have the greatest potential
for southern pine plywood floor systems.

The southern pine plywood price trend is projected to increase
at about the same rate as western softwood plywood prices until 2000
(Haynes and Adams, 1970). As a result, southern pine plywood will
maintain its in-place cost advantage in the eastern United States.
Higher value added products, rising population in the South, and in-
creasing end-use activity should combine to provide larger markets
for southern pine plywood in the future.

Supply factors will influence the diffusion of plywood, too. The

138
most notable factor is the price and availability of softwood stumpage
in the South. Hair (1980) has estimated that stumpage price in the
South will rise at an average annual rate of 2.5 percent between 1976
and 2030. Though this is slightly below the rate experienced between
1969 and 1978, it is still a significant increase. Upward pressure
on stumpage price is predicted due to increasing demand for all southern
forest products, expected declines in softwood roundwood inventories,
and low levels of regeneration on private non-industrial lands (Anony-
mous, 1981b and Hair, 1980). Eventually, the stumpage price increases
will cause southern pine plywood to lose its competitive edge. Newer

panels, such as waferboard and composite panels, will gain as a result.

Newer Product Innovations

 

Newer panel products that have been successfully introduced in
the market give an indication of the direction of product innovation.
As has been the case in the past, the new direction has important rami-
fications with regard to wood inputs. The introduction of Douglas
fir plywood in 1907 as a door panel has had remarkable consequences
(Cour, 1955). Likewise, a faulty press valve in 1924 that resulted
in hardboard first being produced during William Mason's lunch hour
has had a significant impact on wood utilization (Maloney, 1977). Par-
ticleboard and southern pine plywood are other examples of panels whose
acceptance has altered perceptions of wood resources.

Rather than forging a new direction, the most recent product inno—
vations simply are continuing a process that has been underway for
years, namely, the utilization of more diverse wood materials. From

an economic standpoint, there is a dynamic shift from scarce materials

139

to less scarce materials. The increasing importance of southern pine
plywood vis-a—vis Douglas fir plywood is a case in point. The use
of manufacturing residues in particleboard is another example. Mere
recently, waferboard, medium density fiberboard, and composite panels
have added to the panel diversification trend.

A conceptual starting point for future wood-based panel products
is Dr. George Marra's (1969) "Non—periodic Table of Wood Elements."
He identified 14 elements beginning with logs and ending with cellu-
lose. Veneer, chips, flakes, particles, and wood flour are among the
elements. By combining wood elements with other panel inputs, such as,
binders and chemicals, boards with a variety of characteristics can be
developed. End-use requirements and processing technology will be key
factors in determining how these elements will be used in the future.

The product innovations discussed in this section are not new in
the sense that they were just recently introduced in the market, but
they are new in the sense that they are only recently gaining wide—
spread acceptance. Waferboard, for instance, has been manufactured
in Canada since 1961 (Annonymous, 1962). Domestic medium density fiber-
board production began in 1966, and composite panels were available
as early as 1956 (Ananymous, 1980b; Elmendorf, 1956). Changes in
standards, processes, panel prices, and the wood situation are factors
leading to greater diffusion of these products.

The newer panels primarily compete with existing wood-based panels.
One indicator of the expected competitiveness of newer panels with
existing panels is industry's capital expenditure commitments. New
mills and modernization projects scheduled for completion in 1979 and

announced specific projects requiring capital expenditures are listed

140

in Table 34. These projects accounted for over one—quarter of the
committed $1 billion (Anonymous, 1980a). Waferboard was the most men-
tioned type of new panel plant.

The diffusion rates of these newer products will strongly influ-
ence the potential diffusion of existing products, such as, particle-
board and southern pine plywood. For example, waferboard is gaining
acceptance rapidly in the Midwest. As a result, the southern pine
plywood potential in that area may be reduced. On the other hand,
waferboard and southern pine plywood may act in concert to reduce the
market share of western softwood plywood. Since few quantitative data
are available on waferboard, medium density fiberboard, and composite
panels, a qualitative review is presented in the remainder of this
section. The review will aid in assessing the potential success of

these panels.

Waferboard or Structural Flakeboard

 

Waferboard or structural flakeboard characteristics can vary due
to a number of factors. Some of the most important factors are: wood
species flake geometry, flake quality, flake alignment, average density,
density gradients, layer thicknesses, and resin content (Geimer and
Price, 1978).

Experimental flakeboard panels have been made with lodgepole pine,
loblolly pine, sweetgum, southern red oak, Douglas-fir, sweetbay,
red maple, black tupelo, white ash, post oak, yellow-poplar, and mixed
species. Board properties for these panels are reported in General
Technical Report WO-S, "Structural Flakeboard from Forest Residues"

(U. S. Department of Agriculture, Forest Service, 1978). In general,

141

 

oususm osmonuomms muoxmn zosom .ou panama moawm oumSom
coaumuficuoooa

mama “swag eumonumwme m.z .smeumeo .eua .maaaz wooseuuoz

owmfi cumonuowmz omnmso .muumm ma .ouq .couuom xofiEMoz

owmfi oumopuommz oonmso .uo.o Hm> .ouq .xouon

museum osmonuoumz oaumuco .ou monasq mmcwm cumsom

meH oumonummmz mcflmz .ocmaoooa .auoo ofiwflommImeuomu

mmmfi oumonpmmmz .mflz .oum3zmm .muoo owwaommlmcmwmflooq

swam eumonummms .ccaz .xooo .auoo nuumHuom

swam eumonummmz .casz .HneaEmm .auoo noumauom

mesons eumonummmz .ccuz .Hmeaamm .eua .maaaz sooseuuoz

owafi osmonuowmz \M.ccaz .flmanom .auou cameommlmcmflmfisoq
cofiumNchmooa

owmfi woman osmonummmz .CGHE .mofiamm ocmuo .ouA .muosoonm umouom .o.m

coaumHano ucmam mo maze COHumooa zamafioo

 

uamaomam omucsoccm cam mmmfi CH cowumHano

.mmunufiocmaxm Houwamo wcwuflsvmu muoomopa
mow omaoomsom muomHOpa soaumNficumooB ocm mHHfiE 3oz .qm maams

142

.mome .msoE%coc< "mousom

.GOfiumuoa m>HumauouHm HOW wafixooq.l

\s

 

@maH

mmofl

owmfi

Hwofi

mususm

Hmcma muwmoaaou
Hmcma muwmonaou

Hmcma m>ufimoq
IEou .oumonmxmam

oumon
ocmuum omucmwuc

coauMNchmooz
uamHm oumonummmz

.mp0 .mHHH>mc3oum

.mp0 .umxmm

.o.z .smaeso

.m.z .ucoEmumHU

.muae .mema m>mam

.auoo osmonmam

.ou umnEHH commcfiaam

.auoo cameommImeuomo

.auoo oumom umuoocmeam

.cua .memamo mo wooseama

A.u.ucoov em canoe

143
a variety of species can be utilized to create boards with aCCeptable
stiffness and strength for various end-uses.

In order to gain wide acceptance and to facilitate diffusion,
code approval is needed. Canadian waferboard for sheathing was first
approved by the New Jersey State Building Code in 1972. Subsequently,
Canadian boards were approved by the Building Officials and Code Admin-
istrators International Conference of Building Officials (Jorgensen,
1978). Roof sheathing and wall sheathing are approved uses.

The American Plywood Association has begun to promote end-use
or performance—rated standards. The new standards do not require parti-
cular types of wood materials to be used in the panels. Instead, they
specify performance characteristics. For example, sheathing can be
conventional veneered plywood, composite panels, or unveneered panels
such as structural particleboard, waferboard, and oriented strand board,
so long as the performance standards are met (Lambert, 1980). As a
result, waferboard and other newer panels will penetrate the market
more readily.

Waferboard production began in 1961. The first product was pro-
duced by Wizewood, Ltd. at Hudson Bay, Saskatchewan and was sold under
the trade name of "Aspenite." The process used in the first plant
was developed by Dr. James d'A. Clark to utilize green poplar logs
in the production of a weatherproof phenolic-binded board (Anonymous,
1962).

Though waferboard production started in the early 1960's, con-
struction on a second waferboard plant did not begin untill969, and
the first United States plant did not come on-line until 1973. By

the end of 1980, eleven plants in Canada and the United States with

144
1.25 billion square feet of capacity (3/8—inch basis) were expected
to be operating (Pease, 1980). As indicated in Table 34, continued
expansion of waferboard capacity is planned.

The unique aspect of the waferboard manufacturing process is thick,
manufactured wafers are used rather than traditional wood materials
found in particleboard or hardboard. Since the wafers are large in
size, less resin is required in the manufacturing process. In the case
of structural flakeboard, the flakes are thinner than wafers, but no
specific delineation exists (Dickerhoof and Marcin, 1978).

Several existing means for manufacturing wafers and flakes hold
promise for the future. These include the spiralhead clipper, the
shaping-lathe headrig, disk flakers, drum flakers, and ring flakers
(Arola, 1978; Koch, 1978; Price and Lehmann, 1978). The best method
for flaking will be dependent on the species and wood form (e.g., round-
wood, chips, etc.) used.

One flakeboard variant is oriented strand board. Mechanical ori-
entation of strands yields improved flexural strength in the aligned
direction. Several existing waferboard plants have the capacity to
orient wafers. The Elmendorf Board Corporation plant in Claremont,

New Hampshire, will be the first plant constructed in the United States
solely for the production of oriented strand board (Anonymous, 1979b).

Regardless of whether the panel is structural flakeboard, wafer—
board or oriented strand board, delivered price to major market areas
will play a critical role in the diffusion of these products. In east-
ern Canada, waferboard's delivered cost advantage over softwood ply—
wood produced primarily in British Columbia is an important factor in

its market success (Dickerhoof and Marcin, 1978). The regional

145

advantage is only partially offset by the greater shipping weight of
waferboard.

Feasibility studies indicate that flakeboard production sites
in the Eastern and Southern United States have the greatest potential
for competing with softwood plywood. Higher plywood prices in the
eastern markets, lower freight costs for flakeboard from the East and
South, and lower labor and raw material costs are cited as factors
contributing to this potential (Adams, 1978). The locations of exist-
ing and announced plants in Table 31 corroborate the eastern advantage.

Aspen is the primary wood species used by existing waferboard
manufacturers. Aspen has been a little utilized species in the past,
but competition for resources is increasing. In fact, one potential
manufacturer decided to look for a new plant site after a competitor
announced plans for construction in the same timbershed (Lambert, 1981).

There are, however, significant volumes of aspen in the East,
especially in the Lake States. From 1970 to 1977, aspen and cottonwood
growing stock on commercial timberlands in Lake States increased from
9.0 to 9.3 billion cubic feet (U.S. Department of Agriculture, Forest
Service, 1974; U.S. Department of Agriculture, Forest Service, 1979).
Aspen and cottonwood accounted for a greater percentage of Lake States'
growing stock volume than any other species, hardwood or softwood.
By considering other species for which experimental boards were tested,
the potential wood resources for structural flakeboard can be expanded
significantly.

The outlook for waferboard is promising despite current economic
conditions. For instance, a Lake States' waferboard plant that started

production in late 1979 is now producing panels on a seven day, 24

146
hour per day schedule even though most panel plants in the United States
are operating well below capacity (Pease, 1981). Desired board char-
acteristics, code acceptance, delivered price advantage, and large
volumes of potential wood inputs will contribute to waferboard's rapid
diffusion. Lack of consistent consumption data precludes quantitative

estimates of waferboard's potential.

Medium Density Fiberboard

 

The second panel product currently gaining acceptance in the mar—
ket is medium density fiberboard. In the United States, two types
of medium density fiberboard are manufactured: one type is for exter-
ior siding and the other is for interior applications, especially fur-
niture (Food and Agriculture Organization of the United Nations, 1976).
The panels are composed of compressed interfelted fibers, binders,
and other material added to improve board properties.

For exterior siding, the boards are finished and coated for dura-
bility. Due to board surface properties, a high quality finish is
obtained. Competing products include plywood, aluminum, and other
hardboard siding. Dry, wet, and wet-dry processes can be used to man—
ufacture this board (Food and Agriculture Organization of the United
Nations, 1976).

In furniture applications, medium density fiberboard has a very
smooth surface for finishing and good edge properties. The surface
has less tendency for showthrough than particleboard, and the edge
properties allow finishing without edge banding (Maloney, 1977).
Particleboard is the primary competing product in this end-use. Dry

processing with supplemental high radio frequency heating is a common

147
means for manufacturing the furniture board (Food and Agriculture Organ—
ization of the United Nations, 1976). Numerous rocessing improvements
have occurred in machining, sanding, laminating, embossing, and finish-
ing medium density fiberboard (Anonymous, 1980a).

NPA 4-73 "Standard for Medium Density Fiberboard," is the standard
promulgated by the National Particleboard Association for this product.
In addition, medium density fiberboard is included in the industrialite
class of Voluntary Product Standard PS 58-73, "Basic Hardboard" and
in PS 60-73, "Hardboard Siding." Physical property requirements such
as modulus of rupture, thickness tolerances, and screw holding ability
are included in the standards (Maloney, 1977).

As mentioned previously, the first medium density fiberboard plant
began production in 1966 and was located in Deposit, New York. Between
1969 and 1976, ten plants began production (Anonymous, 1980b). Thir-
teen plants were in operation by the end of 1979 in the United States
(Lambert, 1980). Production figures for the 1975 through 1979 period
are presented in Table 35. Market share increased from 0.75 percent
to 1.3 percent during that period. Though the diffusion process has
begun, too few data are available for a quantitative estimate of med-
ium density fiberboard's potential.

Both hardwoods and softwoods are used in medium density fiber—
board production. Ponderosa pine, southern pine, Douglas fir, and
mixed hardwoods are used as furnish. Pulp chips, which are later fiber-
ized, are the preferred form of raw material (Lehmann and Wahlgren,
1978). Therefore, producers must compete with the pulp and paper,
particleboard, insulation board and hardboard industries for supplies.

Medium density fiberboard is more expensive than particleboard

148

Table 35. Medium density fiberboard production, 1975-1979 (in millions
of square feet, 3/4—inch baSis).

 

 

Year Domestic production
1975 215
1976 280
1977 441
1978 480
1979 490

 

Source: Lambert, 1981.

149
on a comparable per unit basis. However, some furniture manufacturers
are finding that it is very competitive in end-use cost. This is due
to processing advantages such as machining, finishing, and shaping
(Anonymous, 1980e). Computerized financial analysis is one tool being
used to estimate end-use cost. As computer analysis increases, medium
density fiberboard diffusion is expected to increase.

The outlook for this product is good in siding and furniture appli-
cations. Substitution for particleboard in furniture manufacture is
the most promising market. Though there have been no new plants opened
in the last few years, an economic recovery may lead to the construc-
tion of new plants and an erosion of particleboard's market share.
Excellent panel characteristics, standardization, and a competitive
end-use price are factors that will have a positive effect on medium

density fiberboard's diffusion.

Composite Panels

 

Composite panels were made as early as 1955; hardwood veneer faces
and a resin-impregnated, softwood fiber core were used (Elmendorf,
1956). However, little interest was shown in composite panels until
recently. With rising wood costs, products utilizing more low cost
wood and retaining a high quality appearance are becoming desirable.
Composite panels, which have particleboard or oriented strand board
cores and veneer faces, fit these requirements.

The composition core and veneer faces make composite panels slightly
stiffer than plywood. The solid cores also reduce moisture penetration
and subsequent cracking of plys receiving exterior exposure (Blackman,

1980). Manufacturing processes vary by producer; the components can

150

be layed up separately or in the same process.

Composite panels meet the standards promulgated by the American
Plywood Association for sheathing and flooring applications. Code
acceptance has also been gained for some panels (McSwain, 1977). In
those end-uses, the panels compete primarily with conventional plywood.
Though price and production data are unavailable at this time, it is
expected that the panel will be competitively priced. The use of low
cost wood in the composite core is a major factor aiding this panel's
ability to compete.

By making composite panels, manufacturers are able to utilize
the entire veneer log in one product. As a result, increased yields
from a given supply of logs are possible. Lehmann and Wahlgren (1978)
estimated that a 40 percent yield increase from a veneer log supply
is possible. By using small logs and culls for the composite panel,
one manufacturer estimated that composite panel output would be twice
the plywood output for a given volume of logs (Bryan, 1981). Either
oriented strand board or particleboard can be used as the core panel
(Blackman, 1980; Lehmann and Wahlgren, 1978). Therefore, composite
panels offer flexibility with regard to core materials and stretch
existing veneer supplies.

The first modern composite plant started production in 1975 and
produced panels with softwood veneers and oriented strand cores (Anon-
ymous, 1975). The same plant closed in 1979 due to environmental con-
trol costs and poor market conditions (Lambert, 1981). Two addition-
al plants began production in 1979 and 1980. In addition, the U.S.
Forest Service and the U.S. Department of Housing and Urban Develop-

ment cooperated in a research and development program for composite

151
panels in the mid-1970's (McSwain, 1977).
The outlook for composite panels is excellent. By early 1981,
nine plants in the United States, including those listed in Table 31,
indicated they were producing composite panels. Given the economic
climate, this increase in capacity is quite remarkable. Code accep-
tance, resource utilization, and similarities to conventional plywood

are factors that will promote the diffusion of composite panels.

Other Factors Influencing Panel Diffusion

 

Though a multitude of factors other than those discussed in pre-

ceding chapters and sections may influence product diffusion, only
five have been selected for brief discussion in this section. The
five factors are: (1) formaldehyde emission, (2) market conditions,
(3) U.S. Forest Service policies, (4) research and development, and
(5) nonwood product competition. Other influencing factors such as
antitrust laws, patents, capital availability, and internal company
policies are beyond the scope of this study.

Formaldehyde emissions during the production and use of panels,
notably particleboard, have been acknowledged for some time. The
specific effects of long-term exposure are not known with certainty,
but some individuals are affected by the emissions. If regulations
are mandated that limit allowable emissions, particleboard diffusion
could be adversely affected. Publicity regarding formaldehyde release
may have a similar impact on diffusion (Lambert, 1981). Foam insulation,
however, may be affected to a greater extent due to its formaldehyde
content. Adhesives product innovations, e.g., isocyanate-based binders,

provide one means of overcoming diffusion barriers due to emissions

152
problems (Ball, Redman, and Adams, 1979).

A second factor that may influence product diffusion is poor mar—
ket conditions (Cox, 1974 and Dickerhoof, 1977). Plant managers may
be unwilling to try unproven products during periods when economic
activity in end-use markets is low. However, if the new product has
been successful as was the case with southern pine plywood and wafer-
board, diffusion proceeds. In essence, the new products may gain in
market share despite the poor performance of the entire wood-based
panel sector.

U.S. Forest Service policies influence product innovation and
diffusion through research activities and harvesting practices. Pro-
cess and product research performed or funded by the U.S. Forest Ser-
vice may lead to new processes, new products, or the dissemination
of information (Fassnacht, 1964; Lehmann and Walhgren, 1978; U.S. De-
partment of Agriculture, Forest Service, 1978). A less discussed im-
pact on product diffusion may result from harvesting policies. If
harvesting is restricted in a manner that increases final product cost,
the diffusion of competing products is enhanced, ceteris paribus. Un-
fortunately, forest industry generally views harvesting policies (e.g.,
the allowable cut) as factors constraining their growth locally rather
than promoting their growth nationally (Lambert, 1981).

Research and development is cited as another factor affecting
innovation and diffusion (Cox, 1974). Research and development expen-
ditures in the forest products industries are traditionally below those
in other manufacturing industries (Risbrudt, 1979). No research and
development expenditure data specific to the particleboard, southern

pine plywood, and other new panel product industries is available,

153
but it is likely that expenditure levels relative to other manufactur-
ing industries are low. Research and development activities by govern-
ment, universities, the equipment manufacturing industry, and the adhe-
sives industry may help offset these low expenditure levels and mitigate
any adverse effects on the diffusion process.

A final factor affecting product innovation, diffusion, and consump-
tion is competition from non—wood products (Zivnuska, 1963). The prices
of most competing materials have increased little relative to all goods
since 1967 (U.S. Department of Agriculture, Forest Service, 1979).

With the exception of western softwood plywood, most panel products
have declined in relative price over the same period. Therefore, from
a price standpoint, wood-based panels may be improving their position
relative to competing materials. However, the substitution of other
products for wood—based panels (see Table 36) will continue to be a
major factor limiting the size of the panel market and the diffusion

of panels within the industry.

154

Table 36. Examples of products substituting for wood-based panels

in various end-uses.

 

End-use

Substitute material

 

Floors and foundations

Roof sheathing

Siding

Wall sheathing

Furniture

Concrete, boards
Boards in single-family roof decks;
concrete and metal in multifamily

roof decks

Brick, aluminum, stucco, boards,
steel, plastic

Plastic foam, foil—faced paperboard,
gypsum products

Plastics

 

Source: Anderson, 1980; Anonymous, 1979a.

APPENDIX

DATA USED IN ECONOMETRIC AND DIFFUSION MODELS

155

 

«HH.H N H mNH.H Roma
wma fl mom coma
Now q mow momfi
owe N wmo «omfi
wmq H omq momfi
woe woe Nomfi
on on Nomfl
moN woN ooafi
omN ooN mmmfi
omN omN wmmfi
mwfi mwfi mmmfi
HHN HHH ommfi
On On mmmfi

.Ifloauaasmcoo coauosoopm

\N ucmpmaa< muuoaxm muuoaeH ofiumosoo you»

 

. IAmHmmn LocHIq\m .ummw mumsvm mo mGOHHHHE

ch manImmmH .coHuQESmcoo ucmumamm ocm .muuoaww .muuoaaw .CONuosooua oumonmaofiuuma .m.: .H< mHQmH

156

Nwm.m
qq~.m
ch.m
mm~.m
qu.N
mom.N
qmq.m
owo.m
Nwm.N
mmm.fi
«HN.H

oNq.H

.unwfioa >3 omuuoamu memo upomEH aoum omumawumm

.msmsmu mnu mo smmunm mnu an 3mfi>mu poocD.I

3
\m.

.muuoaxm I muuoasw + aowuoswoua uwumoEoo u coauQEDmcou undamaa<.l

\N

.oumon mmmooua doom: mam oumonumpwm huflmcmo aswoms monaocfi uoc mmom.l

OOH

mm

om

ow

NHH

um

mq

ON

OH

«H

NH

«H

NH

Hom.m
oHc.m
Nmm.m
NON.m
mmm.N
qofi.m
qmq.m
NHH.m
«om.N
cen.fi
oHN.H

qu.H

2

mmmfi
wmofi
mmod
ommfi
mmmfi
qmmfi
mmmfi
Nmmfi
Hmmfi
Onmfi
mood

womd

A.o.ocoov H< manma

157

.ncmH .mmamsm vow uLwfluz mowafi .mUNEocoom HmfiuumnocH mo common .mouoEEoo mo
ucmauumnwc .m.D ”on mummh omuomamm .msmcmo one we smousm .moumaaoo mo ucmEuumama .m.: ”on
mumm> omuomfiom .msmcmo mzu mo smmusm .mouoEEou mo ucmEuumamo .m.D mnmmfi .maamnm mowmfi .uumnEmA "mmounom

A.e.ucoov H< magma

158

qu.m Ne mmq Noo.m NomH

 

oom.m me Noe mmo.m ocmH
HN¢.m mm mwm NHm.N momH
onm.m om mHm Now.N «omH
HqH.m mN mmq NON.N mooH
«ow.N NH HNm on.N NomH
qu.N NH NwN qu.N HomH
wNH.N NH moN 0mm.H oomH
on,N NH NNm HNo.N mmmH
on.H NH qu moo.H wmmH
qu.H oH NmH 0mm.H NmmH
owo.H mH HHN qu.H omoH
moo.H NH NmH oNq.H mmmH

.lcoaumasmaoo coauonooua

\H ucmumaa< \Mmuuoaxm \mmuuomEH owummEoo mom»

 

.wammn nocHIw\H .ummm mumscm mo
mcoHHHHE ch mNmHImmmH .COHuaanmcoo ucoumaam mam .muuoaxm .mupanH .GOHuosooua oumonoum: .m.D .N< mHan

159

.ome .moHEocoom HmHuumaocH mo smmusm .moumEEoo mo unassummoo .m.D MNNmH .um>mmMoz "mmousom

.muuomxm I muuanH + GOHuosooua oHummaoo n GOHuQESmcoo ucmumma<.l

\N

.AonHV uw>mmxoz Eoum dump mCOu ou ummm mumnvm COHHHHB Honccm mSu wcwm: wmumEHumm.l

\H

 

mmo.w
Now.m
moo.N
NNq.c
mHm.m
mqq.o
mNo.N
woo.o
omo.m
oNN.<
ONw.q

ooN.q

oHH

mm

ooH

owH

NmH

mNH

omH

HoH

«w

NN

No

Nm

Hoo.H
mNH.H
coo
Hoe
HNN
mNN
@OH.H
woo.H
«mo
mow
ooN

mNo

mwo.N
on.N
wom.o
mNH.o
HON.m
mew.m
oqo.o
HNo.m
oNH.m
oqm.q
NmH.¢

moo.m

mNmH
wNmH
NNmH
onH
mNoH
quH
mNmH
NNmH
HNaH
onH
mcmH

womH

H.e.ucoov N< «Hana

160

mmN.m «q we ch.m NmmH

 

mmo.m we we mNo.m oomH
NNm.m Hm Ho Nom.m momH
mNN.m Hm Ne NoN.m womH
ch.m Hm om amo.m momH
Nww.N me oe www.N NomH
wa.N Ne «N omw.N HomH
mew.N om Hm wow.N OQNH
NHH.m Nm oq «HH.m mmoH
HNw.N mm mN «ww.N wmmH
qu.N em oH mNo.N NmmH
«em.N om om mNm.N ommH
mHm.N em «N mqo.N mmmH

lboHuaEDmcoo GOHuonooua

\N ucmumna< \Mmuuoaxm \mmuuanH oaummaoc umow

 

.Amfimmn LUGHIN\H .ummm mumscm mo mGOHHHHE
cwv mNmHImmmH .GOHuQEDmsoo ucmuwaam vow .muuomxm .muuanH .GOHuosooua oumon cowudemcw .m.D .m¢ mHan

161

wcH.m
MNm.m
Hom.m
HNm.m
HHw.N
qNN.m
NNm.m
qu.m
ocw.m
omN.m
omo.m

mmm.m

.omma .muancoom HmfiuumsecH to ammuam .muumeaoo mo unassumamn .m.a “mums

.um>mm¥oz "mmuusom

.muuoaxo I muuoaafi + GOHuosooua UHuonoo n GOHuaasmcoo ucmumaa<.l

.AoNoHV uo>moxoz Boom owumu mcou ou uoom mumsvm coaHHHE Hmsacm mnu wcwm: woumafiumm

me

me

ow

0N

0N

HOH

wN

No

we

om

we

Nm

me

mm

on

oq

Hm

me

ow

om

mm

Nm

Hm

HHH

on.m
NNm.m
Hmm.m
Noo.m
www.N
NwN.m
«Hm.m
mHm.m
mmw.m
qu.m
MNo.m

oN¢.m

\N

\w
sham
mass
Hams
sass
mass
«Nos
mama
mean
Haas
oNaH
mesa

wcmH

H.e.uaoov me magma

162

 

NmH.m m qu.H ch.H NomH
HNm.m w qu.H oNo.N oomH
omo.m o Nqo.H mco.N momH
aqo.N N oq0.H NH¢.H qoaH
NHc.N H mmm mwo.H momH
qoq.m N How on.H NomH
wNo.N m NNN mom.H HomH
«Hw.H N mHN NoH.H oooH
HmN.N m wmm oqm.H mmmH
NoN.H N moo HmH.H wmmH
mNN.H H Nmm NNH.H NmmH
qu.H H wmq qu.H ommH
mmN.H N Nee mmm.H mmmH

ItOHuaeamcoo SOHuosooua

\H ucmumaa< muuoaxm muuanH oHumoEon snow

 

.AmHmmH LocHIm\m .ommm mumswm mo chHHHHE
GHV mNmHImmoH .:0Huaenmcoo ocmumaam can .wuuoaxm .muuoaaH .aoHuosooua oooS%Ha ooozoumn .m.: .q¢ oHan

163

.ome .moHEocoom HmauumaoaH mo smousm .mouoaaoo mo unmauumaon .m.: ”va mummz

wouoonm .msmcmo msu mo :mmusm .oouosaoo mo unmauumama .w.D “Nov mumm> wouoonw .msmcmu mnu mo

smmusm .moumeaoo mo unmauummma .w.D mwNwH .mow>umw ummuom .muduHsoHuw< mo unmeuumamn .w.D "mmounow

NwH.w
qmw.m
NNm.w
owm.m
Nom.N
NNm.N
mmN.q
mmH.m
omq.q
qu.m
owm.w

Nww.w

oq

om

ON

mN

we

we

oq

wN

mH

wm

wH

qH

wwo.N
qu.N
qu.N
wmm.N
me.H
qu.H
NNm.N
me.m
qu.N
Nqo.N
NoH.N

www.H

mHH.H
wNH.H
NwH.H
wwo.H
Nmo.H
Hoq.H
Now.H
owo.N
«N¢.H
me.H
www.H

moo.N

.muuomxo I muuoaaH + cowuonwoua owummEow u COHuQSsmcoo ucmumaa< I.

\H

meH
wNaH
NNmH
wNmH
mNmH
«NoH
NNmH
NNoH
HNmH
onH
mme

wwwH

H.e.ucouv q< mHnma

164

 

 

me.NH mw m oew.NH Nme
«cw.NH we m mew.NH wme
Noq.NH Om m qu.NH mme
qu.HH wN m mmq.HH «wwH
Nwm.cH wH oH mNm.oH mon
HHm.o NH mH mHm.¢ Nme
mae.w «H mH wmc.w meH
NmN.N mH HH mmN.N ome
qu.N NN I wwN.N mmmH
mine 2 \N :36 was
mm: 2 \M 203 $2
qu.m mH I qu.m wmoH
ENK m \N. «38 HS
.IcOHuaasmcoo COHuoswoua
\H ucmumaa< mononxm mouoan ofiumoaon poo»
.AmHmmn nocHIw\m .umow mumswm mo mcoHHHHE
CHV mNmHImmmH .COHuQEDmcoo acoumaam wow .mupoaxo .muuoaaH .cOHuoswoua voozuHm wooauwom .w.D .m< oHan

165

.ome
.moHEocoom HmwuumnwaH mo ammusm .moumfiaoo mo ucmEuuman .w.D “Nov whom» omuomHmm .mSmamo mnu mo

nmmusm .moumEEoo mo ucmauumama .w.D “Nov wummh wmuomHmm .mnmcmo map mo :mmpam .moumeaoo mo ucma
qumaom .m.D “any mummh wouomHom .mamcmu msu mo amousm .moumfifiou mo unmauumamo .w.D “NNmH .mmHmnm "mmousow

.uomm oumnvm ooo.oom cmnu mmmH.l

 

.muuoaxm I muuanH + SOHuoswoua ofiummEow u GOHuaEnmcoo usmumma<vm
qON.wH Noe wH wwm.wH aNmH
NNN.mH woN mm Nm¢.mH wNmH
wow.wH NwN wH NNw.wH NNmH
NON.NH wHN NH wom.NH wNmH
NNo.qH HmN N woN.mH mNmH
mwN.qH New q wom.mH quH
NNm.NH HHq a mNm.NH mNmH
me.NH HNN w mew.NH NNmH
me.wH mm m mmm.wH HNmH
wmo.qH qHH N qu.qH onH
cmm.mH mmH mH www.mH mme
Nmm.qH «w oH mwm.qH wme

A.w.ucoov m< mHan

166

owN «Hw.w wNmH

 

mHN wNw.m mNmH
owH.m quH
owm.m MNmH
mHm.m NNoH
qu.q HNmH
mHm.m onH
me.N mon
wNw.N wme
mNN.H Nme
qu.H wme
Noe mme
ow «waH
AmHmmn LusHIq\wv cowuoswoua AmHmma nocwIw\mV cowuoswoua
wumonumnfim Nuwmcmw asfiooz ooo3%Ha mafia cumzusom pom»

 

.Aummw mumavm mo
mGOHHHHE :HV mNmHIqme .GOHuoswoua osmonponflm NuHmsoo eusmE was wooshHa mch suonusom .m.D .w< mHan

167

 

.meH .uuonEmH mmNmH .comumw:< "mooudow
owe omw.N meH
owe wow.N wNmH
see Nee.s seas

H.o.uaoov we mHan

168

 

 

 

HNqo. HmNo. wqu. NNw.MN Nme
ONqo. quo. NHaq. NNN.MN wwaH
aqmo. HNHo. mem. NqH.mN nme
meo. oqoo. NwNm. wow.HN qme
mmNo. cNNm. on.mH mon
wNNo. HHNm. www.NH Nme
NoNo. wNNm. ooH.wH meH
wNHo. HNNm. wmw.qH owwH
meo. Nwwq. www.mH mmmH
meo. quq. me.mH wmmH
mmHo. HHNq. on.HH NmmH
Nmoo. mqu. woo.NH wmmH
owoo. Nch. mw0.HH mmmH

\mwumoanoHunmm \mwooazHa mafia cumzusow \mtoo3NHa voo3uwom cumumoa \mtOHuaasmcoo
\mmumzm umxumz Hocma Hmuoe pmow

.AEuow HmEHooo .ucmopmn :HV mNmHImmmH .oumoanoHuuma cam .woozmHm mcHa

cumsusom .mooBNHa wooBuwom cumumm3 pow cowuaenmcou mo mumsm umxuma ocm cowuaESmcou Hmcma HmuOH .N< mHan

169

mean cumfiunom I GOHuaEDmcou ucmumaam eooS%Ha eoozuwome

u mumnm umxume eoo3%Hm wooBuwom cumummz

.auow Hmafiooe .ucoouoa cH

3
NM

.eoau

Iuneoua eumoeuonwm Nuwmcme Enfiema msHa cowunESmcoo ucmumaam eoo3mHa eooBuwom eam eooSNHa oooBeum: .eumon

GOHumHSmsH .eumoneum: .epmoemHoHuuma mo coHumEDm mcu ”mummmwe enmeamum .ummw mumsvm mo mGOHHHHB CH

\H

 

memo.

memo.

mooH.

«moo.

mewo.

ewmo.

aqwo.

eqwo.

OMNo.

wmeo.

MNeo.

meo.

«moN.

oHoN.

Nqu.

eHON.

Nme.

NweH.

emmH.

wqu.

HmmH.

quH.

equ.

mewo.

quN.

wwwN.

weom.

«New.

aNNm.

meHw.

momm.

qum.

wNem.

mmwm.

Hme.

wmmq.

«QN.Nm
oom.mm
qu.em
NmN.mm
Nme.wN
«He.ow
«ON.em
qu.em
Nee.Nm
Hem.NN
waq.NN

mmq.NN

onH

wNwH

NNoH

eNaH

mNmH

«NoH

mNmH

NNmH

HNmH

onH

memH

wemH

H.e.scooe Na magma

170

.e<IH< mmHan "moounow

.coHunEScho Hound Hmuou + cOHDQEDmcoo ucmumaam eumoemHoHuuma u mumsm umxuma enmoeoHoHuumm\W
.GOHuaEDmcoo Hmcma Hmuou + COHuoneoua eooBNHa mafia cpmnunom n oumnm uoxuma eooB%Ha mafia cumnusow.l

3

.COHumEDmcoo Hmcmm HmuOu + A aOHuoneoua eoo3NHa

i.e.uaouv Na magma

171

 

 

momH. mqu. oqu. eewH
quH. wNeH. mmmH. memH
NHmH. ommH. NemH. eeaH
mqu. NmmH. ommH. memH
mHeH. moeH. mqu. NemH
mmNH. NomH. oeNH. HemH
«HmH. eeqH. HNNH. oemH
Nme. quH. qqu. mmmH
meHN. oHeH. NmmH. wmmH
eONN. mqu. quH. NmmH
quN. Nqu. emmH. emmH
NomN. eNmH. mme. mmmH
.IeumonumeHw
\q %uwmcon
EaHemz \mepmon COHumHsmcH \Meumoeeumm \meooBNHm oooBeumm
mumsm moxumz “mow

 

.Aauow Hmeflooe .ucmouma CHV aNaHImmmH .eumonumeHw Nufimame
EsHeme ecm .eumoe coHumHsmcH .eumoneumn .eooBNHa eooaeuws How SOHuaanmcoo wo dunno umxumz

.w< mHnt

.cowuaasmcoo Hmamm HmuOu « GOHuaanmcoo ucmumamm whoop cowumHsmcH I dunno umxuma eumon aoHumaamcH I

 

172

.:0HuQEsmcoo Hocma HmuOu v cofiuaesmcoo ucmumaam eumoeeumn u mumzm umxuma eumopeummvm
.COHumasmcoo Hmcma HmuOu v COHuQESmcoo ucoummmm eooahHa eooBeumn u mumnm umxumfi eooBNHa eooSeumm\fl
NwHo. Hmwo. meN. mqwo. mNmH
NNHo. eqwo. meNN. mNmo. wNmH
HNHo. wooo. eNmH. NNmo. NNwH
mwoo. Nmmo. NooH. «moo. eNmH
mNoo. Nwmo. eme. mHoH. mNmH
oeoH. mHHN. mNmo. quH
mon. quH. ewHH. wNmH
oon. wNwH. quH. NNmH
quH. quH. memH. HNmH
mNHH. eHNH. «NmH. onH
NNmH. mmNH. oqu. memH
owNH. mmmH. quH. wemH
eemH. mqu. NmmH. NemH

H.e.scoov w< mHan

173

.N<IH¢ WQHDQH quUHflom

.GOHuaESmcoo
chma Hmuou « coHuoseoua eumoeumeflw NuHmcoe EDHemE u oumnw umxume eumoeumeflm euHmcme Enwemz I.

\q

H.e.ucooe m< mHan

174

HNeo. HmNo. wemH

 

 

oNeo. quo. NemH
memo. HNHo. eemH
emNo. oqoo. memH
mmNo. oooo. eemH
wNNo. memH
NoNo. NemH
wNHo. HemH
meo. oemH
meo. ammH
mmHo. wmmH
Nooo. waH
oeoc. emmH
eumoemHoHuumm woosmHa mcHa cuo£usow
\Mmpmem umxums emwme How»

 

.Aauom HmEHomo
.ucoouoa CHV mNmHIemmH .mume mumsw poxums eumoeoHoHuuma ecm eooBNHm mafia cumSusom emwme .m4 oHemH

.N< mHan "mouaow

.ouw .emmH u emwme NmoH .mnmsm poxumfi mmoH u mpmzm umxuma eowme emmH.I

 

175

\H
mmao. «mom. seas
mooa. Neom. mama
«moo. oHoN. Haas
memo. mesa. seas
ammo. News. mesa
memo. emma. «Nam
eewo. mass. mesa
omeo. HmmH. News
wmeo. acme. Haas
mmoo. oeoa. cess
mama. memo. some

H.e.ucouv m< mHan

176

 

 

wN em w.mee NemH
mN wm e.mqe eemH
NH Nm «.NHe meoH
m om m.NNm qemH
Hm N.mmm memH
mm e.HNm NewH
mm e.oom HemH
Hm m.Nwe oemH
we «.NNq mmoH
we o.mme wmmH
em a.mme NmmH
Nm @.q¢q emmH
mN m.qu mme
mucmHm eooSNHm mucmHa
mafia cumnusom eumoeoHoHuuma cmuwuoe< ImEoocH Hmcomuma
mo Honesz fiuuoz mo popesz \wHemmoamHe .m.D umow
.onHImme .mucmHa eoo3kHa mafia cum
Izusom mo gonad: ecm .mucmHa eoo3NHa cmofiume< Luuoz mo poeasc .msooaH Hmcomumn mHnmmoamfle .w.D .oH< mHemH

177

.NewH .mmHon ecm uswfip3 mome .moHanoom HmHuumnecH mo amwusw .oouoEEoo
mo ucoEuumaoo .w.D ”Ame mummm eouomHmm .msmcmo osu wo ammunm .moumaaoo mo ucmauumamn .m.D mmNmH
.moH>umw umouom .ousuHsopr< mo unmauummmm .m.D mome .uuermH mnNmmH .msoamco:< umNmH .comuoe:< "mmounow

.mumHHoe Neon mo mcoaaafln cww

 

NN NN w.¢mm mNmH
He NN m.NNm wNmH
mm Hw m.mNm NNmH
Nm mN w.Hmw eNmH
Nm NN N.mmw mNmH
mm NN o.qu quH
mm NN N.qmw NNmH
Nm oN w.How NNwH
we qe o.oeN HNoH
oq «e e.HqN ONmH
em Nm m.NHN memH
mm em N.mae wemH

H.e.u:ooe os< mHamH

178

 

 

 

H.mw m.me o.Nm m.mMN mNmH

w.em o.mN m.we m.mON wNmH

e.H¢ m.mN N.¢m N.qu NNmH

o.ww w.HN w.qm o.me eNmH

w.Nw w.Ne q.eq m.qNH mNmH

e.mw N.mN o.mm H.oeH quH

m.om H.wN m.we N.emH mNaH

o.mm w.mw w.Ne H.mHH NNoH

¢.ooH N.ww w.Ne o.qHH HNmH

H.ocH e.Nm «.me q.oHH oNoH

N.NoH N.ww m.moH m.eoH memH

m.ooH m.mm m.wm w.NoH wemH

o.ooH o.ooH o.ooH o.ooH NemH
\Meumon cowumHsmcH \meumoneumm \MeumoanoHuumm moHuHeoEEou HH< ummw

.AooH u NemHv meHINemH
.eumon coaumHnmCN ecm .eumoeeum: .eumoanoHuumn .mmHuHeoEEou HHm pow mmxmecH oofium nooseoum .HH< mHan

179

.mnmm% emuumHmm .moflumwumuw poemH wo smmusm .uoemH mo ucmEuumamn .w.: HNNwH .maHmsm “mmousow

cw emm: EfiuwumwOH

.mdoammouwou meow
.mmHuHeoEEoo HHm new xmecH ooHua nooseoua he emumeme xmecw mowua umoaeoum.l

\H

A.e.uaooe Hs< mHamH

180

 

 

e.oNH o.Ne N.mHH w.qu m.mmH wNaH
o.oNH w.me o.eHH e.qu w.NmH NNoH
o.moH m.ee o.HoH w.NmH m.mmH eNmH
m.em m.we q.ow o.eHH N.eHH mNmH
w.wm m.Hw o.qw e.NHH N.eHH «NoH
e.mHH N.ww H.moH o.qu o.qu MNaH
H.HHH e.Nw w.NoH «.mNH H.0mH NNmH
w.NoH N.ww w.ow w.oHH q.HHH HNwH
N.ww w.Nm H.Hw w.NoH w.NoH onH
N.qHH e.Nm o.coH w.NoH N.omH memH
m.mHH o.wm I o.eNH o.eNH wemH
o.ooH c.ooH I o.ooH o.ooH NemH
.I-Immu:uHumesm .IeooBNHa
\mmwmoanoHuuma .leoo3%Ha leooz%Ha \mooauHOm .IeooBNHQ
emustmz \wooseumm ocmm sumsunow cumumoz \mooauwow ummw
.AooH u NeoHv mNaHINemH . ImquuHumesm eumonoHoHuuma wounwwms ecm .eooBNHa eoozeumz
. leoo3NHa mcHa cumzusom .e003NHa eoobwwom auoumms .eoosta eooaumom pow mmxmesfi mowua umoneoum .NH< mHemH

I

181

.mummx emuuoHom .moHumHumum uoemH mo smmusm .uoemH mo unmauummon .m.: HNNmH .maHmsm "mmuunom

.moumsm uoxuma Ne
wouanms mmxoecfi moHua umoseoua eooBNHa eooseums ecm eoo3mHa eooBuwom .eumon :OHumHSmcH .eumonwumm.l

\m

.mcoflmmouwmu maom
CH emm: ELuHumwOH .mmwuweoafiou HHm you xmecH moflua Hmoseoua Ne eoumHeme xmecw moaum umo:eoum.l

\N
2: u $ma

 

q.on w.HN o.qm m.mmH H.NmH onH

A.e.ucooVHH< mHan

182

 

m.oNH mq.we N.wm we.N H.qu m.mo wNmH
N.eNH No.mm w.Nm Ne.N e.mmH N.HoH NNmH
m.mHH qe.om m.em we.N H.qu o.eoH eNmH
e.mm HN.¢w e.Nw we.N H.qu «.moH mNoH
N.ww oN.om e.qm qe.N H.omH w.mw «NNH
m.eoH mw.mN e.wm eN.N N.ww ¢.we wNwH
w.NoH we.wm N.Nm wN.N e.mm m.qN NNmH
o.wm mN.wq e.Nw eN.N N.coH o.wN HNmH
w.Nw we.Ne N.qw we.N m.m¢ H.Nw onH
e.¢w oN.Hm w.mm Ne.N m.qm o.qw meaH
m.HN I m.em Hm.N m.em N.ww weoH
m.me I N.ww Nm.N o.ooH o.ooH NewH
Iooaua

.INuH>HuoCeoua mmmaewuw .IxmeCH moHCa .IxoeCH moHCC

\e HHHE on3mm mCHC ICOHuoCeoua mu meoua emumHmC mmmmou oHummHC
eooshHa Cumnusow \q umnECH \mmmme eCm mHmCm eCm mm>Hmm£e< Cmmw

 

.onHINemH .mCOHumCUo NHCan CH lemma mama dH< mHan

I

183

.mpmoh emuomHom .mUHumHumum CoemH mo
Cmmusm .CoemH mo quEquamm .w.D mome .moHEOCoom HmHuumCeCH mo Cmmunm .monEEoo mo quEuumamn
.m.D mmNoH .moH>umw ammuom .oCCuHCoHCw< mo quEquCmm .m.D HNNmH .mCHmcm qu< eCm e< moHan “mmoudom

.mquHQ eoo3hHa mCHa Cumnunom mo ConECC mcu Np emeH>He COHuoseoum eoo3zHa mCHC Chonuaow\m
.ooH n memH .mmHuHeanoo HHm Cow xmeCH mUHua noose

IOCC msu %e emumHeme mummuom HmCOHumz Eoum mmoHHC mwmmECuw CmnBHuamm mCHC Cumcusow .mumHHoe CH\N

.ummH eumoe Ho mCOHHHHn CH\w

.ooH u NeoH .moHuHeoaaoo
HHm COM xoeCH ooHCC Cooseoum men we emomHeme mpmxuos mooseoua e003 eCm CerCH Cow meHCCmo NHCCom.I

 

\m
.AooH u NemHv mmHuHeoEEOU HHm new xmeCH moHCC Cmoaeoua msu CuHB emumHmon\M
.mume NuHquCv eCm moHCQ Cmnu Conu0\fl
m.eoH em.Hm o.em wm.N m.MNH N.ooH wNmH

A.e.ucoov mC< mHamH

184

ommH mNe mHmH NNwH

 

wqu Hmm NHNH eNmH
HNHH qu Nmm mNmH
mmmH NHq ewm «NmH
meN omw mmHH mNmH
aNmN owm mmmH NNmH
mwON eqw mwNH HNmH
meqH wNm How onH
oomH mHe Nww memH
memH New mwm weoH
NNmH eoe eHm NewH
emHH Nmm wwN eemH
onH wmq NmoH memH
HemH mom emoH qeaH
Imuumum ImuHCC +m ImuHCC NwH
wCHAmmBe HmuOH .mmwmum wCHmCom .mmwmum wCHmCom Com»

 

.AmuHC: mo meCmmCocu CHV mNmHIqemH .mCOHuvao eCmEme CH com: mume wCHmCO£ .m.D .qH4 mHan

185

.oHHeCC eCm mum>HCm.l

 

\m

.zHCo oum>HCm\M

.muHCC +m .muumum wCHmCo: mCCHE mupmum wCHHHm3e Hmu08\fl
quH mmq emNH mNmH
wNON mNm wqu wNmH

A.e.uCoov «HC mHan

186

 

e.me q.qqm.N q.e¢e.mN m.emH.¢ N.NNm.mm meH
N.mNH m.eNe.w N.eme.ON H.mom.o m.eqH.me wNwH
m.emH N.wwH.w m.wmm.wH m.omo.o ¢.meo.Nq NNmH
w.NqH m.Hem.o m.HmH.wH N.mmm.w H.qmw.Nm eNmH
e.me e.wwH.HH N.owo.mH w.mmw.N m.Hew.qN meH
m.eNH w.qu.HH H.oew.mN q.oqm.e N.wNo.Nm «NwH
N.on H.wH¢.HH m.eNm.mN N.Nme.e N.wNo.eq wNmH
H.ooH e.wwq.HH o.NHo.qN e.NHq.N N.qmw.qq NNmH
w.qm N.NNo.NH o.NHN.wN e.wwH.N m.mwm.em HNmH
e.ww N.wNo.NH N.NNH.QN H.emo.N o.mmm.NN oNoH
N.Nw N.mNm.mH e.owm.mN e.NHH.N e.Nem.Hm memH
H.eN w.NHN.mH e.wew.mN e.oem.e m.eNm.Hm wewH
«.NN w.NwN.mH N.qu.qN m.wew.N «.NNN.eN NeoH
xmeCH ouHmoaaoo meHeHHCe meHeHHCe mCOHumumuHm ICOHuUCCumCoo
muumaaoo mo oHHACC HmHqueHmmCCOC eCm mCOHuHeem \NHmHqueHmmH
quEuumCoo mo mCHm> mum>HCQ mo oCHm> HmHqueHmmu mo mCHm> mo mCHm> Cow»

 

.\flAmCmHHoe NNmH mo mCOHHHHE CHV mNmHINemH .mmumuw emuHCD mcu CH COHuuCCumCoo mo mCHm> Co mama .mH< mHan

187

.va mummm ewuumHmm .mCmCmu ozu mo Cmmusw .muumfifioo wo quEquCmn .w.D "mmuunom

.momHa CH use mCHm> .NHCO mom:on.3mZ\M

.ooH we eoHHaHuHCE eCm xmeCH muHmanoo ooCmEEoo mo quaquamo he emeH>He mCmHHoe quHuCu.l

\H

 

A.e.uaooe mCe mHan

188

NHN N.ww w.Nm o.ooH eemH

 

 

NHN «.mN N.ww o.HoH memH
HoH N.ww o.Hw H.wm qemH
HmH N.em w.mN m.mw memH
wHH m.mN w.HN e.ow NemH
om m.mN e.we e.NN HemH
«OH N.wH w.ee e.oN oemH
HNH N.mH N.ee w.HN mmaH
HoH o.eH e.Nw e.NN wmmH
HHH o.mH N.He o.me NmmH
oHH o.m w.ee w.ee emmH
cHH o.w N.wm N.ee mmmH
IwCHCCuommCCmE
ImquECHLm Imuumum wCHmCOL HHm IwCHCCuomeCmE \H mmusume
\m oEo: \wo owmuCooCmC m mm \H HHm eCm muCuHCHCm
mHHeoz muHCC NHHammHuHCZ wo xoeCH mo xoeCH Com»
.mNmHImmmH

.mCOHuvam eCmEme e003hHa mCHC Cumsusom eCm eumoemHoHuumC CH om: How common eCm com: mume Cmsuo .eHC mHemH

189

NNN

eNN

NNN

eqN

mHN

mNm

Nem

eNm

Noe

Hoe

qu

me

OQN

.ooH >3 emHHCHuHCE eCm muumum wCHHHmBe HmuOu he eoeH>He muHCC +m .muumum wCHmCom

N.wN

o.mN

m.wN

e.oq

N.ww

w.ee

w.em

N.ww

e.me

w.eeH

e.me

m.owH

w.eHH

«.mNH

w.0NH

a.wHH

N.on

q.eoH

o.HHH

q.eoH

o.ooH

w.HeH

w.mmH

o.meH

N.NmH

N.wHH

e.NmH

N.qu

e.omH

o.eHH

H.on

«.NoH

«.moH

o.ooH

muHCC mo meCmmConu CH

.ooH n NemH

\m

\N
D.

meH

wNmH

NNmH

eNmH

mNaH

«NaH

mNmH

NNoH

HNmH

onH

weaH

wemH

NemH

A.e.uCooV eH< mHan

190

.ome .moHEOCoom
HmHuumCeCH wo meusw .monEEoo wo uCoEquaoC .m.D ”mummN emuoonm .mHmNHmC< UHSOCoom mo Cmmunw
.moCmEEoo Ho quEuumCmo .w.D mmNmH .moH>me ummuom .mCCuHCoHCw¢ mo quEquCoQ .w.D H¢H< mHan "mmousow

A.e.uc00e ea< mHamH

191

 

mm.m wNm.N Noq.¢H e.Nw wNmH
Ne.N NaH.N NNw.wH e.ww NNmH
Hm.m wHH.N eoo.NH H.Nw eNmH
wo.m oeo.N eoN.mH m.qm mNmH
mo.w er.H eow.mH N.H¢ quH
wm.m Nme.H me.NH N.Nq mNmH
Ne.N eme.H mew.NH H.qq NNmH
HH.m woe.H wmm.eH N.Nm HNmH
Nw.N meq.H oeH.eH e.NN onH
Nw.N Nem.H www.mH H.om memH
I mNm.H mwm.qH m.mN wemH
I mwN.H oqw.NH m.mH NemH
aw”... fimfim. gamma ....flmma
eooBNHm ustma emouHHmm eoo3umow Com uCCuCO Com»

 

.mNmH
INemH .mCOHuvao NHQCCm eoo3zHC mCHn Cuocunom eCm eumoemHoHquC CH mm: Cow emummu mume Conuo .NH< oHemH

192

.ome .moHEOCoum HmHHumCeCH mo Cmmunw .muumfifioo wo quEquamQ .w.D moH< eCm m< .H< mmHemH "mmoHCom

.ooH u memH .mMHuHeoEEoo HHm COM xmeCH onua mHmmmHoz3
mcu Ne emumHeme meHCCmm kHCCo: mumHuoB COHuoCeoum eooBhHC eCm CmmCm> eoo3umow .mCmHHoe CH\.m
.mquo CH.I

\m

.mHmme fioCHIw\m .ummw dpmsum mo mCoHHHHa CH\M

.mquHa mo CmeECC we emeH>He COHuoCeoua UHummaom\fl

 

oq.m mNm.N wwm.wH w.mq mNmH

H.e.ucoov Nae mHan

193

 

 

 

m.mmm quH N.me qu NNm oNoH
e.omm wqu H.Nq me qoe wNoH
m.on quH o.oq weH wwm NNmH
N.mow NeHH o.om ew qeq eNoH
w.mwN Now o.He oe eem mNoH
o.HwN www w.Hq qu New quH
H.NNN NmHH N.Nq qwm qu wNoH
e.me oomH w.ee Hoe NHe NNoH
o.wMN HmHH N.ee mom eNm HNoH
o.oNN me w.ee NON NNm oNoH
o.oNN HHw N.Nq HNN New meoH
Ifiusom Imupmum Imuumum wCHmCo:
\msu CH \wCHmso: uHCmNmCo .w.D Hmuou mo .Isusow onu CH lzuoow
mEooCH uHCC oco quoCoa m mm nuCow moo CH m wmum wCHHHmse m£u\mH muumuw
HmComCmm .w.D Hmuoe muumum wCHmCon uHCC mCO uHCC +m wCHm50£ uHCC mCo Com»
.oNoHIoeoH .cusow mnu Ou onHuoCm mumo eCmEmQ .wH< oHemH

194

.mummh emuomHom .mHmNHmCm oHEOCoom wo CmmCCw .moumaaoo
mo quEuumamO .m.D “Ame whom» emuomem .mCmCmO mnu mo Cmmusm .moumano mo quEqume .m.D "mmousom

.OoH u NNoH .moCCuHOCmoxm
COHuaECmCoo Hmcompoa COM Coumeoe ooHCC uHUHHaEH one :oHs emumemo .mCmHHoo NNmH mo mCOHHHHe CH

\m

.oos an
emHHmHuHCE ecm muumum mCHmCOL uHCC mCo .m.D Hmuou we emeH>He Lunom mSu CH muumum mCHmCo: uHCC mCO\M
.muHGD mo mficmmDOLu CH\..IH-

A.e.u:00e ms< mHamH

195

 

eomN m o.Nw HHOHHoz\CmE:on
mmoN o o.mm Mao» Bmz
eHmm HH o.Nw HamHz
«EN 0 men «2325::
quN em o.<w mCmmHCO 3oz
8mm 8 N ...a mac—dam:
mme o N.wm oemHoa\uHouumo
NHHN o w.Om omeHno
NQON om m.oN cones .um\mmHHmO
mmNN om o.mw COCmCom
omeN oH m.mw mwOOCmuum£O\mquHu<
HNwN HH w.ew muCOHCmCO

wwumum quommem eoo3NHm eoo3umom

mmHHE CH CH eCm <92 CH HHm mo uCooCoC

mo .eCmHuCom
Bone moCmumHo

mquHo woosta mCHo
CCoLuCom Ho CmeECz

m on eooBNHa
mCHa Cumsusom

A<Hze mono wCHemuu
HOnmE NHHmzozIeCmm

 

.wNoH

.HoeoE ConCemHe HmHumom CH Odo: mumo .oH< mHemH

196

mwom

Now

wNNH

ooeH

mNH

Nme

oNo

quN

«wow

mqu

mOHN

NoNN

NmmN

weON

eOON

em

HH

om

HN

HH

w.NN

H.O

m.N

H.OH

0.0

0.0

o.o

0.0

N.wN

w.eN

e.Ne

N.Hw

w.Ne

e.Nw

e.ww

o.Ne

Coumow
NuHo memo UHmm

Cm>CoO

Hnmm .uw\mHHoammCCHz
mHuumom
eCmemo\oumHquum Cow
mmHme< mOH

eCmHuuom

xoom oHuuHH
wusemumumm .uw\maEmH
eCmHm>mHO

0HC0uC< me
mCoEHuHmm\C0uwCH£mm3
anwCHECHw
mHHH>Comxomo

mHsoa .om

A.e.ocoov mae mHan

197

.ONoH

.CompmeC< "mousom

 

eNON

N.ww

mmstmsHHz

H.e.uaoue ma< mHemH

198

 

.mquEoH:m eooBNHC wNoH HmuoH

1m

\C

uumum wCHmCom

emo.eHm Nem.mH mazes .um
moN.mNN mom.HH xHoHCoz\Cmaum£on
moN.oHN ooO.N xuow 3oz
qu.on NwH.oH HamHz
Cmm.maq NmH.mH «HeaHmemHHem
HNm.Nmm eoq.N mCmmHCo 3oz
omq.mem ooe.q mHnaEmz
on.mem qu.N~ oewHoH\oHoCumm
wmw.NmN ONm.Hm owmoHCO
MNw.Hem mNN.NN euuoz .uC\meHmo
ooH.MNm OHm.wm Coumso:
mHN.Nem mwo.ms mwooamuumeo\mocmHue
NNm.Noe qu.m wouOHCmso
mHmmeI:w\m .uooe
oCvam mo mCOHHHHE CH wNoH CH A<Hze mono wCHemuu

COmmE eHHszzIeCmm

 

.meH .Hmeoa aOHmsmmHe HmHomam was ca emummu mama umeuo .ome mHan

199

mee.on
woo.qu
Nee.Nom
Oom.MNq
Nmo.qmq
eNm.mHN
eew.Hmw

mqo.mqo

on.oeH.H

eom.HwH
omw.ooH
mmN.mNN
NwH.ooH
HNe.Nom
HNm.oHN

qoo.oqN

mNm.m
NNo.m
wmm.HH
qu.HN
omo.wH
eoN.mH
ewm.wH
mNo.eH
NoN.qH
mHm.H
qu.oH
on.e
mow.w
meN.wN
eNH.q

NNo.m

moxHCmaHHz

COumow

NoHo mama uHmm
um>Cmo

Hsmm .um\mHHommoCCHz
mHuummw
eCmemo\oomHoCmCm me
mmHowC< mOH

eCmHuCom

xoom mHuuHH
wusemumuom .um\mCEmH
eCmHm>mHO

OHCOuC< me
mCoEHuHmw\C0uwCH£mm3
EmcwCHEHCm

mHHH>Comxomo

H.e.ua00v o~< mHan

200

. m mummx emuum mm .mCmCmo m u o Cmmunm .moCmEEoo o quEuumamo .m.D m .Cowum C< “mmousow
V H n w w oNoH e

.muumum wCHmCos Cow Nxouo m mm com: mum3 meHHHmSO uHCC q eCm m.N.H Com eosmmH muHECma wCHeHHCm.I

\H

 

A.e.ocoov ome mHan

201

Table A21. Projected high, medium, and low levels of U.S. disposable
personal income, 1990 and 2000 (in billions of 1972 dol—

 

 

 

lars).
Year U.S. disposable personal income
High Medium Low
1990 1540 1450 1360
2000 2110 1880 1690

 

Source: U.S. Department of Agriculture, Forest Service, 1979.

REFERENCES CITED

REFERENCES CITED

Adames, Darius M. and Richard W. Haynes. 1980. The 1980 softwood
timber assessment market model: structure, projections, and pol-
icy simulations. Forest Science Monograph 22. Washington, D.C.:
Society of American Foresters.

Adams, Thomas. 1978. Production opportunities in the Pacific North-
west. in Structural Flakeboard from Forest Residues, USDA Forest
Service General Technical Report WO-5. Washington, D.C.: U.S.
Government Printing Office.

Akers, L. E. 1966. Particleboard and hardboard. Oxford, New York:
Pergamon Press.

Ambler, John. 1980. Personal communication regarding Current Industri-
al Reports, Series MA-24H, U.S. Department of Commerce, Bureau
of the Census.

Anderson, Robert G. 1978. Softwood plywood use in non-residential
construction. American Plywood Association Market Research Report
R41. Tacoma, Washington: American Plywood Association.

. 1979. Regional production and distribution pat-
terns of the softwood plywood industry. American Plywood Associ-
ation Market Research Report E27. Tacoma, Washington: American
Plywood Association.

 

. 1980. Plywood end-use marketing profiles 1979-
1981. American Plywood Association Market Research Report E28.
Tacoma, Washington: American Plywood Association.

 

Anonymous. 1957a. Fiber-clad building board plant. The Lumberman
84(11): 58-62, 71-72.

Anonymous. 1957b. Hardboard, particleboard, flakeboard plant direc-
tory. The Lumberman 84(5): 159-165.

Anonymous. 1957c. Two men operate entire production line in plant.
The Lumberman 84(5): 57-59.

Anonymous. 1962. Wizewood makes wafers from green poplar to form
a new board. The Lumberman and Wood Industries 89(6): 40-42.

202

203

Anonymous. 1966a. Los Angeles story spotlights problem of noncon-
formance. Forest Industries 93(1): 58-59.

Anonymous. 1966b. New plywood standard simplifies specification.
Forest Products Journal 16(12): 14.

Anonymous. 1971a. Board line uses rotary press. Forest Industries
98(7): 47.

Anonymous. 1971b. Associations backstop makers with promotion, tech-
nical help. Forest Industries 98(8): 40-41.

Anonymous. 1975. Oriented wood strand panels open new era in utili-
zation. Forest Industries 102(3): 46-48.

Anonymous. 1978. Softwood plywood, the south—-1967—1977. Forest
Industries 105(3): 30.

Anonymous. 1979a. PET, HDPE to lead plastics parage. Purchasing
World 23(12): 46-47.

Anonymous. 1979b. Upturn in production coming for composite, OSB
panels. Forest Industries 106(9): 11-12.

Anonymous. 1980a. Capital spending jumps in solid wood industry.
Forest Industries 107(1): 30-31.

Anonymous. 1980b. MDF plants, location, etc. (advertisement).
Forest Industries 107(4): 28.

Anonymous. 1980c. Particleboard: the versatile panel product of
the '80's. Wood and Wood Products 85(2): 12-13.

Anonymous. 1980d. Unrest at Georgia-Pacific. Business Week 2664:
147-148.

Anonymous. 1980e. Using particleboard and MDF II. Wood and Wood
Products 85(2): 8-95 (Part 2).

Anonymous. 1981a. Directory of panel plants--U.S. Forest Industries
108(4): 63-67.

Anonymous. 1981b. South's 'third forest' falls short of goals. For-
est Industries 108(1): 42.

Anonymous. l981c. Study, APA disagree on plywood industry's fate.
Forest Industries 108(1): W6.

Arola, R. A. 1978. Fiber for structural flakeboard--mechanized
thinning and topwood recovery. iE_Structural Flakeboard from
Forest Residues, USDA Forest Service General Technical Report
WO-S. Washington, D.C.: U.S. Government Printing Office.

204

Arrow, Kenneth J. 1962. The economic implications of learning by
doing. The Review of Economic Studies 29(3): 155-173.

Baldwin, Richard F. 1977. Southern pine plywood--from 0 in '64 to
33% of U.S. output in '76. Forest Industries 104(3): 40-43.

Ball, G. W., R. P. Redman, and A. D. Adams. 1979. Particleboard
binder promises performance without emissions. Forest Indus-
tries 106(4): 76-79.

Bentley, William R. 1970. Technological change in the forest indus-
tries--a problem analysis. The University of Wisconsin Forestry
Research Notes, No. 151.

Blackman, Ted. 1978. Electron beam system cures coated particleboard
panels. Forest Industries 105(3): 70-71.

1980. Commercial production starts on new composite
Elcoboard. Forest Industries 107(4): 54-55.

 

Brown, Daniel H. 1979. How building codes affect wood markets. For-
est Products Journal 29(1): 14—15.

Bryan, Richard W. 1978. Continuous rotary process makes thin particle-
board for wall paneling. Forest Industries 105(3): 52.

. 1981. Composite pine panel plant is the last word
in versatility. Forest Industries 108(4): 34-37.

 

Buongiorno, Joseph and R.A. Oliveira. 1977. Growth of the particle-
board share of production of wood-based panels in industrialized
countries. Canadian Journal of Forestry Research 7(2): 383-
391 O

Buzzell, Robert D. 1966. Competitive behavior and product life cycles.
ig_New Ideas for Successful Marketing, edited by John S. Wright
and Jack L. Goldstucker. Chicago, Illinois: American Marketing
Association.

Carney, Michael. 1977. Softwood plywood used in new residential con-
struction--1976. American Plywood Association Market Research
Report R38. Tacoma, Washington: American Plywood Association.

Coppens, H. A. E., M. A. Santana, and F. J. Pastorze. 1980. Tannin
formaldehyde adhesive for exterior grade plywood and particleboard
manufacture. Forest Products Journal 30(4): 38-42.

Cour, Robert M. 1955. The plywood age, a history of the fir plywood
industry's first fifty years. Portland, Oregon: The Douglas
Fir Plywood Association.

Cox, Lionel A. 1974. Transfer of science and technology in success-
ful innovation. Forest Products Journal 24(9): 44-48.

205

DeLess, Bruce C. 1975. Automated layup line saves wood, cuts costs.
ig_Modern Plywood Techniques, Volume 2: Proceedings of the second
Plywood Clinic, New Orleans, Louisiana, November 1974. San Fran-
cisco, California: Miller Freeman Publications, Inc.

Dickerhoof, H. Edward. 1976. Particleboard production, markets, and

raw materials in the United States. Forest Products Journal
26(10): 16-20.

1977. Particleboard production, markets, and
raw material in the South. USDA Forest Service Research Paper
SE-157. Asheville, N.C.: Southeastern Forest Experiment Station.

 

1978. Use of wood in mobile homes is increas-
ing. U.S. Department of Agriculture Resource Bulletin FPL 4.
Madison, Wisconsin: Forest Products Laboratory.

 

Dickerhoof, H. Edward and Thomas C. Marcin. 1978. Factors influenc-
ing market potential for structural flakeboard. ig_Structural
Flakeboard from Forest Residues, USDA Forest Service General
Technical Report WO—5. Washington, D.C.: U.S. Government
Printing Office.

Dickerhoof, H. Edward and David B. McKeever. 1979. Particleboard,
medium-density fiberboard, and mende process board plants in
the United States--capacity, production, and raw material trends,
1956-1976. USDA Forest Service Resource Bulletin FPL-6. Madison,
Wisconsin: Forest Products Laboratory.

Durbin, J. and G.S. Watson. 1951. Testing for serial correlation
in least squares regression, II. Biometrika 38: 159-178.

Egan, Mary Lou. 1980. Wood products companies near energy indepen-
dence. Forest Industries 107(9): 79.

Elmendorf, Armin. 1956. New pressed board developments are reviewed.
The Lumberman 83(3): 178.

Farris, Mary R. 1978. The veneer and plywood industry: above aver-
age productivity gains. Monthly Labor Review 101(9): 26-30.

Fassnacht, Donald L. 1064. The story behind the new southern pine
plywood industry. Forest Products Journal 14(1): 21-22.

1965. Southern pine plywood. Southern Lumber-
man 210(2616): 15—16.

 

Food and Agriculture Organization of the United Nations. 1976. Pro-
ceedings of the world consultation on wood-based panels. Brus-
sels: Miller Freeman Publications.

Frashour, Ronald. 1961. An approach to particleboard standards.
The Lumberman 88(7): 64-66.

206

Geimer, Robert L. and Eddie W. Price. 1978. Construction variables
considered in the fabrication of a structural flakeboard. In
Structural Flakeboard from Forest Residues, USDA Forest Service
General Technical Report WO-S. Washington, D.C.: U.S. Govern-
ment Printing Office.

Greenwald, Douglas and Associates. 1973. The McGraw-Hill dictionary
of modern economics. Second edition. New York, New York:
McGraw-Hill Book Company.

Griliches, Zvi. 1957. Hybrid corn: an exploration in the economics
of technological change. Economietrica 25(4): 501-522.

Hair, Dwight. 1967. Use of regression equations for projecting trends
in demand for paper and board. USDA Forest Resource Report No.
18. Washington, D.C.: U.S. Government Printing Office.

1980. Timber situation in the United States—-1952-2030.
Journal of Forestry 78(11): 683-686.

 

Hall, Bronwyn and Robert Hall. 1978. Time series processor (TSP),
version 2.8, user's manual. Available at Michigan State Univer-
sity Computer Center.

Haynes, Richard W. and Darius M. Adams. 1979. Possible changes in
regional forest product output and consumption during the next
50 years. Forest Products Journal 29(10): 75-80.

Hirshleifer, Jack. 1976. Price theory and applications. Englewood
Cliffs, New Jersey: Prentice-Hall, Inc.

Holley, D. L. 1969. Plywood's ability to compete for southern pine
stumpage. Southern Lumberman 219(2728): 177—179.

Irland, Lloyd c. 1972. Labor trends in southern forest industries,
1950—1969. USDA Forest Service Research Paper SO-81. New
Orleans, La.: Southern Forest Experiment Station.

Johnston, John. 1972. Econometric methods. Second edition. New
York, New York: McGraw—Hill Book Company.

Jorgensen, Richard N. 1978. Requirements, codes, and Forest Service
goals, 13 Structural Flakeboard from Forest Residures, USDA For-
est Service General Technical Report WO-5. Washington, D.C.:
U.S. Government Printing Office.

Kmenta, Jan. 1971. Elements of econometics. New York, New York:
MacMillan Publishing Co., Inc.

Koch, Peter. 1972. Utilization of the southern pines. USDA Agricul-
tural Handbook No. 420. Washington, D.C.: U.S. Government Print-
ing Office.

207

Koch, Peter. 1978. Two methods of acquiring residual wood for south-
ern flakeboard plants--the shaping-lathe headrig and the mobile
chipper. in Structural Flakeboard from Forest Residues, USDA
Forest Service General Technical Report WO-5. Washington, D.C.:
U. 8. Government Printing Office.

Kotler, Phillip. 1976. Marketing management. Third edition. Engle—
wood Cliffs, N. J.: Prentice-Hall, Inc.

Lambert, Herb. 1980. Late '79 downturn in housing had modest effect
on panels. Forest Industries 107(4): 30-33.

1981. Understatement: panels did not score high in
1980. Forest Industries 108(4): 30—33.

 

Lancaster, Kelvin. 1971. Consumer demand: a new approach. New
York: Columbia University Press.

Leefers, Larry A. 1979. Forces affecting technological change, inno-
vation, and diffusion: a literature survey. Survey prepared
for U. S. Forest Products Laboratory under Extramural Research
Agreement 12-146.

Lehmann, W. F. and H. E. Walhgren. 1978. Status and prospects of
residue utilization in board manufacture in the United States
and Canada. Forest Products Journal 28(7): 24-29.

Lekvall, Per and Clas Wahlbin. 1973. A study of some assumptions
underlying innovation diffusion functions. Swedish Journal of
Economics 75(4): 362-377.

Lewis, Wayne C. 1965. Wood-residue utilization. Forest Products
Journal 15(8): 303-307.

Maddala, G. S. 1977. Econometrics. New York: McGraw-Hill, Inc.
Maloney, Thomas M. 1977. Modern particleboard and dry-process fiber-
board and manufacturing. San Francisco, California: Miller

Freeman Publications, Inc.

Mansfield, Edwin. 1961. Technical change and the rate of imitation.
Econometrica 29(4): 741-766.

. 1963. The speed of response of firms to new tech-
niques. Quarterly Journal of Economics 77(2): 290-311.

 

Manthy, Robert S. 1974. Opportunities for technological advance in
the timber industry: an assessment. Lansing, Michigan: Green-
tree Associates. Unnumbered monolith.

Marra, G.G. 1969. A vision of the future for particleboard. in
Proceedings of the Washington State University Particleboard
Symposium, No. 3. Pullman, Washington: Washington State Univer-
sity.

208

Mast, Fred R. and David A. Pease. 1969. Suddenly they're everywhere—-
automatic panel layup systems abound as wraps are taken off. For-
est Industries 96(1): 50-52.

McCarthy, E. Jerome. 1975. Basic marketing: a managerial approach.
Fifth edition. Homewood, Illinois: Richard D. Irwin, Inc.

McKeever, David B. 1979. Hardboard and insulation board plants in
the United States-—capacity, production, and raw material require-
ments, 1955-1978. USDA Forest Service Resource Bulletin FPL-
7. Madison, Wisconsin: Forest Products Laboratory.

McKillop, W. L. M. 1967. Supply and demand for forest products--
an econometric study. Hilgardia 38: 1-132.

McKillop, William, Thomas W. Stuart and Peter J. Geissler. 1980.
Competition between wood products and substitute structural pro—
ducts: an econometric analysis. Forest Science 26(1): 134-148.

McSwain, George A. 1977. Technical developments in the wood-based
panel products industry. Paper presented by George A. McSwain,
Assistant Director, Forest Products Laboratory, Forest Service,

U. S. Department of Agriculture to the Food and Agriculture Organ-
ization of the United Nation's Committee on Wood-based Panel Pro—
ducts, Fifth Session in Rome, Italy in November 1977.

Mills, Thomas J. and R. S. Manthy. 1974. An econometric analysis
of market factors determining supply and demand for softwood lum-

ber. Michigan State University Natural Resources Research Report
238.

Moslemi, A. A. 1974. Particleboard. Volume 1. Carbondale, Illinois:
Southern Illinois University Press.

Mottet, Arthur. 1962. Particleboard now, tomorrow. Forest Industries
89(7): 94-96.

National Research Council. 1978. Technological innovation and forces
for change in the mineral industry. Washington, D. C.: National
Academy of Sciences.

Neter, John and William Wasserman. 1974. Applied linear statistical
models. Homewood, Illinois: Richard D. Irwin, Inc.

Oliver, F. R. 1964. Methods of estimating the logistic growth func-
tion. Applied Statistics 12(2): 57-66.

Pease, David. 1977. Computers raise output of particleboard line.
Forest Industries 104(7): 62-63.

. 1980. Structural panels in limelight as capacity
growth approaches. Forest Industries 107(7): 52—53.

 

209

Pease, David. 1981. Waferboard plant finds market growing despite
housing drop. Forest Industries 108(4): 54-57.

Phelps, Robert B. 1977. The demand and price situation for forest
products, 1976-77. USDA Forest Service Miscellaneous Publication
No. 1357. Washington, D.C.: U. S. Government Printing Office.

Price, Eddie W. and William F. Lehmann. 1978. Flaking alternatives
ig_Structural Flakeboard from Forest Residues, USDA Forest Ser-
vice General Technical Report WO-5. Washington, D.C.: U.S. Gov-
ernment Printing Office.

Quandt, Richard E. 1964. Estimating the effectiveness of advertising:
some pitfalls in econometric analysis. Journal of Marketing Re—
search I(2): 51-60.

Reid, William H. 1958. The particleboard industry. The Lumberman
85(7): 38-41.

Risbrudt, Christopher D. 1979. Past and future technological change
in the U. S. forest industries. Unpublished Ph.D. dissertation,
Michigan State University.

. 1980. Personal communication regarding par-
ticleboard furnish, U. S. Department of Agriculture, Forest Pro—
ducts Laboratory.

 

Rogers, Everett and Floyd Shoemaker. 1971. Communication of innova-
tions. New York: The Free Press.

Schmookler, Jacob. 1966. Invention and economic growth. Cambridge,
Mass.: Harvard University Press.

Stern, Louis. 1964. Market share determination: a low cost approach.
Journal of Marketing Research 1(3): 40-45.

Stone, Robert N. and Thomas C. Marcin. 1978. Surveys for forest re-
source demand analysis. Paper presented at the Society of Ameri-
can Foresters' National Workshop on Integrated Inventories of
Renewable Natural Resources in Tucson, Arizona, January 9, 1978.

TenWolde, Anton. 1979. Personal communication regarding purchased
energy in the wood products industries, University of Wisconsin.

U.S. Department of Agriculture, Forest Products Laboratory. 1974.
Wood handbook: wood as an engineering material. USDA Agricul-
ture Handbook No. 72, revised. Washington, D.C.: U.S. Printing
Office.

U.S. Department of Agriculture, Forest Service. 1974. The outlook
for timber in the United States. USDA Forest Resource Report
No. 20. Washington, D.C.: U.S. Government Printing Office.

210

. Department of Agriculture, Forest Service. 1977. The nation's

renewable resources--an assessment, 1975. USDA Forest Resource
Report No. 21. Washington, D.C.: U.S. Government Printing Office.

. Department of Agriculture, Forest Service. 1978. Structural

flakeboard from forest residues. USDA Forest Service General
Technical Report WO-5. Washington, D.C.: U.S. Government Print-
ing Office.

Department of Agriculture, Forest Service. 1979. An analysis
of the timber situation in the United States, 1952—2030. Review
draft. Washington, D.C.: U.S. Government Printing Office.

Department of Agriculture, Forest Service. 1980. An assessment
of the forest and range land situation in the United States.
USDA Forest Service, Fs-345. Washington, D.C.: U.S. Government
Printing Office.

. Department of Commerce, Bureau of Industrial Economics. 1980.

Statistical series--tables. Forest Products Review 36(1): 27-50.

Department of Commerce, Bureau of the Census. Selected years
(a). Particleboard. Current Industrial Reports, Series MA-24L.
Washington, D.C.: U.S. Government Printing Office.

Department of Commerce, Bureau of the Census. Selected years
(b). Softwood plywood. Current Industrial Reports, U.S. Govern-
ment Printing Office.

. Department of Commerce, Bureau of the Census. Selected years

(c). U.S. exports--commodity by country. FT 410. Washington,
D.C.: U.S. Government Printing Office.

. Department of Commerce, Bureau of the Census. Selected years

(d). U.S. imports--commodity by country. FT 135. Washington,
D.C.: U.S. Government Printing Office.

. Department of Commerce, Bureau of the Census. Selected years

(e). Housing starts. Construction Reports C20. Washington,
D.C.: U.S. Government Printing Office.

. Department of Commerce, Bureau of the Census. Selected years

(f). Value of new construction. Construction Reports C30. Wash-
ington, D.C.: U.S. Government Printing Office.

Department of Commerce, Bureau of Economic Analysis. Selected
years. Survey of current business. Washington, D.C.: U.S.
Government Printing Office.

Department of Commerce, National Bureau of Standards. 1966a.
Mat-formed wood particleboard. NBS Commercial Standard 236—66.
Washington, D.C.: U.S. Government Printing Office.

211

U.S. Department of Commerce, National Bureau of Standards. 1966b.
Softwood Plywood, Construction and Industrial. Product Standard
PS 1-66. Washington, D.C.: U.S. Government Printing Office.

U.S. Department of Commerce, National Bureau of Standards. 1974.
Construction and industrial plywood. Voluntary Product Standard
PSI-74. Washington, D.C.: U.S. Government Printing Office.

U.S. Department of Commerce, National Bureau of Standards. 1977.
Tabulation of voluntary standards and certification programs for
consumer products. NBS Technical Note 948. Washington, D.C.:
U.S. Government Printing Office.

U.S. Department of Commerce, Census of Manufactures, 1972. 1975a.
Industry series: millwork, plywood, and structural wood members,
n.e.c., MC72(2)-24B. Washington, D.C.: U.S. Government Printing
Office.

U.S. Department of Commerce, Census of Manufactures, 1972. 1975b.
Industry series: wooden containers and miscellaneous wood pro-
ducts, MC72(2)-24C. Washington, D.C.: U.S. Government Printing
Office.

U.S. Department of Commerce, Census of Manufactures, 1977. 1980a.
Industry series: millwork, plywood, and structural wood members,
n.e.c., MC77-I-24B. Washington, D.C.: U.S. Government Printing
Office.

U.S. Department of Commerce, Census of Manufactures, 1977. 1980b.
Industry series: wooden containers and miscellaneous wood pro-
ducts, MC77—I-24C. Washington, D.C.: U.S. Government Printing
Office.

U.S. Department of Labor, Bureau of Labor Statistics. Selected years.
Producer prices and price indexes. Washington, D.C.: U.S. Govern-
ment Printing Office.

Ventre, Francis T. 1979. Innovation in residential construction.
Technology Review 82(2): 50—59.

Vogel, B. M. 1976. Standards referenced in selected building codes.
U.S. Department of Commerce, National Bureau of Standards, NBSIR-
1140. Washington, D.C.: U.S. Government Printing Office.

Warner, K. E. 1974. The need for some innovative concepts of innova—
tion: an examination of research on the diffusion of innovations.
Policy Sciences 5: 433—451.

Weiss, Doyle L. 1968. Determinants of market share. Journal of
Marketing Research V(3): 290—295.

Wilson, J.B. 1975. Outlook for particleboard to 1980. Forest Pro-
ducts Journal 25(11): 10-16.

212

Wisdom, Harold W. 1977. The impact of the environmental and energy
crisis on the U.S. timber supply. Paper presented at the Third
World Pallet Congress held at Virginia Polytechnic Institute and
State University, Blacksburg, Virginia, October 3-5, 1977.

Wright, Maurice G. and Robert B. Phelps. 1967. Trends in particle-
board, insulation board, and hardboard industries, 1956-66. USDA
Forest Service Research Paper WO-S. Washington, D.C.: U.S.
Government Printing Office.

Zivnuska, John A. 1963. The future of wood in a competitive market.
Forest Industries 90(10): 40-41, 101, 104—106.