DYNAMIC MODELLING AND POLICY ANALYSIS ON TECHNOLOGICAL PROGRESS
AND THE CHANGE OF INDUSTRIAL STRUCTURE
Qingrui Xu
Dept. of Industrial Eng. /Management
Zhejiang University, Hangzhou, P.R.C.
Hanping Li
Dept. of Transportation Eng. /Management
Changsha Communications Institute
Changsha, P.R.China
ABSTRACT
Based on the mechanism of interaction among R&D, technological
progress, the change of industrial structure, and economic growth in
both supply and demand sides, and with the help of dynamic input-
output analysis, in this paper a system dynamics model is constructed,
focusing on the notable impacts of technological change on structural
change in the Chinese economy. Through modelling and policy
analysis,some new findings and patterns of long-term development
including the preferential consequences and opportunity for industrial
development, fulure prospects for China’s S&T and economy during the
next fifty years, the evoluation of industrial structure in the
process of industrialization, and resource allocation to different
industries and R&D expenditure allocation, are obtained.
I. INTRODUCTION
Compared with Hollis B. Chenery’s "standard structure" or "general
large country structure", and based on international comparisions of
industrial structure, China, though a typical low-income large
country, is characterized by an upper-middle-income country with 1000
or more U.S. dollor GNP per capita in terms of the share of industry
in production structure (about 42.6% of GDP in the year 1981). On the
other hand, looking at the share of the labor force in agriculture
(69.6% in 1982), it has the characteristics of a typical low-income
country.. The contradiction and abnormal devirations are related to
defects in China’s industrialization development strategy. During the
past decades, China’s economic growth had been fueled mainly by the
growth requirement of self-ordering heavy industry, and caused the
inbalanced or irrational patterns of industrial structure.Less~
developed agriculture and basic industry acted as a bottleneck have
restrained the growth of the entire economy, and equipment-
manufaturing industry expanded so quickly that the raw material
industry & energy industry and infrastructure could not satisfy its
overdevelopment. At the same time, the consumption goods industry had
been neglected and were usually in serious short supply. It is
estimated that the idle production capacity of the machinebuilding
industry is as high as 20% to 30% because of the barriers of the
unbalanced structure.
Meanwhile, China’s industrialization had been pushed forward thorough-
ly in urban areas. Moreover, most of the country’s population (about
70-80%) was in rural areas producing small portion of national income,
so that the rural population was isolated from the industrialization
575,
process. Many development economists believed that in developing
countries Lhe. economic development process is a shift of surplus labor
from traditional sectors (agriculture with zero or even negetive
marginal productivity of labor owing to its unlimited supplies of
labor) lo modern sectors (manufacturing,etc.) according to the dualism
theory. It is inadequate simply to depend on the expansion of modern
sectors to absorb agricultural labor. Moreover, the expansion of rural
nonagricultural sectors should be included and accelerated. Hence
rural or rural population industrialization become a part of the
entire industrialization. In this way, the deviration or defect
occured during the past course of China’s industrialization.
The lower income per capita is derived from an unbalanced or disequi-
librium and lower-level of industrial structure olher than lower
productivity or lower industrial technical level.In brief, the
crux lies in the lack of coordination or cooperation between the
output structure and demand structure. Furthermore, the profound
reasons hidden behind the unbalanced or disequilibrium structure are
that the transfering capacity of the production structure hindered by
an industrial technology system can not be content with the sharp
change of consumption structure.
Hollis B. Chenery and M. Syrquin concluded, in their "Pattern of
Development 1950-1970", that Engel’s coefficient (the income elastici-
ty of demand for food) demonstrates a significant downward trend as
per capita. income rises over time especially within the areas of 100-
1000 U.S. dollar GNP per capita, in other words, the weight of
consumption of food falls rapidlly. Therefore, low-income countries
will enter the stage of the greater change of industrial structure,
Large countries, however, will do so even at a lower-income level (per
capila GNP estimaled at 270 U.S. dollars or so). Thus, in China (300-
400 U.S. dollars) what is happening is something towards the new
development stage characterized by the great change of industrial
structure. In addision, Petty-Clark’s law and Simon Kuznets’ study
revealed that a country’s economic development is closely associated
with the change of industrial structure, hence growth process is just
one of structural change.
The change of industrial structure may be viewed as the resulis of
overall interactions of such multi-factors as technology, natural
resources ,economy and society. In short, the prevailing factors would
be summarized as both supply and demand factors. The structural change
will appear only when one or all of these factors begins 1o change.
Furthemore, technical factors have a direct effect on both supply and
demand factors, and then further on the structural change. Consequant-
ly, the structure of the technological system sustains or restrains
the change of industrial structure. Interindustry linkages, in fact
are the interration and interdependence among industrial technology.
As a result, technical advance is the governing factor affecting the
structural change.
Therefore, in this paper, the train of thought is developed from
Research & Development (R&D) and innovation to technological progress
to the change of industrial structure and finally to economic
growth, throughout which both supply and demand sides are taken into
account respectively.
In analyzing sources or factors of economic growth, both Edward F.
Denison’s and Hollis B. Chenery’s work is well-known. Denison’s work
attempted 1o enlarge as many kinds of input factors as possible and to
estimate the contribution of various input factors, and output per
unit of input, to growth. Obvisously, Denison, in measuring the
contribution of technical advance to growth, adopted, on the whole, a
576
"plack box" methed. This, as Denison noted, is the measurement of
our ignorance. On the contrary, Chenery’s work applies input-output
analysis o measuring the impacts of demand factors (consumption,
investment, import & export,technical change,ete.) on structural
change and on growth. Furthemore, increases or decreases in input
requirements, represented by changes in input and inverse
coefficients, are interpreted as technological change.
Generally, both works are based on supply and demand side repectively.
Hence, this analysis of single side is definitely limited and
incomplete in portraying the picture of structural change in the
future, as structural change and growth are regarded as overall
effects of the combination of both sides, Therefore, in this paper
both side factors will be taken into comprehensive consideration by
means of System Dynamics.
Il. MODEL DESCRIPTION: AND MECHANISM
S&T and its Application in Economy System
As the staple of S&T activities, R&D is also regarded as the input-
output process in which inputs of various resources bring about R&D
achievement outputs (see Fig.1). Only through its application to the
production process do R&D achievements become practical production
capacity which will lead to technical progress, otherwise, potential
productivity.Consequantly, technical advance is mainly derived from
the application of achievements, in both self-developed R&D and
technology acquisition, to economic system.
R&D ma el reer very
___jApplied | Technology}
——“"tResearch! Diffusion
i E.D be
Achievements Achievements Achievements {Acquisition}
Note: S--Supply support, D--Demand
Fig.1 R&D Input-Output Process and its Application
Factors influencing the application of achievements,that is,the jump
of potential towards realistic technology, are shown as follows:
(1) Demand for Achievements
System and Environment They determine the interest of enterprice
in technical innovation and the demand for achievements, strong or
weak. In particular, the extent of the perfection of market compe~
tition and the integration of S&T with production make a direct or
undirect effect on the demand intensity.
8I7 }
* Investment From the point of view of Soviet Union scholars, the
application of achievements is achieved only through capital
investment in which up-to-date technology is usually embodied, Thus
amount of capital investment, particularly, that in technological
transformation will determine the potential demand of the economic
system for achievements.
* The absorptive capacity for technology. The causal relationship
between industrial technological level and the absorptive capacity
for achievements indicates that the lower technological level is an
obstacle to the enhancement of the absorptive capacity, and vice-
versa. The weak absorplive capacity hinders increase of technical
level. In this way a vicious circle is maintained which prevents
economic growth. The successful experiences of some developing
countries show that it is effective to break away from this
predicament by improving the quality of human capital through
education and technical training. Especially while developing
countries are faced with the insurmountable technical gap, techno-
logy acquision is neccessary for economic take-off.
(2) Supply Side of Achievements
« Qualities of Achievement: advanced and appropriate or not,and
potential economic effectiveness, etc..
* Reserves of achievements. Through proper choices achievements may
be applied to the economic system. Accordingly, the more reserves
of achievements there are, the wider choice margins there are, and
the higher the choice efficiency is.
As a result, Fig.2 presents the interaction mechanism of S&T and its
ele erneere
<e Seu Applied [R&D
System Achievements, Subsystem
+
Industrial 4 Absotbtive R&D Achievement
Technical —___——= Capacity for Reserves:
level Technology
Investment in +Potential Choice
Technological ~~ Demand for Efficiency
Transformation Achievements
Investment in Abundant +
Capital Degree of
Construction “R&D Achievements
Fig.2 Application Process of R&D Achievements
Technological Progress, Structural Change, and Resource Allocation
Mechanism
Technological progress has a directly notable impact on both demand
578
and supply factors, and then further on the structural change. In
demand side, the structure factors include intermediate demand, final
consumption, final investment demand, and export. Here the more
detailed discussion is only on the intermediate demand structure,
Based on the international comparisions of material use, China’s
consumption of raw material and energy is especially high. For
example, its energy use is 5.69 times as large as Japan’s and its
steel material use is 2.8 times Japan and twice United States. In
addition to China’s special GDP structure, the lower intermediate use
efficiency, which is brought about by the lower industrial technology
level and backward management, are the major reasons for this
imbalanced usage. According to interindustry analysis, changes in
input coefficients can be caused by many factors, such as vertical
concentration over time, often combined with a changing product mix
and shifts in industrial product composition. Nevertheless, the
prevailing factor is technical change, which will make the change in
input coefficients tend downwards.
In supply side, dominant factors consist of supply structure and
allocation structure of various resources other than industrial
technology. There are two basic types of resource allocation
mechanisms or economic operational mechanisms--goverment planning
regulation and market regulation. Among government planning
vregulation,demand management is basically composed of fiscal and
monetary policy, and supply management is mainly made up of industrial
policies ( industrial structure policy,industrial organization policy
and industrial technology policy, etc.). Industrial policies can
link up planning regulation with market regulation. Besides, govern-
ment regulation mechanism should work in coordination with market
mechanism as prevailing hand in order to move the structural change
into a higher and more balanced level. In the meantime,government
regulation should wake up for the failure and functional weakness of
market regulation. What’s more, however, it ought to ensure coordina~
tive development among S&T,economy and society, and finally carry out
the national development strategy. To sum up, factors having an effect
on structural change and growth are decribed in Fig.3, and the overall
interaction machanism is given in Fig.4.
{Labor and Capital |
tGlinate Natural
Resources and
Ecological
tEnvironment
{State of Industrial;
1 Structure :
Total factor
Productivity
Technological factors
{Intermediate Demand
jTechnical Equipment } }Consumption Demand
iProduction Techniques} {Final Investment Demand;
jManagement Techniques; {Net Export
Fig.3 Dominant Factors Effecting Economic Growth
579
{Per Capita} ~ !Government}
{GNP i <fgseripansiy— 4Regulation!
{Mechanism |
{Progress | {Structure
S&T appre _ m
Subsystem; iSupply ee ure
Resources | Market
Allocation } !Mechanism;}
& Ulilization}
} Supply =
policy ! {Regulation!
{National Development Strategy in!
S&T, Economy and Society :
Fig.4 Interaction Mechanism of S&T and Economy System
Model Description
With ‘the help of a dynamic input-output model and on the basis of the
‘coordinative development between S&T and economy, a system dynamics
model is constructed, focusing on the notable impacts of technical
change on structural change. This SD model is composed of six
subsystems, including S&T, agriculture ( abbreviated as ‘A’ ), final
consumption goods industry ( ‘FC’ ), intermediate goods industry (
i.e. energy & raw material industry, abbreviated as ‘I’ ), final
investment goods industry ( i.e. equipment-manufacturing industry,
abbreviated as ‘FI’ ), and construction industry & infrastructure
(‘CI’). The above-mentioned industry classification is consistent with
MPS. (i.e. The System of Balances of the National Economy ), and is
based on the Chenery-Watanabe four-way categorization of productive
sectors for China (1981) and the analysis of interindustrial linkage
effects.. These effects are measured by both forward and backward
linkage effect in terms of China’s input-outpul tables available for
1981 and 1983. The interaction between the aggregate system structure
and its environment is shown as follows (Fig.5).
As industrial subsystem is similar to each other, we will concentrate
upon the description of the final investment goods industry. The more
detailed causal-loop diagram of machinebuilding industry is based on
the simplified one(Fig.6).
580
International Environment
Technology Comsumption
Transfer Demonstration Effect
Aggregate Value
of Micro~Econom:
“4
International Environment
Fig.5 System Structure and Its Environment
Import Demand Management
‘ \.
Technology Total -Descripancy 4 Total
Acquisition Supply —” Between Total *——— Demand
?
Demand and
Total Supply
4
Total Factor _+Total _ Flows of Industrial
Productivity Output: Factor Inputs, Policy
Barrier from
Basic Industry .
S&T Stock of 4,
Subsystem Productio;
Factors
Fig.6 Simplified Causal-Loop Diagram of Final
Investment Goods Industry
Five industrial subsystems are inlegraled with S&T subsystem and
aggregate values of micro-economy.
IIl. POLICY TEST ANALYSIS
Basic Behaviours in Base Run
Under a sel of policies and economic conditions, such as the coordina-
live development belween S&T and economy, demand orientations, the
middle accumulation rate at a level of 29% to 31%, and the proper or
moderate demand expansion, the system behaviors in the fifty year span
from 1985 to 2035 are simulated as follows:
« National Income (Fig.7). Annual growth rate of national income
will fluctuate within a 4-7 percent range. However, during the late
stage it will demonstrate a downward trend. Per capita income (yen
581
at 1980 prices) will go from 533 in 1985 up to 1260 or so in 2000.
+4200. 6
100.007
National
Income
07000
75.00F
Annual Growth-Rate
of National Income
06000
50.007
25.007
-03000
a
z
Fig.7 National Income and Its Growth-Rate, 1985-2035
(NI: Hundred Million Yen at 1980 Prices)
* Production Structure (Fig.8). The proportion of agriculture
declines continuously and its gross output value constitutes about
12.8% of total products of society around the year 2035. Converse-
ly, construction industry & infrastructure grows most rapidly since
supported by a vigorous or excessive demand for it. Its index of
gross output value grows in the same way. Contrary to audio-visual
judgement, the share of investment goods industry will’not go up,
but will go down slightly, because insufficient demand for it
restrains its expansion. After the year 2008, however, its share
will begin 10 experience a upward trend due to the support of
demand expansion.
Es
8 3
» &
3
Be
Fig.8 Production Structure
(Proportion to Total Products of Society)
582
* Sectoral Distribution of Employment. The two objectives pursued
by a country’s industrialization are the upgrading of industrial
structure, estimated by the ratio of manufacturing value added to
commodity value added, and preferrably more than 60%,secondly, the
shift of agricultural labor to non-agricultural sectors. The simu-~
lation results show that the share of construction industry. &
infrastructure goes from 12% or so in 1985 up to 25% in 2035, and
that of whole industry (FC. plus FI. plus I.) from 17% up to 33%.
Thus, the shift of agricutural labor will primarily have been
achieved roughly by the end of simulation.
¢ R&D Expenditure (Fig.9) During the period 1978-1985, China’s
R&D expenditure constituted about 1.5-1.8% of national income. ‘The
modelling results show that the R&D expenditure as a percentage of
national income goes from 1.78% in 1985 up to 2.9% in 2008. Later
it reaches its relatively stable value while the demand of the
economic system for R&D achievements rises steadly and S&T is
integalated with economy organically. On the other hand, growth
rate of R&D expenditure , which exceeds that of national income
during the initial few decades, remain at 8-9% owing to the barrier
of limited resources. Then it stabilizes to 4.6-5.7% which is
roughly consistent with that of national income. Generally, the
share of R&D expenditure takes about twenty years to reach its
stable value. However, the stable value is significantly affected
by the application of achievements, or by the effective demand of
industrial sectors for achievements. When the effective demand
increases or decreases by 35% compared with the base run, the
stable share approximates to 3.2% and to 2.7% respectively.
Proportion. of R&D Expenditure
5 to National Income
+12000 6
20.0008 P
4000.0
ee Annual Growth-Rate
28 2 of R&D Expenditure
« 3s soe Maca |
g
2 382
= 828:
if +
a ‘ :
E of. ’
e Behr ence mA . alba
2 55 3 s 2
g Ey g g
& & g FA
Fig.9 R&D Expenditure
Industrial Structure Policy: The choice of Leading Industry
Industrial policies include industrial structure policy, organization
policy and technological policy. The essence of industrial structure
policy as the core of industrial policy is to achieve economic growth
and efficiency pushing industrial structure into the higher level.
Design of industrial structure is primarily to provide the sequence
and opportunity for industrial development. In addition, it is to
583
Considering forward and backward linkage effects (Tab.1),
machinebuilding industry whose linkage effects are sufficient, can
push and pull the development of other industries. For this reason,
some scholars regard this industry as leading industry. However, works
supporting this viewpoint lack quantitative analysis for long-term
projections, which is the key to planning industrial structure. The
base run indicates that during the first few decades,however,
insufficient demand resuls in the serious idleness of its productive
capacity, so thal the share in production structure will fall
slightly. As a result, the demand, as the major barrier, cannot
support its development as leading industry when basic policies are
implemented during the first few decades.
Under the new adjusted policies, such as the increase in each
industry’s depreciation, adding ratio of investment in technological
transformation, and warm incentive to investment demand, new major
behaviours are as follows:
« Machinebuilding industry will enter such a state of operation that
demand for it will be greater than its supply after 1995 (see
Fig.10).
Share in production
Contribution of Structure
-- 245
3 aS Tech.-Progress
2 ggg Descripancy %° Crouth ad
& 28 Between Demand J Tee TI.
Sel Supply = Z of
EIke5,
sses28.
PELE SS ..
, lakes
g ‘
o eee2 be
. R2e82= Hires: . Share in Investment
g FRSESs: &
é a be a oe Structure
“. g¢g22io~" ——
gr SS2283 4 index of Total | ah
BE g Output Value z a
& cal &
Fig.10 Behaviors of Final Investment Goods Industry
* Final investment goods industry will have expanded its share in
production structure progressively, and the index of gross value of
its output becomes second to that of construction & infrastructure.
* Compared with the basic behaviors, national income will go up by
5.1% in 2035,R&D expenditure will grow to constitute 3.05% of
national income around 2030 compared with 2.86% in the base run.
Based on the foregoing results, the preferential consequences and
patterns of industrial development can be achieved. At first,
contributors to bottleneck of industrial structure, such as transpor—
tation, power and raw material, should be given priority in policy~
making and resource allocation. Then, capital inputs and innovations
in agriculiure should be enhanced in order to achieve continuous
steady growth. Even if some industries have certain characteristics of
584
choose leading industry (or strategical industry) which can develop
industrial structure into a more balanced or rational and high level
by the mutual effect when combined with implementing of other techno-
logical and economic policies to subsidize leading industry.
The choice of leading industry is dependant upon a country's reality
and which stage of economic development a country is in, along with
certain criteria. Both ‘the criterion of income elasticity of demand’
and ‘the criterion of growth-rate of productivity’, which describe the
prerequisite of leading industry from the demand and supply side
respectively, are the keystone of Japan’s industrial structure policy.
Through adopting the two critera in the 60’s, the Japanese government
chose ‘heavy and chemical industries’ as leading industry. Japan's
choice was based on such situations as the shortage of natural
resources, so that international trade was placed in the prevailing
position. In China, however, in view of it being such a large country
with significant inward orientation, interindustry linkage can
scarcely be accomplished by international division of labor like
Japan. Consequently, interindustry linkage, in particular, the driving
effects of leading industry on the development of other industries
should be given more weight. That is, the third criterion--‘linkage
effect criterion’-should be take into account. The third criterion is
measured quantitatively by two indicators: dynamic forward linkage
effect and backward linkage effect. The former reflects the impacts or
restraint of one industry on others through the supply of intermediate
goods as well as final capital goods. The latter describes the extent
to which one industry relies upon intermediate inputs of other
industrial sectors.
In terms. of the three criteria and modelling results in the base run,
either fiaal consumption goods industry, intermediate goods industry,
or construction industry & infrastructure is partly characterized by
the leading industry and the criteria. However, technological level
and growth rate of final investment goods industry determine that of
other industries, and furthemore, that of the entire national economy.
Besides, intermediate consumption of raw material and energy is
dependent on the level of technology and new development of R&D,
which, to a great extent,is embodied in capital equipment.
Tab.1 Linkage Effects of China’s Industrial Sectors in 1983
Sector Dynamic Forward Backward
Linkage Effect linkage Effect
Agriculture +43 +17
Metallurgical +73 +33
Power +84 a
Coal +80 +36
Oil +70 25
Chemical +63 +39
Mach inebuilding +58 +43
Building Materials +81 45
Forestry +58 +34
Food ; +19 +62
Textile «19 +27
Paper 42 +40
Construction +93 +74
Transportation +79 +27
Trade Al 46
585
leading industry during the next few decades, leading industry will
not have emerged. Hence, by the end of this centry, industrial
structure should be pushed towards an equilibrium or rational level in
order to estabilish the best starting point for the next centry’s
growth, Later on, with the rise of depreciation rate and ratio of
investment in technological transformation, equipment-manufacturing,
which ,will grow rapidly, will drive the growth and industrial
technology level forward. Thus, during the later stage, the pattern is
one of inclined development of final investment goods industry as
leading industry on the basis of the coordinative development.
Analysis on the Evolution of Industrial Structure[*]
Considering the indicator--'heavy industry ratio’--formulated in terms
of the modelling results under the adjusted policies, we conclude
that, during the initial modelling periods, China will go through a
course of continuous decrease of ‘heavy industry-ratio’. This is
different from developed countries whose portion of heavy industry
rose continuously during its industrialization process. Why is this
so? It is due to China’s previous unreasonable pattern of over-
expansion of heavy industry which will still have a great impact until
the next few decades. Later, however, ‘heavy industry-ratio’ will begin
to go up and reach 55.9% around the year 2035. In this case, China
will still be in the development stage centred upon heavy industry,
compared with ‘heavy industry-ratio’ of various industrial countries
in °1970--57.4%, 61.0% and 62.4% for U.S, Britain and F.R.G
respectively.
Tab.2 Heavy Industry-Ratio and Hoffmann Coefficient
Year 1985 1995 2005 2015 2025 2035
Heavy Industry Ratio «525 +525 +517 524 541 559
Hoffmann Coefficient 2.207 2.26 2.255 2.115 1.995 1.899
Studies by Japan’s economists indicate, that, among those countries
whose per capita income exceeds 200-300 U.S. dollors (calulated in
1950 prices), Hoffmann coefficients of countris with high industriali-
zation level, or post-industry countries, more or less remain
constant. Otherwise, Hoffmann coefficients of industrilizing countries
will go down, and then reach its relatively steady level in a certain
‘period. Contrary to this general rule, however, China's Hoffmann
coefficient will have demonstrated a slightly upward tendency by the
end of this centry. It will then fall steadly, at last, follow a
gradual decline. Therefore, China will still have been in the middle
and late stage of her industrialization during the next fifty years,
From these analysis, we can conclude, that, the issues: which develop-
ment stage China will be in and how to choose and plan the preferent
consequences and opportunity for industry development, are provided
with the basic knowledge and outlook.
[*]: Here and later on we concentrate the discussion upon structural
change among I., FI, and FC.
Acknowledge
Many thanks go to Mr. Steve Peterson for his helps in correcting this
paper.
MAIN REFERENCES
Word Bank Report, China: Problem and Approaches for Long Term
Development, 1984
Xu Chingrui, R&D Management, Higher Education Press China, 1986
Hollis B. Chenery & Moises Syrquin, Patterns of Development 1950-
1970,
G.P.Richardson & Pugh III, Introduction to System Dynamics
Modelling with DYNAMO, MIT Press, 1981
Wassily Leontief, Input-output Economics, New York Oxford Universi-
ty, 1966
Li Hanping, “Combination of S&T with Economy: Study Based on
Allocation Mechanism of R&D Resources", Researches in Science
Management, China, Vol.6, No.4,1988,4
Yang Zhi, Introduction to the Industrial Economics, People’s
University China, 1985
Statistics Bureau of China, Almanac of Chinese Statistics, 1981~
1986, Chinese Statistics Press
