l ofS*
INDUSTRY AND ENERGY DEPARTMENT WORKING PAPER
INDUSTRY SERIES PAPER No. 4
Technological Advance and
Organizational Innovation in the
Engineering Industry '
March 1989
The World Bank Industry and Energy Department, PPR



TECHNOLOGICAL ADVANCE AND
ORGANIZATIONAL INNOVATION
IN THE ENGINEERING INDUSTRY
A New Perspective on the Problems
and Possibilities for
Developing Countries
May 5, 1989
Kurt Hoffman
Sussex Research Associates
Brighton, England
for
Industry Development Division
Industry and Energy Department
Policy, Planning and Research
World Bank



This paper is part of the work on technology change that the Industry
Development Division of PPR/the World Bank is carrying out.
Acknowledgement is due in particular to Roy Pepper, of the Industry and
Energy Division of the Technical Department for Europe, the Middle East and North
Africa; and to Edmund Mangan of the Industry, Trade and Finance Division of the
Technical Department for Asia, for their early encouragement of this work.
The author also wishes to acknowledge the support of Nancy Barry, Chief
of the Industry Development Division of the Industry and Energy Department,
PPR/the World Bank, and to Carl Dahlman, Principal Economist in that division.
Particular appreciation is due to them for their assistance and advice. The
author also wishes to thank the participants at a World Bank seminar given by
the author in October 1988, at which time the basic premises of this paper were
presented for discussion.
The Rockefeller Foundation also provided support to the author to extend
the scope of his research and thinking on the topic.



SABLE OF CONTENTS
Pare No.
INTRODUCTION  ...............................................     i
I.   SYSTEMIC AND FLEXIBLE AUTOMATION IN THE ENGINEERING SECTOR
A.   The Problems of Batch production in the Engineering
Sector . ..............................................     1
B.   Information Technology in Industry.                          3
C.   From Stand-alone to Integrated Flexible Automation           9
D.   FMS:  Impact, Diffusion, Investment .16
E.   Lessons from the Diffusion Process:
the Importance of Organizational Compatibility             21
II.  ORGANIZATIONAL INNOVATION IN THE ENGINEERING SECTOR
A.   The Erosion of Western Competitiveness .31
B.   The Impact of Organizational Change at the Firm Level.      40
C.   Management as the Driving Force--the Workforce as the
Source of Improvement .44
D. The Quality Factor .46
E.  The Elimination of Waste .50
III.  ORGANIZATIONAL INNOVATION IN PRACTICE
A.   Managing the Workforce ................................     57
B.   Managing Machines .....................................     62
C.   Managing the Production Process .......................     66
D.   Managing Supplier Relations ...........................     74
IV.  CONCLUSION
A.  Summary of the Issues .77
B.   Assessing the Implications for Developing Countries         78
C.   Implications for the International and National
Competitive Context .79
D.   Problems and Possibilities for Organizational
Innovation in Developing Countries .83
E.  The Problems Considered .83
F.   The Possibilities Considered ..........................     87
BIBLIOGRAPHY    ....................................................   99



INTRODUGiTION
i.          The engineering sector, and industry generally, is engaged in a
process of fundamental structural, organizational, and technological trans-
formation. The key features of this transformation are growing flexibility,
greater diversity, declining product and, conceivably, plant scale economies
and, ultimately, quantum leaps in productivity. These developments presage a
much tougher competitive environment domestically and internationally in the
engineering sector.
ii.         The leading edge of this transformation currently is centered in
the advanced industrial economies, but the changes will have implications for
all countries. The dominant technical factor at work in this process is the
development and diffusion of a family of automation technologies, flexible
manufacturing techniques, and new ?roducts deriving from computer-based
control and communication systems.J These innovations are generic, highly
adaptable, and thus applicable to virtually every segment of the engineering
subsector. The considerable technical and economic advantages arising from
their successful adoptlon give significant competitive gains to user firms.
iii.        Although most change in production technology so far has involved
the introduction of individual automation technologies (e.g., CAD systems, CNC
machine tools) on a stand-alone or island basis, the trend toward integration
of different components into more flexible manufacturing systems (FMS) is
well-established.  The signiflcance of this trend is that the gains attainable
through systemic integration are considerably greater than those from stand-
alone automation technologies.
iv.         Although the technology factor is clearly a major stimulus for
change in the engineering subsector, other developments are beginning to play
an equally important part in the transformation. Perhaps the most important
are the organizational innovations associated with concepts such as just-in-
time (JIT) inventory control, total quality control, and total preventive
maintenance, which are spreading rapidly through the engineering sector at
inter-firm and intra-firm levels. Much of Japan's early international
economic success in engineering and other sectors is attributed to the
competitive advantages Japanese firms gained from their persistent efforts to
apply these principles to the organization and management of production from
the 1950s through the 19709. In fact, many of the so-called "new management
practices" originated earlier in the U.S., where they were never adopted
widely. Regardless of their historical origins, these organizational innova-
tions represent a sharp departure from the conventional management practices
historically followed by Western firms--many of whom now appreciate that their
future competitive success will depend on successful adoption of "new"
systems. Indeed, a firm's capacity for introducing organizational change has
1/    Simultaneously, other technological breakthroughs are finding widespread
use throughout the sector in both process (e.g., lasers) and product (e.g.,
new materials) applications. Eventually the technical and economic effects
of these innovations will be substantial. Commercial introduction of these
technologies is at too early a stage for predictions regarding their
eventual impact; they will not be discussed in this paper.



eii -
emerged unexpectedly as perhaps the most critical determinant of its success
in using advanced flexible manufacturing techniques effectively.
v.          There is a growing consensus that the combination of flexible
manufacturing technologies and new forms of production management constitutes
a new hbest-practice* system of manufacturing that will replace conventional
approaches in much the same way that mass production replaced craft-based
methods in the nineteenth century. The diffusion of new organizational and
technological innovations is affecting the dynamics and structure of the
engineering industry within the industrialized countries and in the process is
altering barriers to entry and competitiveness at national and international
levels. Because of the engineering sector's vital role in industrial develop-
ment, the profound technological changes taking place are particularly
relevant for countries on the threshold of industrialization.
vi.         In many respects, the changes underway in engineering are part of
a much broader process of change in the global industrial system. This is
manifested at the micro-level in the rapid cross-sectoral diffusion of the new
techniques, as well as in the related rise of Japan as a dominant industrial
power--intent on changing the 'rules of the gamew of global competition--and
the response of the U.S. and European firms to this challenge. The engineer-
ing sector represents a laboratory where the elements of the new technological
and organizational paradigm are being tested and perfected.!/
vii.        This paper is part of work on technology change by the Industry
Development Division of PPR-the World Bank. The first chapter loocks at the
current pattern of technological change in the engineering sector, concentrat-
ing primarily on the nature and impact of what are called flexible manufactur-
ing systepA. Though not at the leading edge of computer-integrated manufac-
turing, FMS represent a significant advance in integration and flexibility
over previous generations of automation technology and are becoming central to
innovation and competition in the sector.
viii.       Although the pattern and effects of information technology on
manufacturing applications have been a dominant focus of industrial analysis,
appreciation of the significance of organizational innovations is more recent.
Chapters II and III explore how organizational change is becoming a key
determinant of a firm's competitiveness in the engineering sector and in
if    For this paper we consider the engineering sector in its widest sense as
essentially including all industries based primarily on metal working, thus
a range of sectors and firms that exhibit enormous diversity. The products
involved range from simple gears or valves to extremely complex multi-
component products such as automobiles and aerospace engines; the scale
of production varies from millions of identical products to limited runs
of special purpose items; the complexity of the production process spans
the gamut from unsophisticated one-step stamping operations to multi-stage
processes requiring extremely fine tolerance and control procedures; and
finally, we include capital goods firms, as well as producers of
intermediates and final products, with the size of these firms varying from
fewer than 50 employees to more than 100,000.



- iii -
man..acturing generally. Chapter II discusses the growing consensus about the
significance of the new practices, reviews evidence of their impact across
sectors and countries, and provides an empirical foundation for the analysis
that follows.  It also esplores the new organizational and managerial ap-
proaches, how they work, and the problems encountered in their introduction.
Chapter III reviews the operational techniques firms use to bring the general
principles into practice and gives an impression of how these techniques
operate in practice, noting some indications of their productive applicability
in developing countries. Chapter IV also focuses on issues related to the
applicability of the new organizational practices in developing countries.
The objective of the chapter is to encourage further examination of the issues
on the grounds that the possibility of introducing these orwAr4-ational
methods could be one of the most important cnailenges and opportunities for
industrial progress now confronting developing countries.



CHAPTERI
I. SYSTEMIC AND FLEXIBLE AUTOMATION Is THE ENGINEERING SECTOR
The Problems of Batch Productior in the Engineering Sector
1.01        In advanced induatrial economies, firms operating in most segments
of the engineerinzv sector face the dilemma of sacrificing the economic -
advantages of hign volume ard long production runs to maintain flexibility for
producing relatively small batches of output. With the emergence of a
powerful range of new technologies based on programmable automation, the
engineering industry is moving into an era when the trade-off between flexibi-
lity and scale will not be necessary. Before discussing the benefits from
this advance, this chapter characterizes the production process in engineering
and identifies areas where flexible automation technologies are beginning to
have a major effect.
1.02        Engineering production falls into three categories: mass produc-
tion of standardized products, with tens of thousands of units manufactured
per year; batch production involving annual volume from tens to thousands of
units; and production of small lots, one-off items, and prototypes. In the
past, producers in each segment evolved a manufacturing technology suitable to
their conditions of production and competition. Mass production firms relied
heavily on automated transfer lines and costly dedicated equipment operated at
high capacity and capable of garnering impressive scale economies. Producers
of one-offs and prototypes, meanwhile, were able to use a highly skilled
workforce in conjunction with flexible machines to create great product
variety at very low levels of output.
1.03        These two groups of firms were among the first in the engineering
industry to benefit from the application of microelectronics to manufacturing
technology--high volume firms, from the further automation of their dedicated
production lines; and small batch firms involved in prototype production, from
computer-numerical control (CNC) of individual machine tools.
1.04        Firms engaged in the middle area of batch production !  faced a
far more difficult task in coping with the demands of their markets. Volumes
were too low to use dedicated machinery and too large for single machines.
Confronted with the need to produce a large range of products at highly
variable levels of output, batch producers constantly traded volume production
for flexibility. This accounts for the basic production setup still in place
among most engineering firms: a variety of function-specific, stand-alone,
i/    These firms account for a large majority of output in the engineering
industry where, on average, 70% of components are produced in batches of
less than fifty. Much of this batch work is done by small firms. Brandt
(1986) estimates that the 100,000 small job shops in the U.S. engineering
industry supply 75% of all the machined parts used by larger engineering
firms. See also the further discussion of the U.S. case in Computerized
Manufacturing Automation, Office of Technology Assessment, Washington, D.C.
(1984); and NEDO (1984) for the U.K.



manually operated machine tools organized in a multi-step, discrete manufac-
turing process.
1.05        The necessity of producing in batch lots--with varying quantities
and specifications--created complications in forecasting sales, ordering raw
materials, balancing, tracking, and in choosing among many possible, but non-
optimal production routes. The level of automation attainable under these
conditions was low, and batch productionl firms never wers able to captura
substantial scale economies. In addition, a range of other problems plagued
these firms and imposed costs in efficiency, capacity, and competitiveness:
*     Frequent breakdowns of machinery, poor maintenance proc 'res, and
long set-up and changeover times between product changes meant low
capacity utilization;
*     Bottlenecks and queueing problems;
*     Large inventories of raw materials, work-in-progress, ard finished
goods;
-     Long lead times in design, production, and product launch;
3     Poor production control leading to excessive scrap and waste;
e     Inconsistent quality and high rejection rates;
3     High production and management overhead; and
v     Poor delivery.
1.06        Because of these problems, in the typical U.S. engineering firm in
the 1970s, machinery stood idle 70-95% of available production time, with
value-added operations on the product accounting for only 2% of tHe time in
the factory. More significant, between 50-70% of total product costs can be
tied up in materials inventory and other overheads. For example, in the U.K.
engineering industry, this accounted for the estimated US$37 billion in
inventory and work-in-progress in 1986.4-
1.07        Through the end of the 1970s and into the 1980s, the competitive
pressure on engineering firms began to change due to market entry by low-cost
competitors, changing consumer preference, the recession, etc. These
pressures translated into demands for lower prices, shorter production cycles,
quicker delivery/response, and greater emphasis on design, quality, and
customization (see Table 1.1). These problems require solutions that allow
the attainment of both high productivity and high flexibility--a set of
i/    Ayers and Miller (1984); and Bessant and Heywood (1986).  The discussion
on production problems facing engineering firms draws on the work of
Bessant (1985), Bessant and Heywood (1986), and Bessant and Rush (1987).



characteristics that has come to be associated with the phrase "manufacturing
agility",5S
Tabi I 1: NAUFACTURING FUTURES SURVEY (1986)
(List of coaptitive priorities In order of laportance)
Europe            USA                Jspan
Consistent qualfty    Consistent quality    Low prices
High perfor nce   High performance  Rapid design changes
Dependable deliveries  Dependable deliveries  Consistent quality
Fast detiveries   Low pricei         Dependable deliveries
Low prices        Fast deliveries    Rapid volume changes
After-sales service   After-sales service   Fast detiverie.
Rapid voisae changes   Rapid volume changes   After-sales service
Source: 1986 INSEAD survey of senior management of 500 to manu-
facturing firms; cited In eessant nd Rush (1987), p.4
1.08        The ability of the engineering industry to respond to pressures
for production flexibility had been inhibited until recently by the limita-
tions of existing technology, though partial solutions were possible (i.e. the
use of CNC tools or Croup Technology). The recent availability of technical
solutions (based on inte-rated production systems) to the flexibility/cost
conundrum coincides with pervasive market pressures for firms to adopt these
solutions. This timely interaction of market demands and technology will
remain an important force in the diffusion process in the engineering sector
of the industrialized countries.
1.09        In turn, as discussed below, both the changing market conditions
and the responses of engineering firms in industrialized countries to these
changes will affect developing countries in major ways--in both export aspira-
tions and in the organization and production for domestic markets.
Information Technologv in Industry: Toward a Systemic Understanding
1.10        Microelectronics enjoys distinct economic and technical advantages
over previous methods of information processing and control technology
(manual, mechanical, electromechanic r, and pneumatic).  These advantages
constitute an extremely powerful set. of economic incentives that virtually
compel the substitution of microelectronics-based systems for earlier
5/    INSEAD (1987), cited in Bessant and Rush (1987).  These trends are well
developed in the motor vehicles sector and in a wide range of consumer
products and high volume continuous processing industries such as
petrochemicals, food processing, textiles and clothing. See Wyatt (1987)
and "Microelectronics Monitor," irsue 21, Fall 1987, UNIDO, Vienna, for
a review.



systems.!1 Because misroelectronics can be used in all information-based
activities, the technology can be introduced into virtually every aspect of a
firm's operations--from production management, administration, design and
process specification, and raw material processing, to packaging, testing and
inspection of final products and manufacturing processes. Due to this
flexibility, microelectronics, or information technology as it also is called,
has found wide application across industrias, witlkin sectors, and within
firms.7/ This capacity of information technology to stimulate successive
rounds of innovation outwardly lies at the heart of the so-called information
technology revolution.!,
1.11        Of more direct intersst is the following model of the pr-gress of
automation at firm level,!' which helps situate past and future (.zvelopments
in automation in a way that relates directly to current patterns of technical
change in the engineering industry.
1.12        Figure 1.1 shows manufacturing divided into three distinct spheres
of activity:  desi, groduction, and cogrdination, with each sphere compris-
ing a set of discrete activities.1_0 Since the Industrial Revolution, there
has been a gradual process of automation of individual activities, but rarely
were discrete activities within spheres linked together, even in the mass
production assembly lines with their fixed automation. The recent emergence
of microelectronics, however, has stimulated a more pervasive process of
activity-specific automation. This is occurring with different degrees of
intensity and rapidity at three levels, as depicted in Figure 1.2.
ik    See Soete and Dosi (1983) for the clearest explanation of the technical
advantages of microelectronics and their economic implications.
2/    See Hoffman (1986) for a review and references.
L/    The macro level effects of this technological revolution for the national
and international economy and the implications for theory and policy have
been widely debated and discussed in the literature. See Dickson and Marsh
(1978); Hines and Searle (1979); ETUI (1983); Freeman, Clark and Soete
(1982); Soete (1985);  r. Kaplinsky (1987) for early reviews and different
perspectives. See James (1987) for extensive references and a critical
discussion of the limpacts" literature.
_/   This model has been developed by Kaplinsky  in a recent series  of
publications. See Kaplinsky (1984, 1985) for the initial presentation,
and Hoffman and Kaplinsky (1988) for a revised version and empirical test.
.1Q/ Design activities Include drafting, copying, basic and final design,
process engineering.   Production activities could be mixing, molding,
cutting,  handling,  testing,  packaging.    Coordination  involves  all
managerial tasks needed to support and guide the firm's operations
internally and in the marketplace.



Pplre 1.1  Pra-,learonle Orgaslration of Factory Production
paper-bied input                              paper-base# Input
(e.g. tender document)                        (e.g.       list)
N.>  paper-based output
I      I ~~(e.g. parts, Iltsesnfrato
Design                             I   coordina Lon
paper-b     output      p     ba       t                ad
(e.g. dra    g)         (e.g. p   uctiotpu
(e.g. markating
flow of          Mauftu\          flow of
inuputs           8      t    __________
Source: gaplnasky (1985)
Figure 12 Three Levels of Autoctin
(a) Ltra-activity
elgn                     Coordination
autosti
t         Manufactunr    uomto
tos       t tion
Design        (b) lntra-sphese
~ Q     Coordination
automat Ha
(c) inter-cohere
Design
o ordination.
automation
Manufacture
Source: Kaplinsky (1985)



- 6 -
1.13        The first level of intra-activity automation involves automation
of individual activities in a stand-alone fashion; application of information
technology has been concentrated at this level so far. The second level is
intra-sphere automation; its key feature is the integration of individual
activities within the same sphere. The third level is inter-sphere automation
in which activities in separate spheres are integrated and coordinated via
their common dependence on digital control systems.
1.14        The crucial components in this multi-stage advance of automation
are a family of automation technologies that emerged early in the industrial
application of microelectronics. Computer-aided desin (CAD) is dominant in
the design sphere. In manufacturing, computer-numerical control (CNC),
appl'.ed initially to machine tools, was the earliest, followed by r2bots.
orogrammable logic controllers. automated materials handling systems, and
Drocess controllers for real-time control of production. In the coordination
sphere, centralized data orocessing and office technologies were first to have
an impact on managemen;..
1.15        As with microprocessors, these new technologies are highly
flexible, with applications suitable for use in a number of sectors.l-/ User
firms have been able to capture substantial technical and economic gains, and
the pool of users has steadily increased' as unit costs have declined and
performance improved. These factors contributed to the rapid diffusion of
automation technologies beginning in the late 19705.12/
1.16        While the large-scale, mass production segment of the engineering
industry uses CAD units and robots extensively;L31 the sector has played a
.11/ Various issues of UNIDO's "Microelectronics Monitor" provide details of
a wide range of applications for stand-alone automation technologies. See
also Hoffman (1985).
12/ Sales of CAD systems grew 85% annually over 1978-82 and by 1985, some
27,000 firms used CAD systems worldwide.  The CAD world market totaled
about US$3.3 billion in 1986, with an expected annual growth rate of
between 15% to 20% in the medium term. For industrial process controllers,
U.S. demand has been growing at 30% annually; in 1985 some 115,000
programmasle logic controllers were in use in FRG, France and England;
77,500 programmable robots were in use worldwide in 1984, with more than
140,000 'steel collar" workers in place in factories in Japan, Europe and
the U.S. by 1986. Total 1986 sales of robots were about US$1.88 billion
and are expected to nearly double to US$3.5 billion by 1990.   U.S.
Department of Commerce (1985) for CAD numbers; ECE (1986) for machining
center figures; Northcott and Rogers (1985) for European figures; and UNIDO
(1987) for robot figures. See also The Economist (1987) and Bessant and
Rush (1987).
13/   Producers of design-intensive products in the aerospace sector (where CAD
systems were first developed), such as Boeing, typically use many hundreds
of CAD systems, while GM has some 200,000 programmable  tools.   The
Economist (1987).



- 7 -
very important role in the development and diffusion of numerical control
technology. Expanded demand for numerical control machine tools (NCHTs) has
led to a large increase in the share of NCMTs in total production of machine
tools throughout the OECD (Table 1.2). Table 1.3 shows that the engineering
industry was the major user of NCHTs in the U.S. and Japan in the early 1980s.
The growing intensity of NCMT use relative to conventional tools in the U.S.
metalworking industry is indicated in the statistic that by 1983, approxi-
mately 103,000 NCHTs were in use.14/
tabte 1.2: SHARE OF NCNTS IN TOTAL PRODUCTION
OF SELECTED NETALCUTTING MACHINE
TOOLS IN OECD IV 11976 and 1982)
1976               1982
.............  .........  Growth Rte,
us.       X        usso     X           1976-1982
Boring machine
NCNT              92       35        297      57            223
Conventional       171       65        226     43             32
Milling mwchines
MCNT              145      23        633     53             337
Corwentional       493       77        S57     47             13
Drilling mchines
MCNT              34       13         93     34             173
ConventionaL       229      87         178     66            -22
q/ USA, Japan, FRO, France, Italy, UK
Source: Hoffman (1986)
4J/   Large firms have always been major users of NCMTS, but smaller engineering
enterprises have now become consistent investors in the new technology.
Small firms with fewer than 100 employees accounted for about 40% of the
U.S. stock of NCMTs in the early 1980s. In Japan, firms with fewer than
300 employees accounted for 62% of the investment of NCHTs in 1981.
Jacobson (1985).
In NCMTs, unit price declines have been the result of declines in the cost
of the CNC unit, which in 1985 was less than four times its cost in 1978.
Likewise with CAD units: A typical workstation can cost from US$50,000
to US$125,000, but prices are now dropping to below US$20,000 and even
US$10,000.  Similar performance could be cited with regard to robots.
NMTBA (1986); the Economist (1987); Electronics Week (July 1984).



- 8 -
tab(e 1.3: OISTRIBUTION OF NCITS BY SECTORS IN JAPAN (1981)
AND THE USA (1983)
Japan t/              USA /      X
General machine         11,394     43        52,541      51
Etectricat machinery     4,262      16       10.772       10
Tramsport equpmnt        6,276     23        15,284      15
Precision machinery      1,775      7         4,874       5
Metal products           1,460      5        14,463 h/    14
Casting/forging products   580      2         2,662 g/    3
Miscellaneous             978       4         2,102       2
Totat                   26,175     100      103,308      100
j/ The Japanese inventory covers ptants with 100 or more eaployees; the USA
inventory covers plants of all sizes.
F fabricated metal products
S Primry metals
Source: Hoffmn (1986)
1.17        More recent data indicates that engineering firms of every size
continue to invest in all aspects of automation technology. For instance, the
most current annual survey of computer use in the U.K. engineering industry--
based on a structured survey of over 2,000 firms--shows about 20% annual
growth in overall applications of computer hardware and software between 1985
and 1987.15/ In the U.S. 25% of companies planning to buy robots in 1986 had
annual sales below US$10 million; with most of the firms involved in the
production of machined metal parts. Computer purchase by small job-shop
manufacturers are expected to increase by 35% a year through 1990.162
1.18        In summary, the diffusion pattern for all three of the main
automation techniques--NCMTs, CAD systems, and industrial robots--has followed
the typical S-curve. NCMTs entered their growth phase in the mid-1970s and
are now in their mature phase as evident in metal cutting activities where
their penetration is greatest. NCMTs now account for over 76% of all produc-
tion by value in the OECD countries, and small firms have accounted for well
over 50% of investment in recent years.
1.19        CAD systems entered their growth phase in the 1982-84 period and
will remain in this phase for some time.  Their take-off was spurred particu-
larly by the development of PC-based systems, which have registered annual
i2/ Benchmark Research, Ltd., "Surveys of U.K. Computers in Engineering, 1985,
1986, 1987" as reported in Engineering Computers (1987), Findlay
Publications, Horton Kirby, Kent. Cited in Bessant and Rush (1987).
j2/   Brandt (1986).



sales increases above 70% since 1984.17/ Finally, sales of industrial robots
began to grow more rapidly in 1984/85. Annual investments in the OECD
countries are now about US$1.5-$2.0 billion. Robots' growth phase is continu-
ing, suggested by the growing investments by small firms and, more important,
because robot applications have begun to move out of strictly welding opera-
tions and into assembly activities.18/
From Stand-alone (Islands) to Integrated Flexible Automation
1.20        Although automation technologies are undoubtedly major innovations
in their own right, two additional aspects of their diffusion should be noted.
First, automation technologies have been introduced largely on a stand-alone
basis, advancing the level of intra-activity automation within the firm but
not involving any linkage either within the same sphere or across spheres.
Much of the multi-billion dollar investment in automation made by the U.S.
auto transnational corporations (TNCs) in the early 1980s, for example, was in
stand-alone equipment. The bias towards installing islands of automation was
typical of much of the engineering sector ;And manufacturing industry through
the 1980s and is a pattern that will continue as the range of applications
expands, functions increase, and relative unit costs decline.19'
1.21        Second, and more important, the focus of innovation and investment
has shifted increasingly toward the intra-sphere and inter-sphere integration
of these islands of automation.  The emphasis now is on capturing systemic
gains through exploiting the flexibility inherent in the technology.  Flexible
automation's implications for firm-level gains and international competitive-
ness are qualitatively different from the issues raised by stand-along automa-
tion.2_/ This evolution toward more flexible, integrated systems is feasible
because automation technologies share a common knowledge in the use, process-
Jr/   One U.S. producer of PC-based CAD systems sold more than 100,000 low-cost
terminals in 1986.
13/   See Jacobson, Steffan and Charles Edquist (1988), Flexible Automation: The
Global Diffusion of New Technology in the Engineering Industry, Oxford:
Basil Blackwell Ltd, for the most comprehensive diffusion review available
on stand-alone automation techniques.
12/   This wave of stand-alone automation has attracted a great deal of attention
in the analytical literature -- not only in relation to industrialized
countries but also, to a much more limited extent, in implications for
developing countries.   See the articles  in Hoffman  (1985) and the
references in James (1987).
2Q/   These recent developments are the subject of most of the current research
by policy analysts in the industrialized countries. See, for instance,
Haywood and Bessant (1987) and Bessant and Haywood (1987), as well as New
(1986), Voss (1986), Boody and Buchanan (1986), and Senkar and Beesley
(1987). So far the literature has hardly begun to interpret this new trend
from the perspective of developing countries -- an issue to which we return
in the Conclusion.



- 10 -
ing and transmitting of information.  The wav is ooen now to link together
automation technoloeies within the separate soheres of oroduction. design and
coordination and to integrate manufacturing with design. under the coordina-
tion of management. This process is being driven by the underlying tech-
nological logic of microelectronics that, in principle, the greatest gains
from the industrial use of information technology occur when the highest
degree of integration is achieved.
1.22        =I .  The generally accepted name for this trend toward full f
manufacturing systems is computer-integrated manufacture. The CIM concept,
strictly speaking, refers to the ultimate manufacturing installation, where
all the relevant activities in a company's operations are brought together
under integrated computer control, a development that used to be referred to
as the "factory of the future". However, the concept also is used more
generally for the many emerging configurations and levels of integrated
automation within the factory. Considerable strides are being taken toward
ever higher levels of CIM, such as the integration of CAD with computer-
directed machining operations (known as CAD/CAM); CAD links into computer-
based inventory control and purchasing systems; and the linking of all three
(inter-sphere automation) via CIM. These leading edge CIM applications have,
naturally enough, attracted a great deal of media attention._/
1.23        Moreover, integration has gone beyond the factory and company
bounds. Electronic communication has led to the development of a range of
computer-based links in design, production scheduling, purchasing, and
shipping among suppliers, producers and customers. This level of integration
between producers and suppliers is represented in Figure 1.3, with indications
of the different forms that intra-firm automation can take.fi
21/ The firm-level factory automation landscape exhibits similar characteris-
tics across Japan, Europe and the United States. In each region there is
a select group of leading-edge firms, perhaps 30-40 in total, followed by
a much larger groups of firms now aggressively following, albeit at a less
advanced stage--and benefitting from the lessons of the leaders. Arbose
(1985) gives details of the SKF installation in Sweden and the massive GE
CIM complexes in Pennsylvania and Kentucky; Handke (1982) and Jelinek and
Godhar (1986) discuss the well known CIM facility at Messerschmitt-Bolkow-
Blohm in West Germany; Hoffman and Kaplinsky 1988) give details of the
extensive CIM setup at FIAT in Italy; and ECE (1986) describes a number
of other CIM installations including those at Fujitsu Fanuc and Yamasaki
Machinery Works in Japan.
22/ Hoffman and Kaplinsky (1988) describe one of the most advanced examples
of computer-based intra and inter-firm automation at Nissan's Murayama
plant where computers control all aspects of production, including the
operation of flexible automation technologies, production scheduling,
components ordering and in-plant JIT. Computer-based ordering accounts
for 90% by quantity and 90% by value of all components used.



Figure 1.3
Organization in the "Factory of the Future"?
E    InitiaOPEN GaSThS INTERCaNgCSpiONf
CA    Computer-A*.i deA.P. D Ges
E .    tC Stiatitial Prcs  Control
FM.AS.PFe.bl Manufacturing            Sysmtinrtemsl
CA.G..      AuomauticrGuided  Vehicle
T.Q.C.      Total Quality Control
M.R.P.      Material Requirements Planning
J.I.T.      Just-in-Time
,C.I.M.      Computer-Integrated Manufacturing
Source: Hoffman (1988)



- 12 -
1.24        The trend toward full integration will almost certainly define
manufacturing in the long term, but the most immediate and relevant technical
changes involve the design and introduction of flexible systems within the
manufacturing sphere--particularly intra-sphere automation. In this area the
distinctions usually drawn among the three levels of integration are:
*     flexible machining units (FMU) which are one-machine units,
typically a machining center (combining functions previously
carried out by individual CNC tools) equipped with automatic tool-
changing and material-loading devices that allow some degree of
unattended operation;
*     flexible manufacturing cells (FMC) comprising two or more machines
(machining centers and/or individuals CNC tools) plus material
handler, all controlled by computer; and
*     flexible manufacturinif svstems (FMS) made up of two or more FMC,
some form of automatic transportation system to move pallets,
workpieces, and tools between machines, and all controlled by
computer.
Figure 1.4 captures these different levels of intra-sphere integration in
metalworking._/
ZJ/  Within metalworking,  developments are proceeding in three directions.
First, where most FMS are now used on "prismatic" products, new generations
can handle rotational products as well.  (Prismatic parts are based on
cuboid shapes such as gearboxes, and rotational parts are cylindrical in
shape, such as axles and shafts. Currently these are handled separately,
with the majority of FMS used for prismatic machining. Future FMS will
be able to handle both shapes, thus eliminating a completely separate stage
in processing). Similarly, FMS have been developed that can handle sheet
metal work whereas previously they had been confined to activities
involving metal cutting of one form or another.



- 13 -
Figure 1.4 The Trend towards Integration in Metalworking
\  <    <    </<    n    iing     le purpose
\ \<_,/ Idedicated machine
o     o     o    (3 C       tools and skilled
/ overators
Multipurpose
machines and tools
U  O)O                  and skilled
operators
Q\\               INC & CNC multipurpose
000 Q   and programmable
machines and actions of
o \       Z skilled or unskilled
operator
\\\  2       | CNC machining centers
under CNC comouter control
as manufacturin2 cell
Q          IFully flexible
*manufacturing system
Source: Bessant and Rush (1987)



- 14 -
1.25        Technological developments in this area are multi-dimensional, and
although diffusion is proceeding at a reasonable pace, flexible manufacturing
is still an immature technology. Users and suppliers continue sizeable R&D
projects, while considerable research is being supported by the public sector.
Despite this degree of experimentation, there is little doubt that the
orientation of industrial innovation within engineering and manufacturing has
shifted irrevocably in this direction. Three aspects are discernible for
future progress:
(i)   Increasing flexibility.  Changes under way in manufacturing
technology represent a truly major advance in flexibility and are responding
to several factors: increased international competition, combined with a
change in the nature of market demand for greater quality, more variety,
custom product specification, improved delivery times, and a shorter product
life cycle; factory automation suppliers offering flexibility and higher
integration at lower cost--a trend that will continue as output levels
increase and standardization prevails;24/ and most important, the sizeable
competitive advantages offered by the enhanced flexibility of integrated
systems, advantages that enable firms to meet changing market demands without
sacrificing scale economies.!!/
1.26        This new concept of flexibility has a number of dimensions,
providing technological advantages not available previously:
.     Product flexibilitv--for easier changeover from product to
product;
*     volume flexibility--for efficiently accommodating changes in
volume;
*     routing flexibility--to process parts via different routes within
the plant in response to breakdowns or other factors;
*     machine flexibility--to make different parts within a product
family;
*     operation flexibility--to vary the sequence of operations; and
*     2rocess flexibility--to produce a product family in different ways
using different materials.
2_/  Haywood and Bessant (1986) show that the average cost of FMS installed in
the U.K. has declined from US$8 million in the early 1980s to US$1.5
million in 1986; costs for FMU and FMC have also declined considerably.
2/   This shift in manager's perspective towards flexibil'ty as a source of
economic gains in production is precisely the kind of charge in the
collective "common sense" that one would expect to see as evidence of the
diffusion of the new technological and organizational paradigm. See Perez
(1985) and Piore and Sabel (1984).



- 15 -
Increased flexibility thus yields benefits in reducing lead time, increasing
capacity utilization and throughput rates, expandlng product variety, and
lowerLig labor requirements.
1.27        (it)  Proliferat     of oJtionn.  DependLng on how Che different
elements of the technology are combined, numerous technlques encompassed by a
flexible manufacturing system (FMS) are now avallable for each area of
manufacturing (see Flgure 1.5). These include different types of FHU and FMC
and also techniques for storage, handling, machine feeding, tooling, and the
control system. Concerning FMS, most systems still are put together on a one-
off basis. However, the technology is sufficiently advanced to allow con-
figurations that cover a much wider variety of part/volume combinatlons, which
means that firms of all sizes are able to use FNS; most prevLous users were
large TNCs.
Figure 1.5 Options ln Flexible Hsnufactuzlng Systems
Number
Of
parts
Number of different parts
Soure:l Rasemna and Rusi (19Sr)
1.28        (iii)  Extension of th. scope ogf asgications.  The third charac-
terlstic of current and likely future trends is the penetration of flexible
automation into a widening range of sectors, products, and manufacturing
activities. Users of flexible manufacturing technologies in the late 1970s
and early 1980s tended to be in select segments of the engineering industry,
such as aerospace, vehicles, and agricultural machinery, where FMS were used
primarily on a narrow range of operations on particular types of components.
In metalworking. asseutal F45 are now in operatlon in Italy, the U.S., and
Japan, thus bringing FHS into operation in the largest area of production
activity. Data indicate that over 60% of all production activities within
engineering involve assembly amenable to flexible automation.r,
~/   Bessant (1986).



- 16 -
1.29        These developments are rapidly expanding the range of engineering
sector activities where FMS can be applied, but the process of EMS diffusion
will not be confined to engineering--the philosophy of flexible manufacturing
is relevant to all batch-based industries. Increased flexibility and integra-
tion mean that the scale-related constraints that have operated in batch and
mass production are easing substantially. Minimum scale economies for mass
production at the plant and produc_ level are dropping, and batch producers
are able to achieve scale economies at much lower output levels.271 Hence, a
new group of users of flexible manufacturing technology is now emerging in
consumer appliances, plastics, forging, glass, electric motors, pumps, marine
components, valves, hand tools, switch gear, and even garments and shoes.
FMS: Imuact. Diffusion. Insestment
1.30        Evidence on the diffusion of FMS and on the gains arising from
their introduction is mounting rapidly. Some of the evidence is presented
here in table form. Table 1.4 shows results from some of the first large FMS
installed in the U.S.!!l Even at this early stage, the pattern of systemic
gains in labor productivity, reduction in cycle time, and product flexibility
was established. Table 1.5 gives a partial summary of the more recent (1986)
UN Economic Commission for Europe (ECE) survey of more than 100 FMS throughout
the world. Tables 1.6 and 1.7 provide a comparison between old and new
systems installed by U.S. and Japanese auto component firms who are planning
to introduce another 15 to 20 FMS lines soon.29
j/   See Kaplinsky (1986) and Hoffman and Kaplinsky (1987) for an extensive
discussion of the impact of flexibility on scale economies in batch
production. See Bessant (1986) for a discussion of FMS in other sectors,
and Hoffman and Rush (1988) for a study of FMS and other automation
technologies in the clothing sector. A recent dramatic example of the
reduction in scale economies attainable via flexible automation comes from
Nissan U.K.'s new engine plant where it will be economically viable to
produce 80,000 engines per year, as opposed to the previous break-even
point of 300,000 units annually for an in-house engine plant.  See the
Economist (1988).
2.w/ Ayres and Miller (1984) were among the first to review FMS experience in
the late 1970s.
22/  Hoffman and Kaplinsky (1988).



- 17 -
Table 1.4 Capital Costs and Reported Savings in Operating
Costs for Selocted U.S. Plexible Manufacturing Systems
lnstalled before 1982
User Sector            Cost              Product         Comparisons
(US$. 1982)        volume and       with old system
part variety
Truck axles          $5.6 milion        24.000*        1/4 floor space;
45**       nso-up costs
elLminated
Alrcraft engines    $8.4 mllllon         24,000*        1/3 floor space;
9**       1/4 labor; 50%
number of part-
holding devices
Tractor             $18.0 million        50.000*        Cost: $18 million
components                               8-13**        to replace dedicated
transfer lUe at coat
of $28 million
Truck                $5.0 m llion        65.000*        Cost of PS8 te stme
components                                   5**       as dedicated
transfer line, wLth
comparable cycle
time but less
flexibility
Constructlon         $5.0 million         8,0000*       Total transit tine
equipment                                     8**      through system:
old: 8.5 hours:
now: 0.3 hour
Total numbs: of parts per yoar machined on systes
** Bumber of different part types machlned on systae
Source: adapted from Ayres and Miller (1984)
lgblo l.S: SMUTRY Of SELECTED RESULTS FROM ECE SURVEY OF FPS IN EUROPE
*   For a subet of 39 fnstallatfone, 21 have redueed operator nus  rr
betwen 502 and 752, w*  fn 10 fnstallatlon the number of workers
has Ben redcLed by more than 75S over conwentionel lines.
*   In 28 intallations, 19 have reduced the  orer of machifn  tools
by more than 502 compared to conventional Lins.
* In 18 installations, all have reduced lead time by more than 25X;
of 31 additional installations, eight hae reduced Lead time by
between 50X and 752, and another 18 have reduced lead time by more
than 7'2
Source: Adapted from ECE (1986), Appendix 1



- 18 
Table i.6  FXS  inetalled by Japanese and U S.  Auto COMponUUtS FirMS
Japanoee iNS
Firm A  IS for brake assembly at 40,000 units pec month
Old In.O          INs
Capital costs                  3 x 130u (1)       Y120m (2)
Vorkers (3)                    12                 1
ToolLng                           Seas in both lines
Znerty                            Sass in both lines
Space                                             1/3
Number of produat  tyos        6                  20 (4)
Notes: (1) FM8 output equivalent to tbcee old lines
(2) Plus  devlopent costs associated vith tees of four or
fivs people vorking four to five years
(3) Requires significant increase in maintenance personnel
end capabilities
(4) Parts rccognized by "touch"
Flit b - FffS for assembly of speedonatecr
old line           FXS
Number of lines (1)            3                  1
Nusber of product types (2)   2-3                 50
Worters (3)                    30+                3
capital costs                  YlOO' Y300.
Notes. (1) F1S outpu  equivalene to three old lines
(2) Adjustuent fron one model to another is triggered by
dumy part liserted into line
(3) No etra saintenamco personnal required
Source: inatertws
Table 1.7  110 Installed by Three U.S. Coponents Flrm
FPr A             Firm                Fitu C
Old line  FNS      Old line  FnS       Old Llne  FMS
Product        Compressor          Alr conditioner    Brakeahoes
and compressor
Number            4       1           3        1          2       1
of lines
Capital crats   $2. 5     $ 12.      $300                        $2m
$800.000 $8-Lom
each
workers           24       4          16       2          5
per line
Space                                         50% reduction
Number of         1       10                   86         2       6
produce       per line                      different
types                                        pares
Output                     75%                  30%                 46%
increase             increase            increas
Comment                  35 manyears ot
software R&D
required
Source: tnterviews



- 19 -
1.31        Flnally, Table 1.8 gives details of the most recent survey of FMS
carried out in the U.K.S/ and highlights important trends. First, it
confirms the systemic gains described above: lead times reduced on average by
74% work in progress (WIP) by 68%, stock turnover improved by 3 to 4 times,
and machine utilization up from 50% for stand-alone CNC tools to 80% on
average for the systems listed. The table reveals shifts in FMS parameters,
lndicating what is ahead and confirming the general trends. The average cost
of the systems has declined from US$8-$15 million to US$4 million, with 72% of
the systems in the sample costing less. Batch sizes are being reduced
dramatically, with 58% of the sample handling batches of 10 or fewer units.
Before they could cope only wlth small-parts families; now 22% of the sample
can handle more than 100 different parts. Flnally, the size of user firms is
dropplng, with 69% of the sample FHS in companies with fewer than 1,000
employees.
Table 1.8 Summary of Gains from FMS
Used by 50 U.K. Engineering Firms, 1986
Lead time reduction         74%
Work-in-progress
inventory reduction         68%
Increase in machine use    63%
Turnover improvement        3.5 fold
Source: Bessant and Haywood(1986)
1.32        Trends in investment and distribution.  The few reliable studies
on the pattern of FHS investment document a rising rate of diffusion across a
growing range of sectors and users. The ECE in 1986 estimated world stock of
FMUs to be in the region of tens of thousands and the total stock of FMCs
around several thousands.31' Estimates for the stock of installed FMS vary
widely.H4 One estimate indicates about 1,000 now installed, with an expected
annual rate of growth of over 20%.!
I/   Bessant ar-i Haywood (1986).
2./ Only fragmentary evidence on the rate of growth is available; the stock
of the U.S. installed machining centers (key components of all flexible
systems) grew from 17,000 in 1978 to 24,000 in 1983.
i2/   ECE (1986).
w    Bessant and Rush (1987).



- 20 -
1.33        Expectations are that the shift to flexible manufacturing will
increase significantly over the next ten years, with the share of systems
investment as part of investment in total factory automation rising consider-
ably. Estimates of total investment in automation vary from $32.3 billion to
$200-300 billion by i990--with $100 billion expected to be spent in the U.S.
alone.!Y Table 1.9, drawn from the lower estimate, shows how the amount
spent for systemic elements such as factory computers and programmable
controllers, already quite high by 1985, is expected to increase substantially
by 1990.L5' Obviously, country-specific patterns of future investment in
factory automation will vary in size and in configuration, with the U.S. and
Japan expected to be the leaders (Table 1.10).
Table 1.9 Estimated Allocation of Investments in
Factory Automation
(US$ millions)
1985        1990         1995
Factory computers and software        935        2500         6500
Materials handling systems           2000        4500         8000
Machine tools and controls           3000        4800         7000
Programmable controllers               50          550        3000
Robots and sensors                     65          660        2800
Automatic transfer and equipment      800         2000        4000
Total spending on automation        $6850       $15010      $31300
Source: Dataquest, cited in Port (1986)
14J   See ECE (1986) for a comprehensive review, and The Economist (1987);
Saporito (1987); Port (1986); and Arbose (1985) for different numbers
relating to the future. The Economist quotes Hewlett-Packard executives
saying  that 96,000 factories in the United States are currently in the
process of installing CIM in one form or another. In a year or two, many
of them will become global forces to be reckoned with; some will no doubt
conquer the fbreign competition and become the new market leaders. " Arbose
(1985) quotes a vice president at ADL Ltd. who estimates that "between 50%
or 60% of the manufacturing companies in Europe, the U.S. and Japan have
made some progress (towards integration) by adopting CAD, HRP (Manufactur-
ing Resource Planning), robots and FMS. The remaining firms have little
hope of being internationally competitive in the 1990s."
i2J   Arbose (1985) cites an A.D. Little study that predicts 50% of a company's
in-house computing power will be devoted to manufacturing and engineering
by 1990, compared to 25% today.



- 21 -
Table 1.10 Number of Fully Integrated
Flexible Manufacturing Systems
Estimated to be in Use by 1990
Japan             320
USA               150
West Germany       70
France             45
UK                 35
Italy              30]
Source: The Economist (1987)
1.34        Despite the buoyant optimism underlying these estimates, when the
technological future will arrive in entirety is unpredictable. The integra-
tion of different elements of automation technology--particularly across
different spheres of activity--is in fact technically extremely complex.
Similarly, as discussed below, FMS are difficult systems to operate, requirlng
sophisticated supporting services and highly trained workers. Both these
factors will inevitably retard the diffusion process and limlt the cost-
effective use of FMS to firuas exhibiting a specific set of product charac-
teristics, total output, and scale of production runs. Thus, large numbers of
small and intermediate firms that, in principle, should be able to benefit,
will in practice find the use of full-fledged FMS an unviable proposition for
a considerable period.
Lesss_ from the Diffusion Process, the Imoortance of Organizational
CompatibillL
1.35        As the last set of comments indicates, although a fundamental
technological transformation is underway in engineering, it is proving to be a
slow and painful process at the firm level. Case studies on the introduction
of advanced manufacturing technology indicate significant problems. In many
cases, returns on the investment have been much lower than hoped for, and the
learning period has been protracted and difficult. For example, a study by
the British Institute of Management revealed that of 64 firms with FMS, more
than two-thirds had achieved only a "low payback." A larger survey of 250
firms showed a similar pattern of unfulfilled expectations in connection with
a variety of automation technologies. A number of analysts !!I report
problems with the fntroduction of complex integrated systems in the European
engineering industry mnd hlghlight the growing caution among potential users
that already has caused substantial downward revisions of investment plans.
i§/   Bessant (1986), Senker and Beesley (1986), Boddy and Buchanan (1985), and
Blumberg and Gerwin (1984).



- 22 -
1.36        Parallel stories of problems and disenchantment with advanced
systems also have surfaced in the U.S. The massive wave of investment in
automation from 1980 to 1986 was carried out in full expectation that a
"technological fix" was the key to restoring competitiveness, particularly
against the Japanese. The stand-alone automati.n introduced over this period
was assimilated relatively easily, yet virtually all the firms that pursued
larger, more ambitious automation projects experienced unexpected difficul-
ties. Projects written up in glowing terms in the mid-1980s have since been
revealed to have been fraught with problems.37/
1.37        The best known case of technological snafu involves U.S. auto
industry attempts to repel the Japanese invasion of their markets by relying
on the massed technological battalions of factory automation. General Motors
(GM) led this charge with a US$40 billion investment plan that included US$5
billion for fully automated production of the Saturn car, US$20 billion to
automate old factories and build new ones, and the rest for a variety of
expensive high-technology projects. The advantages gained were meant to allow
GM to compete head-on against the Japanese (and Ford and Chrysler) and to re-
establish its pre-eminent position.
1.38        Instead, the company encountered nightmarish and costly complica-
tions, particularly in showcase projects.!!1 Consequently, eight years and
US$40 billion later, it is not the Japanese, nor Ford or Chrysler, but GM that
is in trouble. Its market share has declined from 48% in 1977 to below 36% in
1987, when it also plummeted to last place in productivity and profits. The
company faces the real possibility of having to shut down hundreds of
thousands of additional units of capacity by 1990 because it cannot match
standards of productivity And quality now being set by Japanese-owned but
U.S.-staffed plants in the U.S.39/
1.39        The need for organizational compatibility.  Explanations by
analysts of what has gone wrong with the introduction of advanced manufactur-
.22/ The Economist (1987) comments on many of these, while Port (1987) and
Whitney (1986) discuss other aspects.
i2J   Despite the $500 million spent to equip its Hamtranck Cadillac plant with
the largest army of robots and computers in any auto plant anywhere, that
plant is still nowhere near the quality and productivity of GM's NUMMI
joint venture with Toyota in Fremont, California, where automation is
limited but where Toyota-style management and organizational practices are
in place. Similarly, although $400 million was spent on refurbishing Buick
City with flexible tooling, the plant has had one of the slowest startups
in history and is still plagued with problems. And finally, there are
serious questions being raised about whether the Saturn project will ever
get off the ground. See the Economist (1987) and Business Week (1987) for
media reports, and Hoffman and Kaplinsky (1988) for an analysis of GM's
problems.
1_/  Womack (1987).



- 23 -
ing technology are illuminating. Western companies' high-tech indigestion
frequently resulted because the absorptive capacity of the company, plant, and
workforce was overwhelmed. Another aspect is that these highly automated
plants were, in fact, over-automated. A variety of technical problems,
particularly in interface software and network communications, proved dif-
ficult to solve at first, although these problems are being overcome through a
forced learning process--as shown by the success of smaller-scale FMS efforts
that followed the first round of projects.401
1.40        Besides technological overkill, two other sets of factors were
found to play major roles in absorption problems. Most fundamentally,
introducing sophisticated integrated manufacturing systems into plants that
operate inefficiently does not eliminate the original problems and often makes
them worse. In the words of one U.K. manager, "all you get when you put a
computer into a chaotic factory is computerized chaos.n4" Thus, automating
existing activities without giving careful thought as to how these might be
rationalized and simplified was found to be a recipe for disaster.42/
1.41        Detailed research on these problems has shown that the key to
success doe- not lie in technological wizardy but in planning and preparation.
Successful _..irms first set out to learn how their factories function and,
operating on the principle that complexity is the root of all evil, then
simplified all procedures and tasks, eliminating everything that did not
contribute to value. This eliminated many steps in the manufacturing process,
including tasks that initially seemed candidates for automation.!!/ Within
this approach, automation takes place incrementally, which is less costly and
much easier to absorb. Such rationalization in fact can do away with the need
for highly integrated systems such as FMS.!!/
1.42        The second set of factors for successful implementation of
flexible manufacturing technology involves achieving compatibility between the
technology and the organization into which it is being introduced. The scope
and depth of the organizational accommodation required in advanced manufactur-
ing systems differ substantially from that previously encountered in stand-
gQ/   See The Economist (1987) for a discussion.
bl1/ Bessant and Rush (1987).
k2/  This is borne out by the successful cases of advanced manufacturing
computerization involving companies such as Allen-Bradley, John Deere,
Caterpillar, Hewlett-Packard and many others.   See Cook (1986), Feder
(1987) and The Economist (1987) for examples.
k3/ Schonberger (1986) carries a number of examples of this process, including
that of one engineering company, which after spending millions installing
a computerized inventory and production monitoring system, found that it
was no longer necessary once the company simplified its production process.
k4/ Schonberger (1986) and Bessant and Rush (1987).



- 24 -
alone technology and go far beyond the need to simplify procedures. In
effect, an entirely new way of thinking about the management and organization
of production becomes necessary when firms move into integrated flexible
automation.45/
1.43        Two types of adaptation are particularly importan=:._
(a)   Functional integration, which implies minimizing of inter-depart-
mental and skill-based boundaries to reflect the integrated nature
of the technology. Such integration is necessary at all levels
but is particularly important in two area: First, to bring the
syztem into use, design and production staff need to work together
to develop suitable products. Enormous gains can result from
"designing for manufacturability" that would otherwise not
occur.47/ Similarly, in operating the system, new groupings of
multi-skilled operators and engineers are needed. These must have
close liaison with scheduling and marketing staff to ensure full
use of the system's flexibility.8'/ The basic objective of
greater functional integration is that specialist skills are not
eliminated but are applied in a more coordinated fashion by groups.
of people thinking in systematic terms about the manufacturing
process.
(b)   Vertical integration, which is necessary to match the technology's
flexibility and rapid response capacity. This implies the
creation of a managerial decision-making structure less hierarchi-
cal than traditional structures. The aim is to have a management
structure that is closer to production but also capable of a much
greater degree of delegated autonomy. The technology's flexibili-
ty allows-those with the most knowledge of the system and its
capabilities to make decisions in response to rapidly changing
market conditions. This also implies that engineering, super-
visory, and other technical staff move into a supporting role vis-
a-vis production groups.
4,/   Dempsey (1982).
L6/   Bessant and Rush (1987).
AZ/   The concept of "design for manufacture" is, perhaps, the latest term to
emerge from the study of Japanese organizational practices. Simply stated,
the idea is that products are designed explicitly to simplify the
manufacturing process. This approach makes eminent common sense but was
ignored by western firms until recently. The improvements recorded are
impressive--number of parts per product are reduced by 40%, 50% and even
80%,  greatly  simplifying  the  subsequent  production  process.    See
Schonberger (1987) for numerous examples.
AJ8   See Jones and Scott (1987) for an excellent case study on these issues,
as well as Handke (1982), Brodner (1985), and Wall (1986).



- 25 -
1.44        A great deal of experimentation is going on within technology user
firms on the best way to simplify procedures and achieve functional and
vertical integration. (Many of these issues are discussed in Chapter III.)
There is impressive evidence that organizational adaptation can have a
profound impact on manufacturing performance wherever new technology exists.
Most recent studies include strong arguments about the critical need for
organizational adaptation before flexible manufacturing systems are intro-
duced.49/ For instance, studies on the U.S. automobile industry recognize
that the organizational changes by Chrysler and Ford--patterned closely on
Japanese practices--as part of their multibillion dollar automation efforts
have been critically important in allowing Chrysler and Ford to sweep past GM
in productivity and profitability.!_/
1.45        Unfortunately for all three of the U.S. firms, they still lag far
behind the best-practice standards set by Honda and its North American
operations. Honda's first U.S. assembly plant (1982) relied on fixed automa-
tion and Japanese-style organization to gain a significant share of the U.S.
market. Building on this, the company has embarked on a massive flexible
automation scheme that will remove two-thirds of all labor input in machining
and assembly by the early 1990s.51/
1.46        Japanese superiority in flexible automation now dominates the
engineering sector--despite the common and mistaken assumption that the West,
particularly the U.S., is ahead in full CIM installations and is, therefore,
A9/   See references given in footnote 18 as well as Senker and Arnold (1983);
Fleck (1985); and Waterlow and Monniot (1986).
IQ/   See Altshuler, et al.  (1984); Hoffman and Kaplinsky (1988); and the
discussion in the next chapter.
IJ/   By this time, Honda will have a total capacity of 500,000 units, a world
scale engine facility, and a number of affiliated component plants bringing
local capacity to 75% for an investment of 1.7 billion. The major advances
Honda is making into the world of flexible automation are typical of
Japanese auto and other engineering firms. The key factor is that Japanese
firms, by and large, already have an organizational structure suited to
flexible automation. They have already won the battles western firms are
just beginning to fight in this area. This gives them an enormous jump
over western firms, even though many of the latter, particularly in the
U.S., invested earlier and more heavily in automation technology. The
evidence presented in Jaikumar (1987) attests to this. See also Womack
(1987) and Hoffman and Kaplinsky (1988). The combination of organizational
efficiency and flexible manufacturing that Honda has established is so
powerful that, in the words of a noted auto analyst, "the demonstration
effect of this facility on American industry in general and the auto
industry in particular is hard to overestimate.. .Senior executives, plant
managers and union leaders in American companies... [have] a simple
imperative: match Honda"s achievements in productivity, quality, and
factory flexibility in the United States by the early 1990s or be prepared
for bankruptcy." Womack (1987).



- 26 -
leading the technology race. In fact, since 1980 Japan has invested in
flexible systems at a much higher rate than other industrial nations--by more
than two-to-one compared to the U.S. Its factories boast over 40% of the
world stock in flexible tools--with small to medium producers using two-thirds
of them. The crucial lesson from this situation is that Japanese leadership
in flexible technology has emerged not because of greater investment but
through more efficient technology use.
1.47           Stunning confirmation of this assessment comes from a comparative
study of the performance of 95 FMS in use in the engineering seetor in the
U.S. and Japan. Comparison of U.S. and Japanese FMS performance are given in
Tables 1.11 and 1.12.
Table 1.11: PERFORMANCE COMPARISON OF FMS IN THE U.S. AND JAPAN
U.S.            Japan
Systems development time (years)                          2.5 to 3        1.25 to 1.73
Number of machinos per system                                 7                6
Types of parts produced per system                            10               90
Annual volumes per part                                    1,727              258
Number of parts produced per day                              88              129
Number of new parts introduced per year                        1               25
Number of systems with unattended operations                   0                1
Utilization rate over two shifts                              52X              64X
Average metal cutting time per day (hours)                   6.3               20
Sourcr Jaikumar (1987)
Table 1.12: MANPOWER REWUIREMENTS FOR METALCUTTING OPERATIONS
TO MAKE SAME NUMBER OF PARTS ON U.S. AND JAPANESE FMS
Conventional System       FNS
............. .................................
U.S.        Japan        Japan
Engineering                        34             18          16
Manufacturing overhead             64            22             5
Fabrication                        52            28            6
Assembly                           44            32            16
Total number of workers            194           100           43
Source: Jaikumar (1987)



- 27 -
1.48        The reasons for the differences lie in the design and operation of
the systems. U.S. managers approach their tasks with the perception that the
FMS is simply another set of machines for high-value standardized production.
They are still operating in terms of mass production and "old-fashioned
Taylorism and its principles of scientific management." (See footnote 79 in
Chapter II for an explanation of Taylorism.) As a result, U.S. firms pursue
precisely the wrong objectives: design is separated from execution; skilled
machinists are replaced by narrowly trained operators; output and uptime are
valued more than flexibility and process improvement.L2/ Such firms attempt
"narrow-purpose production on expensive FMS technology designed for high-
powered, flexible usage."!3'
1.49        The Japanese, however, have grasped fully the potential of the
technology. They have built on their already highly flexible production
organization, to develop the management structure and practices necessary to
extract the full benefit of flexible automation. Maximum flexibility and
reliability are designed into the system; small groups of systems engineers
design and run-in the installations, thereby capturing learning benefits that
feed directly into the design of the next version. Responsibility for
operation and innovation now falls to small teams of highly skilled, multi-
fuvicciun engineers, while senior management, relieved of daily production
conArerns that typically preoccupy their counterparts in the West, can con-
cen_ate on planning for changes in market demand.  These changes, particu-
larly the engineering bias (engineers now outnumber production workers by 9 to
1), "signal a fundamental change in the environment of manufacturing. In the
FMS environment, engineering innovation and engineering productivity hold the
keys to success." '
1.50 For U.S. and European engineering firms, the implications of these
findings roughly parallel the situation in motor vehicles--the choice is to
match Japanese standards for FMS performance or fall further behind.  They
must learn and apply lessons from the Japanese design and management of
flexible systems, and they must become aware of the implications of the
j2/   Jaikumar's conclusions on this are blunt:   "A close look at how U.S.
managers are actually using (FMS) technologies (shows) ... they are buying
the hardware of flexible automation -- but they are using it very poorly.
With few exceptions, the FMS installed in the U.S. show an astonishing
lack of flexibility. In many cases they perform worse than the conven-
tional technology they replace. U.S. companies use FMS the wrong way.
Compared with Japanese systems, those in the U.S. produce an order-of-
magnitude less parts. They cannot run untended, are not integrated with
the rest of their factories, and are less reliable. Even the good ones
form, at best, a small oasis in a desert of mediocrity.   Rather than
narrowing the competitive gap with Japan, the technology of automation is
widening it further.   The technology itself is not to blame;  it is
management that makes the difference," Jaikumar (1987).
ll/   Jaikumar (1987), p. 71.
Ei/   Jaikumar.



- 28 -
increased engineering intensity of production. Whether the response from U.S.
engineering firms and auto makers will be soon enough and sufficiently broad
to head off further Japanese advances remains to be seen.55/
1.51 However, two sets of conclusions emerge. First, the conventional fixed
automatidn technology of mass production and the Taylorist principles of
scientific management are becoming less and less relevant. With the minimum
efficient scale for flexible manufacturing now on the order of six machines--
and with fever people--size and scale no longer are overwhelming barriers to
market entry nor guarantees of competitive advantage. Managerial competence
and understanding of the systemic and strategic implications of flexibility
are the new determinants of market success. National capabilities in these
areas will determine the international distribution of benefits from the
technological transformation. The continued dominance of the U.S. and Europe
as the leading users and suppliers of engineering technology is not guaran-
teed. This portends a shift in the structure of the global system of produc-
tion and exchange in engineering, which is perhaps even more important than
the underlying technological transformation.
1.52        The second set of conclusions bears directly on the focus of the
rest of this paper says. The diffusion of flexible manufacturing technology
is still in the very early stages. However, it is significant that management
issues and organizational factors already are proving to be key determinants
of successful adoption and diffusion of the new technology--and hence are
influencing international competition and division of labor. Analysts
concerned with the implications of radical technical change for developing
countries have not yet understood the primacy of organizational over technical
factors in this regard since many of these observes have been even more guilty
of embracing the myth of technological determinism than have Western plant
managers.
1.53        Even more important are the consistent findings that the majority
of benefits from technology use come not from the technology itself but from
rationalization of procedures and organizational change. This indicates that
organizational change (and its benefits) are, in fact, separable from tech-
nological change. From the perspective of developing countries, this is a
factor of considerable importance. For those countries already operating in
international markets, the shift toward flexible automation as the dominant
production technique appears to imply both an intensifying of competition and
a considerable leap over the technological barriers to entry.
lL/   Port (1987) recounts a number of cases where technological and organiza-
tional changes are being introduced together, for example, in the giant
PPG Industries, which is designing and building a $30 million, highly
automated glass plant expected to be 50% more productive than any other
plant in the world -- and in which Japanese-style management and
organizational practices will be in place from the beginning.



- 29 -
1.54        However, the emergence of these new organizational innovations may
lead to productivity and performance improvement that are immediately within
the grasp of those countries. Equally significant, although trends in
flexible automation are important to engineering firms in some advanced
developing countries, the new approaches to production organization may in
fact be applicable across a broader array of sectors and countries. This
possibility raises many questions about the origin, nature, and characte-
ristics of the new organizational practices and their applicability in
developing countries, which are issues treated in the next two chapters.
[D-276(a)]



- 31 -
CHAPTER II
ORGANIZaTIONL INNOVATION IN THE ENGINEERING SECTOR
The Erosion of Western Comoetitiveness
2.01        Although a common topic of discussion today, the competitive woes
of Western industry only began to surface in the late 1970s, in the form of
dramatic alterations in international trade and economic relations among
industrialized countries. The U.S. economy, which had enjoyed global indust-
rial supremacy in the post-World War II era, experienced the greatest change
as its share of world trade declined and its trade balances plummeted. As
shown by Table 2.1 and Figure 2.1, this reversal in Western, particularly
U.S., economic fortunes coincided with the rise of Japan as a formidable
international competitor.
Table 2.1 Trade Dalano in Naniacteurng
(US billlans)
Year     l.S.          Japa       Vest Cormany
iaos    18.8           93.7          63.21
1981    11.8          115.6          61.7
1982    -4.3          104.0         67.5
1983    -31.0         110.3          58.7
1984   -87.4          127.9         60.5
1985   -107.5         107.7         59.5
Source; NatIonsl Resseach Council (1966) p. 14
2.02        In the late 1970s, Japanese exports established such a powerful
presence in specific segments of Western markets that embattled Western
producers regarded Japan as the main threat to their market share, as power-
fully demonstrated by the experience of the U.S. motor vehicles industry.
Whereas early in the 1970s, as Table 2.2 shows, foreign penetration into the
U.S. market was extremely limited, imported Japanese automobiles captured 20%
of the U.S. domestic market in only eight years.561
6/   These early gains in U.S. domestic market share by Japan marked the
beginning of a profound transformation in the U.S. auto industry   --
virtually forced upon it by the real possibility of having to cede the
majority of market share to Japanese producers. Even though substantial
changes have occurred in both the technological and organization practices
of the domestLc producers, U.S. industry is still being pummelled by the
superior efficieney and products of the Japanese auto makers. The Japanese
could land a car in the U.S. in 1980 at a cost that was US$2,000 less than
it cost U.S. firms to build a similar car at home. (The gap has closed
somewhat but still remains sizeable.) The net result of a decade of this
onslaught was a U.S. auto trade deficit with Japan of US$60 billion in
1986, a growing number of Japanese-owned plants in the U.S. capable of
producing in excess of 2 million units, and a massive overcapacity problem
that could force the U.S. industry to close down 600,000 to 700,000 units
of capacity by the early 1990s.



- 32 *
Table 2.2 Iwport Share of the U.S. Western Europe.
and Japanese Auto Markets
Imports to:        U.S.        Europe      Japan
Exports from: Europe Japan World   Japan  Rest of vorld
1962        4.8   0.1  4.9       n.d.       1.4
1968        8.9   1.6  10.5      0.6        0.5
1969        8.7   2.5  11.2      n.d.       0.4
1970        10.5   4.2  14.7     1.1        0.5
1971         9.0   5.9  14.9     1.6        0.5
1972         7.6   5.7  13.3     2.7        0.6
1973         9.0   6.2  15.2     3.6        0.8
1974         9.0   6.7  15.7     4.0        1.1
1975        8.9   9.3  18.2      5.2        1.1
1976         5.6   9.2  14.8     5.6        1.0
1977         6.3  12.0  18.2     6.2        1.0
1978         5.9  11.9  17.6     6.3        1.2
1979         5.6  17.0  22.6     7.2        1.3
1980         5.4  22.8  28.2     9.8        1.0
1981         5.8  23.0  28.8     9.1        0.7
1982         5.4  23.2  28.6     8.6        0.7
1983         6.3  19.7  26.0     9.9        n.d.
1984         5.2  18.3  23.5    10.L        1.62
1985         5.6  20.1  25.7    10.7        2.17
1986        n.d.  n.d. 28.3      n.d.       n.d.
Source: Hoffman and Kaplinasy (1988)
2.03        Motor vehicles was not the only product group in which Japan began
to amass a sizeable trade surplus at the expense of the U.S. (as well as the
U.K., Canada, and France). The same rapid growth took place in medium-
technology sectors such as machine tools, railroad equipment, construction
equipment, and agricultural machinery. Japanese growth was rapid in high-
technology sectors such as consumer electronics and electronic components, as
well as in the "sunset* industries, as shown by Figure 2.2.57/ This emergence
of Japan in the early 1980s, both as a threat to Western domestic markets and
as a serious challenger for world industrial leadershlp, set the backdrop
against which the features of the new organizational paradigm began to appear.
.2Z/  See the analysis in Cohen and Zysman.(1987) in Chapter 5; see also NAS
(1987) Chapter I.



- 33 -
Figure 2.1   Shares in World Trade in
Manufactures for U.S., Japan, France, West Germany
1960-1982
percent
Source: Cohen and Zysman (1987), p. 64
Figure 2.2 Trade Balance for Manufactures, ifxcluding
FlLgh Technology 1Vxports, for! TT*q*,9 JRen W!rgue
1,2                    ~~~~0
Surc    C and West Grmany   1967- 198
in milos  ac
of US$
Source: Cohen and Zysman (1987), P. 74



- 34 -
2.04        The astounding success. of Japan in the international economy
stimulated a search for explanations for this performance. Early opinion from
plant managers, senior executives, and union leaders was that Japanese market
share had grown because of unfair advantages gained by dumping and predatory
pricing, an undervalued exchange rate, a favorable domestic tax structure and
capital markets, and protected Japanese domestic markets -- or at the opposite
extreme, because Japanese firms employed much higher levels of automation than
Western firms. In short, it was believed that all of these exceptional
factors widely favored Japan; if forced to compete on a "level playing field"
with the West, Japanese firms would lose their advantage.58/
2.05        Detailed studies and extensive visits by U.S. and European
managers to Japanese firms in the early 1980s showed that initial Japanese
gains in market share derived from the extraordinary advantages the Japanese
enjoyed over Western firms in productivity and product quality.59/ This
superiority was not because of illegal actions or advanced automation but
because Japanese producers had developed an entirely new approach to produc-
tion management that differed fundamentally from both the European craft
tradition and the U.S. mass production model.
2.06        In principle, these differences revolved around the following
aspects:
*     Cooperation rather than confrontation between management and
workers and between users and suppliers;
e     Pervasive concern for quality in all aspects of design and
production instead of the singleminded pursuit of cost reduction
and volume;
*     Flexibility in product mix, output levels, and deployment of labor
instead of specialization, strict demarcation, and economies of
scale;
*     Production according to order rather than to stock;
*     Group effort rather than management focus on output (from hostile
individuals).
Responsibility for quality control, maintenance, materials ordering, and
continuous improvement of every task devolved to work groups while super-
visors, technicians, and engineers became support staff.
5/   In America, this reasoning was used by the auto industry to lobby for the
severely protectionist Voluntary Export Restraint Programme, designed to
limit further erosion of the market until U.S. firms could respond to the
threat. See Hoffman and Kaplinsky (1988).
52/   See  Hayes  (1981),  Hall  (1981),  Wheelright  (1981),  Garvin  (1983),
Schonberger (1982) and Altshuler et al. (1984) for the auto sector.



- 35 -
2.07        Many of these organizational innovations were developed by vehicle
assembly firms, led by Toyota. The net Lmpact of their application on
competitiveness in the auto industry was staggering.@1 Tables 2.3 and 2.4,
and Figure 2.3, present some select performance comparisons (standardized for
product and process technology) among Japanese, European, and U.S. car firms.
Table 2.3 Workers per Shift to Produce Similar Vehicles
in Japan and in Two European Countries, 1980
Japan            Europe 1    Europe 2
Daily volume                      360               379          308
Direct labor per shift
Stamping                   135                58           39
Metal assembly and
body shop                                   362          221
Paint shop                  40               151           82
Trim and final             175               453          301
Total direct labor per shift    350                1024          643
Indirect labor per shift
Inspectors:
Body                       1                28           31
Paint                      2                 6            7
Trim                       2                12            7
Final and repair          10                37           23
Total inspectors per shift         15                83           68
Material handling and line
feed                             10              328          113
Other                              35               601          174
Total indirect labor per shift   60                1012         355
Total direct and indirect         410              2036         998
Absenteeism                        1%              6.1%          5.1%
Source: Hoffman and Kaplinsky (1988)
LO/   Studies by Altshuler et al. (1984) and Shingo (1981).



- 36 -
Table 2.4 Porformance Comparison for Japanese and U.S. Engine
and Vehicle Assembly Plants, 1982-1984
Toyota Kamigo            Chrysler          Ford
englne plant            Trenton            Dearborn
Products          2.4 Itr/4-cyl            2.2 ltr/4-cyl      1. 6 ltr/4-cyl
2.0 ltr/4-cyl            inel turbo        HO; turbo; EFI
Plant size        310,000 sq. ft           2.2 million        2.2 mlILion
sq. ft            sq. ft
Hourly            180                      2,250              1.360
employment
Line rate         1,500/day                3.200/day          1.960/day
manhours
per engine         .96 hour                5.6 hours          5.55 hours
Shifts            2                        2                  1 assembly
2 machining
Tnventory         4-5 hours avg            2.5-5 hours avg   9.3 days avg
Wages             $11.35/hour              n.d.*              n.d.*
(excluding fringes)
Robots            none                     S                  n.d.
Suzuki                  Mazda              CHAD
KosaL                   Hofu               Lake Orion
completed 10/83         completed 10/82   cmlatead 1243
Products          Cultus (Chevy Sprint)   626                 CadiUc & 0ls
C-car
Plant size        553,270 sq. ft           1.Sm sq. ft        3.7m sq. ft
Employment        600*                     1,800              6,700
LLn  rate   '    30/hour                   62.5/hour          75/hour
60,000/year             240,000/year       260,000 year
Manhburs
pax car           20**                     L4***              48.7
Inventory         4 hours avg              30 minutes min.   6 hours avg
1 day max.        of hlgh cost
items
gages              $7.40 approz./hours     $10.55/hour        $12.67/hour
Robots            137                      155                157
* One shift
** Includes stamping and engine and transmission assembly
** Includes stamping
Source: Automotive Industries (1985)



- 37 -
Figure 2.3 Total Hours to Produce a Motor
Vehicle, in the American, Japanese, and West German
Motor Vehicle Industries, 1970-1i981
850                ...\  . .
840 _
CL
0 80
300sus  4    U e& 3u4 . 3
a)~ ~ va
-    200 'aUssebe  an  supplie  firms
2s  e .s 
Xsu d b  a ca  .    0  n 1
...              sear
sneluding management and production labor in
-final assembler and supplier firms
source: Altshulen   et Al (pQRf)r n,  let
2.08        As Figure 2.3. shows, Japanese automakers reduced the number of
hours required to build a car from 250 in 1970 to 130 by 1981. Table 2.3
shows the huge differences in both direct and indirect staff used by Japanese,
European, and U.S. auto firms to produce the same type and volume of vehicle.
Table 2.4 shows that differences in plant performance are not due to higher
levels of assembly automation, because the number of robots is the same or
they are not used at all.
2.09        Although the gap has closed in some areas, in 1987 the best
Japanese auto plants using the same number of manufacturing steps could
produce a vehicle with 50% of the unit labor input, and only 50% of the
defects, of average U.S. and European car plants.61/
jl/   Womack (1987).   Abernathy, Clark and Kantrow (1981), showed Ford and
Chevrolet defect rates per vehicles shipped to be 5-6 times that of Toyota
in 1979.



- 38 -
2.10        Although the auto industry provides the most widely publicized
examples of Japanese manufacturing superiority, studies document that the
practices and advantages are present across the whole of the Japanese engi-
neering-and manufacturing sector.!' A process of organizational innovation
leading to continuous productivity and quality improvement has been, in fact,
a feature of Japanese industrial management methods since the early 1950s .A3
There has been a consistent pattern of improvement in quality, productivity,
reductions in lead time and work in progress (WIP), and savings in physical
space, across many sectors, firms, and pLoduct types. Table 2.5 gives a
partial selection of organization-based gains in various engineering sector
firms. Table 2.6 shows that as a product's complexity grows across sectors so
does Japanese efficiency, compared to that of U.S. firms.
2.11        A landmark comparative study of U.S. and Japanese producers of air
conditioners showed that the failure rates of products from the best producers
(all Japanese) were between 500 and 1,000 times less than those of the lowest
quality producers (all U.S. firms).!1 Air conditioners are standard com-
modities produced by typical engineering companies in both countries, using
mature and unchanging technology in almost identical configurations. It is a
product frequently manufactured under license in developing countries.
Japanese firms had man-hour productivity levels up to four times greater than
U.S. firms. The total costs of quality for the Japanese (including preven-
tion, inspection, and failure costs) were less than half of the failure costs
(only) incurred by the best U.S. companies.
2.12        Although the study does not examine the impact of the observed
quality differences on international competitiveness, other industries such as
consumer electronics, semiconductors, photographic products, and agricultural
machinery have learned not to ignore the quality factor when facing competi-
tion from the Japanese.!A,
j2/   Clark (1979), Magaziner and Hout (1980), Hayes (1981), Wheelright (1981)
and Schonberger (1982).
iL/  Deming (1982), Walton (1986), Kobayashi (1986), Shingo (1986), Garvin
(1988) and Ishikawa (1985).
fi/   Garvin (1983).
Ql/   In 1980, Hewlett-Packard reported that after testing 300,000 16K RAM chips
from 3 U.S. and 3 Japanese manufacturers, the Japanese had a failure rate
of zero while the U.S. the rate was between 11% and 19%.  The well-
documented Japanese inroads into semiconductor markets are easy to explain
on the basis of these sorts of figures.   Similar stores are told by
Magaziner and Reich (1982) and Takamiya (1981) about the demise of the U.S.
and European color TV industries. See also Port (1987) and Helm (1987)
for other examples.



39-
Table 2.5 Selected Examples of Gains from Organizational
Innovation in Japanese Engineering Companies
Product         Time    Inventory   Lead-time      Space      Labor
period   reduction   reduction      saving    productivity
Precision     1973-83    (indexed)   significant   factor    (indexed)
parts                     100 - 50   ($15m savings   of 3      100   250
via waste
elimination)
Stamping/     1977-81    100 - 60    20 days  -   significant  100 - 178
machining                              - 8 days
Auto          1974-81    100 - 35    significant   significant  100 - 191
components
Machinery                   30 - 1   1/2 day -        100 - 50    30% gain
parts
Machining                 100 - 61    days to         100 - 32    from 10
line                                  minutes                     to 4
operators
Source: adapted from Suzaki (1987); Schonberger (1982), and interviews
Table 2.6 The Japanese Edge in Manufacturing
Product              Manufacturing steps      Labor index*
Automobile                   1200                    1.98
Fork-lift truck               900                   1.82
Auto engine                   250                    1.62
Automatic transmission        200                    1.41
Color television               80                   1.15
Steel sheet                    17                    1.00
* Ratio of hours of labor, U.S. vs Japan
Source: Boston Consulting Group, cited in Port (1987), p. 134



- 40 -
The Imnact of Organizational Change at the Firm Level
2.13        One effect of the growing awareness of the sources of Japanese
competitiveness was that, starting in the early 1980s, a concurrence began
building among analysts and observers on the necessity for organizational
reform within Western firms if they were to have any hope of competing against
Japanese products.!1 At the same time, a much more significant process of
change was starting on the shopfloor and in the boardrooms of manufacturing
companies in the U.S. and Europe. Western industrialists and managers began
to alter fundamentally their opinions about the Japanese approach. For
example, before visiting a variety of Japanese manufacturing plants, managers
from General Electric in the U.S.
"generally believed that the meaningful differences between Japanese and
American manufacturing were cultural and environmental. By the end (of
the visit), their opinions had changed. 'Nothing amazing was being
done.. .We have known all along how to do these things but have lacked
the discipline to follow through and do what we know how to do... It
would be much easier for us if the differences we saw were technological
or grand-strategy related.. .but they are not...The fact of the matter is
that the Japanese don't work harder. They just work more consistently,
and they work together.' (quotes from GE managers)"67 
2.14        The above sentiment is increasing among Western managers as
understanding of Japanese managerial and organizational practices spreads via
diverse conduits--visits to Japanese plants by Western managers; advice
provided by management consultants and advisors with expertise in the new
practices; seminars, workshops, and a variety of consciousness-raising
activities undertaken by industry associations and university-based academics;
and through the demonstration effect from Western firms and Japanese subsi-
diaries who are introducing the new practices in North America and Europe.6/
2.15        The main locus of change in Western industry continues to be in
the motor vehicles sector, characterized by a "frenzy of adaptation" to the
new standards of managerial best practice, to ensure their competitive
I6/   See Appendix I and the discussion and analysis in Abernathy, Clark, and
Kantrow (1983); Piore and Sabel (1984: Hayes and Wheelwright (1985); and
Cohen and Zysman (1987).
LI   Wheelwright (1981).
iL8    Chapter I disctussed how engineering firms have been forced to embrace the
new practices as part of their efforts to assimilate advanced manufacturing
technology. This process will continue, but diffusion of the new practices
has taken on a momentum of its own quite distinct from the technological
link, as firms increasingly appreciate the inherent value of organizational
innovation.



- 41 -
survival.691 Ford, of course, is the best known example of successful adapta-
tion to the new practices. GM is the best known example of failure to adapt
to the new practices, despite its tardy efforts to do so and its well-
publicized collaboration with Toyota.
2.16        Organizational change in the auto components segment is perhaps
even more frenetic. For the last five years, component firms have been under
tremendous pressure from assemblers (both domestic and foreign) to adopt the
new practices for consistent quality, on-demand delivery, and productivity-led
cost reductions. There are indications that most of the major auto component
firms in Europe and the U.S. are pursuing organizational change at some level
(and in turn attempting to pass the lessons on to their suppliers).70'/
2.17        Non-auto segments of the engineering industry also provide ample
firm-level evidence of adoption of new practices. Table 2.7 gives a compila-
tion of results from well-known engineering sector cases of organizational
change. Outside the engineering sector, large and small firms from virtually
all segments of manufacturing and services are introducing the new
practices. 71
L9/   See Hoffman and Kaplinsky (1988); Mitchell (1987); Bernstein and Zellner
(1987); and Hampton (1987) for the most recent reviews.
7/   Womack (1987) describes exchanges between Toyota and a variety of U.S.
components firms when Toyota was trying to determine the source of defects
in the components it was receiving for use in NUMMI. The exchanges are
enormously illuminating both about the pressure assemblers are placing upon
suppliers to change, and about the fundamentally different nature of
relationships between suppliers and buyers under the new practices. When
Toyota asked how the supplier's plant was run, they expected full
cooperation, as is the case with Japanese suppliers. The answer they moist
often received was that "how the supplier's plant is run was none of
Toyota's business!" Toyota's reply was in effect that "if you want a long-
term relationship with Toyota, your business and Toyota's business are one.
Maintain your present attitude and our relationship is over."
ZJ/   Schonberger (1987) provides a listing of 84 firms that have introduced the
new practices from more than 15 different sectors.  See Helm and Buell
(1987) on Kodak against Fuji; Work et al. (1987) on Goodyear Tire and
Rubber Co.; Port (1987) for AT&T, Corning Glass, Dupont, Ford, IBM,
Westinghouse and Hewlett-Packard; Schiller (1987) for GE, Whirlpool and
Raytheon; Hoerr (1987) on the ACTWU and Westinghouse Furniture Systems;
and Stavro (1986) on Caterpillar. See Fortune (December 19, 1988), "This
Cat is Acting Like a Tiger" for Caterpillar's experience in revamping its
plants and beating back the Japanese.



- 42
Table 2.7 Selected Exa ples of Gains from OrganLzatLonsl
Innovation La U.S. and European Engineering Firms
Product       Time   Inventory  Lead-tLme    Space    Labor
period  reductLon  reduction    saving   productLvity
Engine               15000 units 10 days   (stock turnover)
Small machinery      $20. $3.5u 25 days   (delayed   (hours per
2 days    delLveries)  unit)
40% 2%   330 200
Tractors    1980-84  31% on      signLficant significant  40%
average
Auto        1982-84  turns       I month     25% and     13% in
components          1.9 4.0    1 veek       eliminated   direct
stockroos    Labor
Source: Bessant and Rush (1987); Suzskl (1987)
2.18        Although anecdotal evidence is impressive, no comprehensive and
reliable diffusion study on the spread of organizational innovations exists;
thus, it is impossible to judge either the rate of diffusion or sector and
country coverage. 72/ Many observers argue that U.S. and European firms still
lag far behind Japan in the introduction of new practices and have not yet
perceived their'importance. Moreover, as discussed below, many problems have
attended the introduction of these practices into Western firms. Finally,
there is media and academic debate that Japanese management methods are so
culturally specific that there is little chance of broad assimilation by other
countries 73/
2.19        These qualifications have validity.  Diffusion is probably slower,
more difficult, and more uneven than case studies and the business literature
indicate, and Japanese culture and conditions certainly have played a major
D/   We cannot assess some of the more sweeping statements  found in the
literature, such as the recent claim made by a vice-president of Cummins
Engine Co. who, on the basis of his direct experience, stated "that more
than 75% of the tope 500 U.S. industrial corporations are now moving along
new management lines." Piper (1986).
Dj~/ See Johannson (1988) for reservations about the rated differences in the
U.S. See Fukuda (1986), Murakami (1982), Kono (1982), Weiss (1984), and
Matsumoto (1982) for a sample of the literature on the issue of cultural
specificity, as well as the discussions in Piore and Sabel (1984) and Cohen
and Zysman (1987).



- 43 -
role in shaping Japanese managerial philosophy.74  Since processes of para-
digmatic diffusion from country to country take place over a long period, they
may be obscure and unnoticed by the contemporary observer. Nevertheless,
empirical evidence indicates that large parts of the new system and its
specific practices are transferable to different countries and to firms
operating in sectors greatly differing from the motor vehicle industry.
2.20        The reasons underlying firms' decisions to change their managerial
philosophy are straightforward. Some firms must adopt the new practices to
survive competition from the Japanese and the NICs. Others change because of
demands for greater flexibility and higher quality in the marketplace or
because they recognize, belatedly, that flexible automation cannot be exploit-
ed fully without a complete overhaul of organization and management practices.
Perhaps the most telling reasons why many industrialists now are willing to
change their philosophy are that the logic of the new principles is rooted not
in theory but in clearly superior production practices whose implementation
yields immediate, visible, and substantial benefits.
2.21        There is widespread and growing consensus that the process of
inter- and intra-firm organizational change will be one of the central
phenomena of global industrial development into the twenty-first century.
This possibility raises immediate questions on the relevance of modern
practices to the specific conditions of product.on in developing countries.
2.22        For a balanced assessment of the question of applicability, an
examination of the new approaches to production organization from the stand-
point of both theory and practice is necessary. Two bodies of knowledge can
be consulted for this purpose. First are the writings--and experience--of the
"practitioners," a small but influential group of managers, consultants, and
;gurus who have been instrumental in the spread and adoption of the new
industrial organizational principles in the U.S. and Europe. The second
source of information are the researchers and industry analysts who have
studied the processes of organizational change and the problems and effects of
firm-level efforts to apply the new practices. They provide studies less
concerned with "nuts and bolts" and more focused on general lessons derived
from systematic analysis of empirical evidence. To synthesize these distinct
bodies of knowledge, we discuss in the remainder of this chapter the several
principles and unifying themes that form the philosophical foundations of the
new approaches to management and production organization. In Chapter III, the
techniques and practices them-selves are discussed in detail.75/
Z4/   Piore and Sabel (1984) show how particular circumstances in the American
economy heavily influenced the shape of the mass production paradigm and
how its path of diffusion in other countries was determined by conditions
in those economies. We would expect the same process to be at work in
relation to the inter-country spread of the new practices.
D5/  Much of this material will be known to those relatively few readers already
familiar with these issues. "ey might chose to skip to the last chapter.
For those not familiar with the nature and logic of the new practices, the
following presentation will be enlightening.



44 -
Management as the Driving Force--the Workforce as the Source of Improvement
2.23        The strongest theme in both practitioner and analytical literature
is that the new organizational practices will not be successful unless top
management understands their logic and commits fully to the fundamental
changes involved.751 This transformation involves pervasive shifts in the way
managers view the production process; in their perceptions of responsibility
and control; and in their personal and corporate ideology regarding management
of the workforce. Managers must acknowledge that responsibility for most past
and present problems does not lie with the workers but with the rules,
procedures, working environment, and production systems that management itself
has created. Perhaps even more difficult, management also must transfer an
unprecedented degree of responsibility for quality, innovative effort, and
production organization to line workers.77/
2.24        The philosophy underlying this approach contrasts sharply with
standard (mass production) practice in the West. Its essence is that the most
important source of gains in productivity and quality are the skills, knowle-
dge, and expertise of the people directly involved in the production process
day to day. Thus, rather than seeing workers as the cause of problems,
_/   Achieving those changes is no easy task:  The old idea that a manager's
main function is to control workers is replaced with the concept that a
manager should encourage employees to use initiative ... To accept the
commitment model of work, managers have to go through a 'personal paradigm
shift' which is a deep psychological process."   Fraser, 1986, p. 78,
quoting a U.S. food company executive with direct experience introducing
the new practices.
ZZ/   Schwarz, 1986 sums up the differences between the Western and Japanese
attitudes to the workforce in an article that documents the success of
Japanese managers introducing their methods in the U.S.:  "The Japanese
are saving a lot of time and money in the U.S. because they start with the
assumption that their workers are intelligent, rational, motivated human
being ... Western managers on the other hand often think they are dealing
with dumb gorillas.  Western firms have been forced to build a complex
system to manage an adversarial relationship with its workers.   The
Japanese invasion may help bring peace."



- 45 -
management's job is to create conditions that encourage workers to apply their
creative energy and skill to the solution of problems.!Y
2.25        The Re-Education of Management.  Much academic lite_acure in this
field emphasizes the need for management commitment but is devoid of sugges-
tions on how to bring about this transformation, the most difficult step in
the whole process. Some practitioners recognize this problem and argue that,
as a first step, managers have to experience organized and structured re-
education. This is necessary even with a genuine commitment to change because
the new principles and their implications for action are so alien to standard
practice that managers frequently do not know where to begin.
2.26        Management re-education can be undertaken by means of an arms-
length approach, which involves management attendance at practitioners'
seminars, visits to 'converted" firms, and the close study of codified
practices and principles in the "cookbooks" (case studies) available.
Alternatively, companies can bring in outside consultants to advise and work
actively with managers to implement changes.
2.27        Companies also can enter into joint ventures with firms already
expert in the new practices. This arrangement allows managers to learn how
the principles are applied, by being directly involved in their applica-
tion. 79
8J   If we look back at the fundamentals of scientific management as laid down
by Frederic Taylor, the differences in approach to managing the workforce
are revealed starkly. Accorling to Taylor, four principles were involved
in "scientific management"--all intended to eliminate sources of worker
power and control and enlarge the scope of managerial responsibility.
Management had to absorb and codify the skills of workers and reduce these
to rules.  "All possible brain work should be removed from the shop and
centered in the planning and laying out department."   The increasing
division of labor should lead to the separation of "direct" from "indirect"
tasks such as machine set-up, preparation, maintenance and repair.
Management should specify the tasks of the workers. All of this was to
be achieved via the development of eight functional layers of management.
See Hoffman and Kaplinsky (1987) for an extensive discussion.
D9/   By far the best known attempt at joint venture learning is the GM/Toyota
NUMMI project. GM entered into the project to master the Toyota production
system, and Toyota got involved to learn how to transplant its system into
the North American business environment. The project is expensive (US$200
million by Toyota and an entire plant by GM), but under Toyota management
it has achieved world-class levels of productivity and quality. In the
course of this, Toyota has educated more than 300 of its managers on
training American workers, but GM has managed to train fully only 45 to
50 of its managers in the Toyota system. This might be because of the
greater difficulty of GM's task (re-education of managers firmly entrenched
in GM's mass production culture), or because the companies followed
fundamentally different training procedures. This issue is worth further
examination.



- 46 -
2.28        These management re-education methods have shortcomings that might
prove particularly difficult to overcome in a developing country context (an
issue returned to in the Conclusion). Whatever approach is selected, the aim
must be to engage managers in activities that provide' a definite and visible
learning focus and that allow them to translate into practical terms what the
new practices mean for the workforce and for the way production is organized'
in their plants.
The Ouality Factor
2.29        A pervasive commitment to all aspects of aualitv at all levels of
the firm is the second unifying theme of the new practices. The issue of
quality has a long history in U.S. industry, beginning with a focus on
"quality control by leading management authorities in the first decades of
this century.".80' Following this, the "quality assurance" era spanned the
1950s and 1960s, when concepts such as the cost of quality, total quality
control, reliability engineering, and zero defects became popular. However,
throughout the 1970s, the approach to quality in the U.S. was largely defen-
sive (identifying rather than preventing defects); its attainment was a
specialist function; and management was rarely concerned with it as a source
of productivity gains and market share.
2.30        This was not the case in Japan.  U.S. experts ell introduced
quality techniques to Japan through seminars and workshops in the 1950s.
These ideas were spread widely by the Union of Japanese Scientists and
L0/   Quality control as a concept was first introduced by Taylor (1919) and
Radford (1922) and given the classic treatment by Shewart (1931). In these
early texts, quality control was seen as necessary; it was carried out via
inspection and was a specific function of shopfloor supervision. Shewart
(1931) introduced scientific and sratistical rigor into quality inspec-
tion, elevating it to a task worthy of specialized expertise.
§_l/ Deming, Juran, and Feigenbaum. Statistical and management techniques for
pursuing these objectives on the factory floor were developed by people
such as Juran (1951), Feigenbaum (1956), Halpin (1966), and Crosby (1979).



- 47 -
Engineers (JUSE) and the Japanese Federation of Economic Organizations.82/
With the development of quality control circles and company-wide quality
control in the 1960s, the Japanese approach became an all-embracing concept in
theory and in practice.83! Quality as a source of competitive advantage was
by then firmly entrenched as a principle preoccupation of workers, super-
2_/  See Garvin (1988); Walton (1987) and Deming (1982) for a discussion of the
evolution of Japanese quality concerns. Deming (1982) cites how Mr. K.
Koyangi (founder of JUSE) reported at the 1952 meeting of the American
Society for Quality Control that great strides inn quality and productivity
improvement had been accomplished by thirteen top Japanese companies,
following top management attendance at the 1950 lectures--Furukawa Electric
Company reported that within six months of applying the practices outlined
in the lectures, it achieved a 90% reduction in rework in its cable and
insulated wire plants, as well as a great reduction in the frequency of
accidents; Tanabe Pharmaceutical Company, where productivity went up by
300% via process improvements, identified through the application of
quality principles; and finally, Fuji Steel Co. reduced by 29% the amount
of fuel needed to produce a ton of steel after only three months experience
with quality control.
[This author notes that many plants in developin;g countries today operating
with 1950s technology might well benefit from a similar commitment to
quality.]
_Q3/ Dr. K. Ishikawa introduced the concepts and techniques of quality circles
in Japanese industry in 1960.   By 1962, some 100,000 employees were
involved. By 1978 more than one million workers were involved in QCs; and
by 1984 the figure had risen to 1.5 million. This is impressive but the
important point to note is that quality circles are not crucial to Japanese
quality. They are simply one more mechanism employed in a constant search
to improve quality as part of an overall philosophy. However, as Lawler
and Mohrman, 1985 show, western firms missed this point entirely when they
aggressively pursued introduction of quality circles as one-off devices
to squeeze moire productivity out of workers but did so without a genuine
management commitment to quality. Most U.S. quality circles, in effect,
self-destructed after a short while because of management myopia on this
issue.



- 48 -
visors, and senior management--an approach that has proved enormously success-
ful1 4t
2.31        Elements of the Oualitv Philosoghy.  Four elements underpin the
approach to quality control now being widely pursued by Western firms.  First,
the competitive benefits arising from quality relate to both consumer satis-
faction and productive efficiency. Quality of product and service confer
demonstrably decisive advantages in the marketplace, and these concerns
typically underlie calls for firms to pursue the "strategic management of
quality." Less well appreciated is the fundamental point that the improvement
of quality leads to measurable cost reductions and productivity improvements.
2.32        The "costs" of quality include prevention costs, appraisal costs,
internal failure costs, and external failure costs. On average, about 40% of
total manufacturers' costs relate to quality. Eliminate quality problems, and
total costs decline substantially, both through eliminating activities
associated with quality costs and through improvements in the production
process.
2.33        Second, qu.lity improvement is not a one-off activity but a
continuous process. Improve quality continuously, and cc;ts go down conti-
nuously. The aim of quality improvement "... is perfection; anything less is
regarded as ann interim goal, to be succeeded by progressively tighter
standards ...  Until defects can no longer be found."85/  This attitude, of
course, runs counter to the common Western assumption that quality costs money
1_/   It has not all been easy.  Walton (1982) makes clear that many Japanese
companies had to struggle for years to get their quality practices to
acceptable levels. For example, the Hiroshima plant of Japan Steel Co.
faced many typical problems with quality and widely variable demand
patterns for its machinery products that were causing it to lose
profitability and market share inn the 1970s.   In 1977,  the company
embarked on a TQ effort that involved the use of JUSE consultants to train
staff and line workers in the use of various SQC measures, the introduction
of QCs and monetary rewards for employee suggestions.
Between 1978 and 1981, suggestions per employee rose from 5.6 per year to
28.5 per year, the cost of defects as a proportion of sales dropped from
1.57 to 0.4, production rose by 50% while the number of employees declined
from 2,400 to 1,900, throughout improved 100% and the accident rate dropped
from 15.7 to 2.3 per million man hours. The Hiroshima plant won the Deming
price in 1979, Japan Steel instituted TQ practices on a company wide level
at that time and itself won the Deming quality prize in 1983.
5J/   Garvin (1988).



- 49 -
and that there must be a point at which improvement stops, because further
improvement will not pay for the cost of improvement.86'
2.34        The thir  element is that the problems that cause defects are much
more cheaply and effectively dealt with at their point of origin rather than
later via inspection, rework, and field correction. The economic rationale
for this is straightforward, and the costs of not ensuring quality "at the
source" are staggering. Figure 2.4 (prepared by General Electric) captures
this point. The costs of error rise by an order of magnitude each time a
product or component moves a step further along the production process. Thus,
an error costing US$.003 to correct at the supplier stage costs US$300 to put
right by the time the product is in the field.
2.35        The fourth and final element in the quality equation is that
defects can arise at any point in the production process--design, materials,
machinery operation, poor training of operators, packaging, movement, and sale
of final goods. Thus, all staff and all functions must be involved in the
prevention and correction of defects and the improvement of quality, with two
groups frequently singled out for special emphasis. Line workers play a
particularly important role in the pursuit of quality, not because they cause
defects but because they know the production system and can identify causes of
defects better than anyone else. Design engineers also are a critical element
in quality concerns. This is not just because of the costs of faulty or
.6/   Deming (1986), Walter (1986), Hayes (1981) and Garvin (1988) all discuss
this aspect extensively and give examples of why the logic in fact runs
the other way. The concept that the continual improvement of quality leads
to the continual improvement of productivity and the reduction of costs
is a difficult one to grasp for traditional managers (a4a economists
brought up on marginal cost analysis). How does this happen?
"Quality is achieved by improvement of the process. Improvement of the
process increases uniformity of output of product, reduces re-work and
mistakes, reduces waste of manpower, machine time and material and thus
increases output with less effort ... Defects are not free. Somebody gets
paid to make them and it costs moire to correct them than to make the
product inn the first place ...  From 15 to 40% of the manufacturer's cost
of almost any American produict that you buy today is for waste embedded
inn it--waste of human effort, waste of machine time, loss of accompanying
burden.
Reduction of waste transfers man-hours and machine-hours from the
manufacture of defectives into the manufacture of additional good product
... the capacity of the production line is increased.  The benefits of
better quality through process improvement are the long-range improvement
of market position, greater productivity, much better profit, and improved
morale of the work force because they see management is making some effort
themselves, and not blaming all faults on production workers. (Deming,
1982).



c 50 -
unappealing designs but because, as mentioned in Chapter I, manufacturability
is seen as a quality and, therefore, a design issue. 7/
Figure 2.4 Escalation in Cost of Errors Down the
Production Line
..0.
SM~~~~~~~U
Low*
5Zrtu  1seu   Fwatu   Tea      FhZO      w"m
-^"F  Compt  C;au    C=Ug     r4        c=
e,U                    ca
* Estimated cost per defect per product
Source- cited in Garvin (1988)
The Elimination of Waste
2.36 The third common theme in the new practices concerns the problem of
waste in production--an issue first raised in the 1920s by Henry Ford. Waste
is conceived as "anything other than the minimum amount of equipment,
materials, parts, space and workers' time which are absolutely essential to
add value to the product. !_81 Any action component or procedure that does not
contribute to value must be eliminated. The origins of current concern with
waste derive from Toyota's efforts in the 1950s and 1960s to cope with the
introduction of product variety as a competitive weapon in gaining market
share in the Japanese auto industry.
iZI A corollary of the idea that all parts of the firm are responsible for
quality is that the customer determines and defines quality. The concept
of customer in the context of the new practices is broad, with the final
retail customers being the ultimate arbiter of quality through their buying
decisions. At the same time, every worker at every stage in the produc-
tion process is both "customer" (of the output of the previous stage) and
"supplier" (of inputs to the next stage). Each supplier must strive to
satisfy the quality demanded by each customer in the production chain.
See Walton and Deming (1982).
1-/  Suzaki (1987).



- 51 -
2.37        Though most frequently referred to in the context of inventory
reduction within the just-in-time (JIT) production system, the focus on
elimination of waste-has numerous dimensions that underpin its role as a
unifying concern. Figure 2.5 lists multiple sources of waste. These exist in
all factories, not just because of sloppy management but because they are
built into the system and procedures of the mass production model (emulated
even by firms not involved in mass production)."'I
2.38        The different categories of waste all generate by-products that
inevitably drive up costs and reduce productivity. Figure 2.5 captures this
aspect of waste from overproduction, but similar costs could be identified
each of the various categories. As with quality problems, the elimination of
waste reduces costs and improves productivity.
igure 2.5 The Seven Wates
1).V ate from overproduction
2) Wste of waiting time
3) Transporttion wvst
4) ProcessLng waste
5) Irventory Wate
6) Veset of motion
7) Uat, from product defects
Waste Resulting from Overproduction:
Extrait veniory
/    ~~~~Etxra h'-dILoS
t  eoeExtr sp¢e
Wt Vst of tr nt rt chacges
verproductionreamor 
/                     J  S~~~~~~~~~~Lxtra c ¢hin ry
Extra defeat
ftua papermck        o4xtr  verh zd
Extra p ople
Overpro~ductioan creatoes wee problo s *d obscures thc ro l
cause of the problem.
Sources Suraki (1987)
The function of inventories in mass production is to  -llow continual
operation of the plant. The organizing principle at work is a supply-
driven one, with all attention focused on keeping the line running at any
cost. Quality concerns are secondary, and relations with suppliers are
adversarial because price is the major determinrit of contract placement.
To this end, large inventories of components and work-in-progress is the
norm in case anything goes wrong in terms of (frequent) poor component
quality, late delivery, equipment breakdown, etc.



- 52 -
2.39       Deliberate steps to identify the causes of waste will eliminate
waste-related costs. Some of.these steps are straightforward, such as the
establishment of procedures for workplace organization. Others are more
difficult, such as the elimination of inventories and the reduction of set-up
and changeover times. Figure 2.6, taken from Toyota's 1960 Industrial
Enginearing Manual, illustrates how inventories mask various forms of waste.
Management must systematically drive down inventories (the water level in the
figure) to reveal defects in the system (the rocks), rectify them, and capture
savings in WIP.
Figure 2.6 Toyota's View of the Inventory Problem
IF THE RIVER IS DEEP, THERE IS NO BOTTOM IN SIGHT!
.....   ...  ...   ...  ...    ...  ...    ..., i
-. *--   *- -----            - --  ---* --    -          -.
... ~ ... ..     .. ..    ..     ..._
......   .....  ........  ....    ......     . ....   .. ...  .....     ..... 
0
....    .....    ...   ....      .....  . ..... . ..... .  ....O
M_achinery malfunctionX
.. .... \ ~~....                 ...........  ...... v
Co- eyance          i;tng tin            Faulty  -       -
5 Q wasite                a     C         east a
.0o
Inventory changes with the production syst.m
Inventory changes with the manufacturing method
Source: Toyota Industries



53 -
2.40        Tying it all together.  Although it is unlikely that most Western,
developing country, or Japanese firms will ever approach the level of sophis-
tication of Toyota or Honda, it is useful to show briefly how the principles
of quality improvement and waste elimination are put into practice in a
worker-led production process. Basically, two systems work together, JIT
inventory control and total quality control (TQC). The basic features of JIT
and many aspects of TQC systems emerged initially in response to specific
conditions in the Japanese auto sector.90/
2.41        JIT.  The critical organizing principle of JIT emerged through a
long process of trial and error from the 1950s to the 1970s and is elegant in
its simplicity. Production lines, originally structured according to mass
production principles, were reorganized so that work passed from one stage to
the next without intermediate storage. Machines (or people) at one step thus
produced the pieces required for the next step, from fabrication through final
assembly. The ideal lot size is one piece at a time--i.e., job lots of one.
Consequently, work flow is on a "pull-through" basis, with production at any
stage occurring only when the next stage in the line is ready to receive the
output of the previous stage. Tool and line changeover is short, through
20/   In 1950, Toyota, soon followed by other firms, made a strategic decision
to introduce product variety to gain market share in the auto sector, since
it could not compate via the pursuit of scale economies. However, the
firm had to deal with a range of problems arising from the decision to
produce a variety of models in small lots.  Product variety introduced
complexity into a production process whose equipment was typically designed
for high-volume dedicated production of, basically, only one model.
Product variety demanded that what was, in effect, a demand-driven
production process be highly flexible to minimize inventory carrying costs
and the costs of line and tool changeover, both major categories of costs
inherent in (supply-driven) mass production.
Reducing production runs exerted great pressure on materials handling
because of the resultant need to handle and store the multitude of parts
required to make a variety of products. Materials had to arrive at just
the right station at just the right time. The total volume involved could
not be stored near the line, the case in mass production. Otherwise, the
advantages of small-batch production are lost and machines and workers sit
idle. Hence, the marketing decision in the 1950s that targeted product
niches and demanded production line flexibility was also the first step
toward the JIT system.



- 54 -
engineering "quick change" capabilities, thus making it economical to run
small lots.911
2.42        TQO.  This system is most visible in workers' attitudes and
responses to the quality of components coming to and leaving their work-
stations; for example:
"Say that a worker makes one piece and hands it to a second worker whose
job is to join another piece to it, but the second worker cannot make
them fit, because the first worker made a defective part. The second
worker wants to meet his quota and does not like being stopped, so he
lets the first worker know about it right away. The first worker's
reactions are predictable; he tries not to foul up again--and tries to
root out the problem that caused the defective part.
The typical western way by contrast, is to make parts in large lots--two
weeks' worth maybe. The second worker might find 10% to be defective
but he does not care. He just tosses a defective part into a scrap
rework bin and grabs another. There are enough good ones to keep him
busy, so why complain about defectives! The Japanese cut the wasted
hours and wasted materials by not allowing large lots of defectives to
be produced."921
2.43        A Schematic View of JIT/TOC.  Figure 2.7 shows a JIT/TQC produc-
tion line as a cause-e£fect chain reaction of events. The starting point for
the chain reaction is the deliberate lot-size reductions shown in the double-
bordered rectangle A. This forces a reduction in set-up times to facilitate
fast changeover (not shown), then leads to less inventory carrying costs (A)
and reductions in other indirect costs (I).
2.44        Smaller lot sizes lead to a round of direct and indirect savings
due to scrap and quality improvements (B, C and D). This occurs because of
the interaction between workers over quality, depicted by the E-F feedback
loop. Workers (organized in small teams), supervisors, engineers, all search
for solutions to delays or quality problems which inhibit the smooth flow-
through ideas for improvements because of their higher awareness of the
21/   In mass production, a change in the specification of output necessarily
involves the time-consuming task of resetting (inflexible) machinery. The
costs involved in this (due to lost output) are very high and the
changeover variable one of the major factors accounting for scale economies
in mass production--which meant that in the early 1980s a new engine
production facility would cost in the region of $500 million. Flexible
product schedules would prove an extremely costly innovation unless
changeover times could be reduced. Abegglen and Stalk, 1985, describe in
detail precisely how Toyota set about doing just this in the 1950s. The
now legendary reductions in changeover times were the result.. Changeover
times on multi-ton does and presses that on western mass production lines
took 8 hours, even days sometimes, to switch over were reduced to minutes.
2_Z. Schonberger (1982).



- 55 -
consequences. "Each good idea ripples through the JIT cause-effect chain.
Improvement feeds upon improvement."931
2.45       The right side of Figure 2.7 shows the effects of perhaps the most
surprising of the techniques for managing the JIT system. On its own, the
lot-reduction process moves the system toward a smooth flow of output.
However, there are (and always will be) irregularities; in the normal course
of events, buffer inventories build up between stages and cushion the effect
of production irregularities. Japanese managers eliminate these buffers. As
soon as the system settles down, they deliberately remove buffer inventories
and/or workers (H) to set the cause-effect process off again. The result is a
continual round of productivity improvement--and the relentless heightening of
performance pressures on the workers.
2.46       In this model, TQC greatly enhances the quality control aspects of
JIT via its effects on B, C, D, E and F.  "Quality at the source" captures the
TQC practice of immediately solving problems as they appe.ar, with workers
taking steps to correct the defects or problems.
2.47       In the model represented in Figure 2.7, the JIT system clearly is
more than an inventory control system. It functions also as a quality and
scrap control tool, as a technique to help in reconfiguring plant layout, as a
means of raising process yield, as a method to achieve production line
balancing, and as a mechanism for employee involvement and motivation. The
output of the self-sustaining JIT/TQC system is thus a continual round of
improvements in productivity and product quality. Of course, this system did
not develop overnight, nor is it practiced fully in most plants. However, the
JIT/TQC model has tlecome the new best-practice standard that most Western
engineering (and manufacturing) firms will spend the next twenty years
emulating.
fipte 2.t JIT San TMC  as n tfttegtad System
teLtberaa withbdrevsk
(D)       oE buffer  dVancortLee
Hi tghtei evafenfss  torkes
of problee. and  <     \
ptobl_ cMM      (1)
Rduced buffet  tnwnettes
tnd/Og  ockers
Id   fo tS e I~to 6  deoa fordeas
.Cuttng  JIT deLL%&ry  ConctotLi (  C9)
lot stage  pertforance  detectS c   ta   to
i  co"grol  * rare.
CA/ (^    (t)    /C  (D)
/1..  1        dt eZt  feztwee rovr I ' / L"   "Catv I
"0 I Itav a   tn  for- tnt   ooe t o eutabor   (e lees mater
the wysc_  Lo Invetorty.
SPat mOd equipUO
ao handtl  tmncoy.    1
Inventory aCeounttft
hystcal Inventory cent
Loe matertal, tabor. and Indliret c   puts gor the U_ o  hiho
output   hMoer produstwtty.
Lee, invetory to the system - fetter aeckt cospoet. becter
forecasting. t   Los adt nitcraeton
io.cco Schonhetrgr (t962)
22/  Schonberger (1982).



- 57 -
CHAPTER III
ORGANIZATIONAL INNOVATION IN PRACTICE
3.01        This chapter focuses on the specifics of the new organizational
practices, giving the reader sufficient detail about the substance as well as
the underlying theory of these practices. Because of this concern with
detail, Chapter III concentrates on particularly relevant areas.!/
3.02        The discussion in this chapter is organized under four aspects of
management: management of the workforce, of machines, of the production
process, and of relations with suppliers. Under each topic is a discussion of
the techniques that have been developed to bring the principles discussed in
Chapter II into practice. The issues raised pose challenging implications for
developing countries and for firm and country strategies for industrial
competitiveness. These are discussed selectively as they arise, with most
analysis of these issues presented in the last chapter.
Managing the Workforce
3.03        Effective management of worker relations within the firm is
perhaps the most critical area because of the key role of the workforce in the
:ntroduction and daily implementation of the new practices. Both workforce
and management have to be prepared to alter attitudes. The complex way in
which individual, corporate, and social characteristics interact within the
workplace makes it difficult to specify steps or techniques, however. Case
studies show that worker suspicion of management objectives has represented an
obstacle to change, sometimes unjustified; in other cases, worker resistance
has been legitimate because management has fraudulently donned the attitude of
a "new era of cooperation with labor" while in effect pursuing the old
confrontational policies. /
2_/  Much of what is discussed is drawn directly from the work of practitioners,
such as Deming (1982, 1986) and Schonberger (1982, 1987).
2_/  Hoffman and Kaplinsky (1987) document a number of these instances in the
U.S. motor vehicle sector, most of them involving G.M. See also Shaiken
(1985), Katz (1985) and Jones and Scott (1987). There also have been cases
reported of labor suspicion and antagonism to management practices in the
U.S. subsidiaries of Japanese firms. One of these relates to an extreme
episode of labor-management strife at a Sanyo T.V. plant in Arkansas. When
it was initially taken over by Sanyo in the late 1970s, the company
received rave reviews (Schonberger, 1982) for JIT/TQC improvements that
supposedly had positive effects on productivity and employee morale.
Carson (1986) however reports a complete breakdown in relations, two
vicious strikes, and quotes labor representatives who claim that the much
publicized productivity improvements mainly came from an old-fashioned
speed-up of the assembly line. Clearly in this case and in others, more
information is needed to make a judgment.



- 58 -
3.04        Creating the Right Conditions.  Once the re-education of top
management is well underway or completed, the practitioners propose a number
of macro-management steps to introduce change and to begin convincing the
workforce of management's seriousness and commitment to the new approach.
Some of these steps are one-off measures introduced as part of the start-up
process, but others 'are meant to take place over much longer time or on a
continuous basis.
3.05        Forming teams and proiects to manage change.  The first step is to
create a team structure within the firm to ensure implementation of proposed
changes. Teams should include people with direct and indirect responsibility
for the problems being tackled. Management-led teams will play a major role
in the early stages, but line-worker teams, formed for problem solving or as
part of the reorganized production process, should handle shopfloor
projects. 96  A central idea here is that changes should be introduced in-
crementally via small, manageable projects that follow a "learning" cycle of
observation, planning, action, feedback, and reaction.!' The ideal practice
is to have a series of projects always underway in all parts of the company,
and that are aimed at improving performance via a manageable, attainable
process.
3.06        Inform the workforce at all levels.  The second step is to use
seminars, workshops and discussion groups to inform the entire workforce about
why changes are necessary, what principles are behind them, how changes are
to be implemented, and what the effects are likely.to be on the firm and its
workers. The aim is to create a critical base of people in the firm who
understand and support the changes.
3.07        Open channels of communication.  This education process must not
be a one-way process. Management at all levels needs early exposure to the
comments and criticisms of the workforce about the way the firm is run, its
products, processes, and working conditions. Management has to demonstrate
its willingness to respond positively to these criticisms. The aim is to lay
the groundwork for daily communication across all staff levels on production
problems and solving them. The problems that arise from lack of staff
communication and responsiveness are legendary and well-documented, but such
ii/   The most common model for organizing workers into teams in the engineering
industry is teams consisting of five to seven members led by a team leader
who usually has no direct production line responsibilities.   The team
leader is responsible for activities performed by supervisors, industrial
engineers, QC staff, and other specialists. These include organizing the
team's production and project responsibilities, as well as quality control
audits, team training, work standardization, preventive maintenance, and
paper work.  The team leader also fills in for absent team members--thus
creating strong peer pressure against absenteeism.
2_/  Deming, Suzaki and other practitioners use what is known as the Shewhart
Cycle to illustrate how this learning process moves forward.



- 59 -
situations persist throughout industry.! A crucial part of any attempt to
introduce the new management practices is to open communications among staff
and to create a sese of mutual responsibility and shared goals. Bringing
managers and engineers physically and mentally down to the shop-floor--desks
and all--so they can respond more effectively to line workers and solve
production problems is one solution.!D
3.08        Eliminate fear in the workRlace.  Two sets of problems charac-
terize working conditions in most Western--and developing country--plants and
factories. First, too often employees are afraid to point out problems, to
ask for help, or to make suggestions. Second, many obstacles exist in the
working environment that prevent workers from doing their jobs properly and
taking pride in their work.
3.09        Management can take a variety of specific steps to identify and
overcome these problems. These include addressing physical problems, such as
repairing faulty machinery, clearing up crowded working areas, or even halting
mindless exhortations to work harder. But the problems often derive from
management's adversarial relationship with its workforce, and solving such
2-8/ Typically, purchasing departments place orders based on written specifica-
tions they do not understand, buyers do not understand how the materials
they buy are used, designers constantly come up with products that give
engineers problems, engineers are often made to feel unwelcome on the
shopfloor.   Walton (1986) gives a description of what happened when a
Honeywell plant in Massachusetts "launched a wave of task-oriented project
teams that cut across department lines. As soon as departments started
talking to each other, some of the thorniest problems dissolved. There
was, for example, a case of delays on the Honeywell loading dock. Tracking
the shipments, a team discovered that there were traffic jams when outgoing
orders arrived simultaneously at a single freight elevator. An elevator
schedule was drafted immediately... In a related problem, a team learned
that orders often went to the wrong destination because of labels that were
difficult to read, for any number of reasons ranging from sloppy
handwriting to smeared printing. Sometimes the labels were slapped into
a side of the box that was hidden when the items were not properly stacked.
Eventually a computerized system was introduced for labelling and
instructions were issued for positioning the labels."
2_9/ Schonberger (1986).



- 60 -
problems will require actions that fundamentally alter the climate of
work. o
3.10        For example, workers must feel secure about making suggestions,
offering criticism or asking for help from supervisors--without fear of
retribution. Supervisors must focus on quality as their main objective and
must take problems on quality to upper management--and expect action.
Managers, supervisors, and engineers also must learn that their primary
attitude vis-a-vis the workers is to be supportive rather than dictatorial.
Performance incentives tied to quality and skill acquisition should replace
numerical quotas and output targets. This will stimulate workers to aim for
quality, pursue further training, and become active in improvement efforts.
The practitioners' literature provides examples that demonstrate how common
and costly problems rooted in fear are, and the gains that accrue when there
is a conscious attempt by management to deal with them.12 /
3.11        Institute adequate training and set clear Derformance standards.
One of the most common causes of production problems is insufficient worker
training and guidance because supervisors are more concerned about increasing
production or because management has not clearly defined quality and output
standards. A training period must be sufficient to qualify the worker fully
for the job. Quality and performance standards must clearly define what is
expected of the worker, and there must be explicit provision for continuous
training of workers to upgrade skills and develop capability to handle
malfunctions.
3.12        Provide Skills and Incentives.  Beyond addressing fundamental
issues of labor-management relations (even more problematic in developing
..I/  Deming (1982 and 1986) is one of the few practitioners to tackle this issue
head on. A quote from Deming (1982) and a newspaper reference illustrate
this point about fear: "The economic loss from fear is appalling. People
are afraid to point out problems for fear they will start an argument, or
worse, be blamed for the problem. Moreover, so seldom is anything done
to correct problems that there is no incentive to expose them... they re
afraid superiors will feel threatened and retaliate if they are too
assertive or ask too many questions.. .They are afraid to admit they made
a mistake, so the mistake is never rectified. In the perception of most
employees, preserving the status quo is the only safe course." Schandler
(1981) cites a U.S. government survey of 13,000 randomly selected employees
of 15 government subcontractors. Of the 8,600 who responded, 45% (4,000)
said they had personally observed or obtained direct evidence of wasteful
or illegal activity in the last 12 months. One in ten claimed the cost
of these events was in excess of $100,000. However, only 30% who claimed
knowledge of wasteful activities reported this to anyone else. Abour; 73%
of those who failed to report an impropriety said it was because they "did
not think anything would or could be done to correct the problem." Another
20% said they didn't say anything because it was "too risky."
101/ Walton (1987) provides a number of case studies that show how management
has attempted to deal with the issue of fear in the workplace.



- 61 -
countries), four other points need to be made about this topic. The first
relates to the shopfloor organization of workers. The basis for craft/job
distinctions disappears when workers are multi-skilled and perform a variety
of tasks. Typically, this means the negotiated disappearance of the job
classifications so common in Western plants. One example of this is at the
NUMMI plant where there are only four job classifications compared to 183 at
GM's Massachusetts plant organized along traditional lines.
3.13        The second point relates to the need for continual improvement of
machines and processes. Production under the new practices is highly engin-
eering-intensive because of continuous problem-solving and machine alteration
efforts. Firms must use their engineers more intensively and effectively.
More engineers per worker are necessary under the new practices; and the ratio
of engineers to production workers is likely to be much higher than current
ratios typical of Western and developing country plants.!2/
3.14        Third, under the new management philosophy, production workers are
central to quality control and productivity growth. To fulfill this function,
their skill base has to deepen and broaden (to meet quality control responsi-
bilities, allow multi-machine/assembly operation, lead improvement projects).
Table 3.1 captures the transition in worker skill levels from traditional
practices to the new systems. Average entry-level education and skill levels
need to be higher than under the old practices. Firms need access to con-
siderable training resources in order to raise workforce skills to the levels
demanded by the new system.
3.15        Fourth, incentive pay schemes are as important under the new
approach as they are under traditional practices. The major difference under
the new system is that workers are rewarded for producing quality goods, for
offering suggestions for improvement, for attaining higher skill levels, and
for achieving productivity gains. They are also compensated for length of
service. Many ways exist to structure such a merit-based payments system; the
key feature is to make employees aware of the connection between performance
and pay.!2/
1QZ/ Weiss (1986). The same phenomenon is identified in Jaikumar's study of
FMS usage in the U.S. and Japan.
103/ Under the old practices, firms pay the same hourly wages to their
employees.   Incentives offered are for production (incurring inventory
costs for the firm) or to overcome management mistakes (work overtime to
get late orders out or to replace poor quality products.



62 -
Tabte 3.1: TRANSITION IN EMPLOYEE SKILLS
UNDER NEW ORGANIZATIONAL PRACTICES
Old                                         New
Machine operation
Skfil is in the set-up               Sktll is in simplifying the set-up
Sometimes set-up technicians or      Operators Lead projects; technicians
engineers are needed                 and engineers help
Operator watches machine run         Operation is a well-timed rout1ne;
operator is thinking about next
improvement
Assembly
Assembly jobs were simplified        Assemblers acquire:
so unskilled labor could perform     o multiple job skitls
them                                 o data colLection duties
o diagnosis & problem-solving
talents
Source: Schonberger (1986), p. 38
3.16        Finally, although incentives have a role to play in encouraging
worker participation and attention to quality, they need augmentation from an
effective supervisory structure. This implies supervisor re-training and re-
orientation toward quality and away from increased output alone. Japanese
firms and Western companies that have implemented the new practices have
devised various techniques to stimulate management toward quality and conti-
nuous improvement. =
Managing Machines
3.17        The new organizational practices have a direct impact on virtually
every aspect of firms' operation, purchases, maintainance, and improvement of
machinery. Total preventive maintenance (TPM) is at the heart of this
concept. Under a full JIT/TQC production system, the need for completely
reliable machines capable of producing perfect quality output is paramount--
because machine-related downtime or defects soon cause stoppages. Japanese
industry's early concern with quality and quick change-over led to a set of
practices that now constitute TPM. The two most significant aspects of TPM
are (i) that maintaining machines in perfect quality makes good sense under
any circumstances, and (ii) that such practices are largely common sense and
transferable to any situation where operator-controlled machinery is used--
provided there are appropriate training and working conditions.
104/ Schonberger (1986) contains a number of examples on the day-to-day use of
these techniques by management in small companies.



- 63 -
3.18        Input from machine operators for TPM is wide ranging and includes
the following basic steps:
o     Daily checking of machines, involving operator checks on basic
functions and carrying out a number of minor tasks such as
cleaning, oiling, tightening, adjusting, sharpening
o     Keeping careful TPM records that include details on hours of
operation, preventive maintenance schedule, and the results of
tool, performance, and quality checks--all essential for keeping
track of major maintenance procedures
o     "Good housekeeping" of the machine areas, which involves cleaning
the area as well as taking care of tools and setting up. /
3.19        For the maintenance crew under TPM, similar procedures need to be
followed. Figure 3.1 captures the central principle of the "five whys"--in
practice there should be persistent questioning to identify causes and specify
solutions to problems. Figure 3.2 shows how the operator and the maintenance
efforts work together in TPM to F event machine breakdowns.
3.20        Constant imorovement of existing machines.  Beyornd TPM implica-
tions for operator training and skill demarcation, there are a number of other
machinery aspects under the new methods of organization, many linked directly
to the requirements imposed by the new practices.
(i)   Attaining a quick changeover capability means that equipment often
has to be modified; in many firms this can be a major engineering
job when the machinery varies by year, manufacture, and operating
parameters, all of which have to be standardized.2!/
105/  In addition, machine operators are expected to be alert for unusual sounds
or machine behavior that indicate wear or the need for repair. Operators
are expected to carry out minor repairs, such as changing and tightening
belts, replacing oil seals, even changing motors and bearings--usually
simple tasks that do not involve major maintenance.   Inherent is the
assumption that workers will come to feel pride in their responsibility.
106/  Hall (1981) relates that "when the Japanese explain in detail how they
achieved their big increases in productivity, the biggest war stories from
the plant floor involved hard fought battles to reduce set-up times on a
piece of equipment at first regarded as an insurmountable obstacle.
Accounts of these battles detail changing the design of bolts, and the fit
of pieces together on the machine, and the building of special tools to
speed changeover."



- 64 -
Figure 3.1 Example of the Applications of the
"Five Whys" Principle of Machine Maintenance
Problem: Malfunction of digital controller for NC machine
Why: Defective printed circuit board
Why: Lack of cooling
Why: Lack of air
Why: Lack of pressure
Why: Dust on filter
Solution: Clean filter every month
Source: Suzaki (1987), p. 116
Figure 3.2 How to Prevent Machine Breakdowns and Trouble
Maintain the normal machine conditions (e inspection
to cleaning
(e tightening bolts
(- maintain correct
operating procedu-es
Detect abnormal conditions as early    to inspection/use of
as possible                                 operator's five senses
(o inspection/use of
diagnostic equipment
4o by maintenance crews
Develop and implement counter-           (e ask why, five times
L.measures to regain the normal            (o develop new standard
machine conditions
Source: Suzaki (1987), p. 114



- 65 -
(ii)    Maintaining production line flexibility means that machines need
to be light-weight, in order to be moved easily, with fixed
storage racks and conveyor systems eliminated or minimized.
(iii)    In keeping with the JIT principle of producing only as much as
needed, machines are operated at low speeds or, if there is no
immediate demand, not at all (anathema to the Western mass
production tradition of pushing out the product)--thus extending
machine life, reducing breakdowns, and preventing inventory
buildup.
(iv)    Problem-solving under the pursuit of quality implies a continuous
process of machine improvement and adaptation that is likely to be
engineering intensive. The Japanese have a word, kaizen, which
connotes a philosophy distinct from the search for high-tech
quantum leaps in productivity gains and which stresses the
historical importance of incremental productivity gains through
organized learning.
3.21        Kaizen and the new practices also imply different criteria
regarding capacity expansion and mechanization. Stemming from gains from
flexibility, the new approach dictates that capacity expansion should be
pursued incrementally by adding general or special purpose low-volume
machines. Small volume, modular machines can be added as sales grow. They
are cheaper, incur lower downtime costs, and provide the crucial flexibility.
3.22        Automation versus improvement.  Under the new practices, the
rationale for scrapping conventional machinery in favor of more automated
technology also needs reexamination. Equipment in most factories has been
badly neglected and its capacities poorly exploited. Thus, there are so many
wrongs to be corrected that (a) the potential for gains in productivity and
quality is enormous and sustainable over a considerable period, even in the
face of automating competitors; and (b) if more automated and expensive
equipment is acquired without reforming bad organizational practices, the
equipment will fail to overcome the original problems of poor quality and low
productivity.
3.23        Instead, new maintenance and quality control procedures can make
current equipment reliable in a uniform, dependable cycle with no diminution
of quality--and a likely substantial rise in productivity. In addition, the
introduction of work standardization and pre-automation improvements can
standardize and shorten reach and flow distances and make tools, parts, design
packages, racks and fixtures easily accessible. These are all aspects of



- 66 -
operations that contribute to improved efficiency without resorting to
automation. !5
Managing the Production Process
3.24        Production line reorganization and materials movement.  The major
physical change under the new approaches involves a shift from a process-
oriented to a product-oriented production layout. Production lines in
engineering firms are typically organized as job shops, with equipment
arranged by process. All the milling machines are together as are all lathes
and presses. Work usually goes.back and forth to allow completion of dif-
ferent operations. Such clustering, illustrated in Figure 3.3, creates many
conditions inimical to organizational flexibility.L2!/
3.25        The ideal under the new approaches is to organize by flows rather
than by processes, with sets of closely related activities grouped into
product families. Workstations are close, so that work can pass between
workers without intervention. Various configurations of processes organized
by flow are depicted in the lower part of Figure 3.3. Figure 3.4 shows how
the flow principle was applied in practice in a Rockwell manufacturing plant
in the U.S.
1Q2/ Chapter II pointed out that many Western firms have either eliminated or
greatly reduced their requirements for automation technology after pursuing
practices along these lines. The same situation almost certainly applies
in many Third World firms, where the belief is that automation is the only
means for survival. Indeed, as discussed in the final chapter, the logic
of the new machine practices raises fascinating questions about the way
Third World governments and firms, as well as development economists,
planners, and international agencies, approach issues of competitive
strategy. This is not only in response to automation in the industrialized
countries but also in relation to plant operation, worker training, R&D,
techniques, supplier selection, technology transfer, project design, etc.
lOBi Schonberger (1986) emphasizes in particular that perhaps the main drawback
of clustering is that it tends to create a "gang" or "us and them"
mentality among workers, which translates into a tendency to blame others
for problems and a reluctance to take action to overcome the assumption
that the responsibility lies elsewhere. When work stations are distant
from suppliers, coordination of workflow is difficult and impedes rapid
communication between workstations, which is crucial to flexibility.
Product flow is long and involves much handling, thus increasing chances
for damage; inventories build up, thereby hiding defects and evidence of
their causes; inventory storage crowds the work area, as do the carts and
other implements required to move it from one site to another.



- 67 -
Figure 3.3 Examples of Clustered Production and Flow Lines
1. Clustered, jumbled                             a0ao0aa 0  0^
e Clusters of generic workstations         oa:aOa      ^
* No attempts to organize by product        Oo0 0%
flows                                    0 00
* No easily identifiable flow path,
or organization contrary to flow path
2. Clustered, flow-line                         OAaa,4^o   Oa6Q_^^^
* Clusters of generic workstations        X XGG £AA   o
* Organization by product flow            *o
3. Cellular
* Unlike workstations grouped into cell to         Q9 qp pQ Qp qp
produce a product family                              &4 Q
. Only one workstation of a type, except           * I   't J   g
where more needed for balance
* Cell-to-cell organization by product flow
4. Unitary machine or assembly station                      A A  . 0.1
* Complex module/product made at one machine,
station, transfer line
5. Dedicated flow-line
- Unlike-workstations grouped into flow-line
e Only one workstation of a type, except where
more needed for balance
* Organized by flow of a single product or a
regular mix of products
6. Combined
* Mixture of any of the above for a product or
group of products
Source: Schonberger (1986), p. 105



. 68 -
Figure 3.4 Reorganization of Production Line and Gains
Achieved, U.S. Components Firm
Before                                  After
.t                                    .~~~~~~~~~~~~~~~~~~~~~~~~~~~I
Summary of Improvements
Before.                 After            Improvement
Floor space 1148 sq ft                  720 sq ft           37%
In-process
inventory    4000 pcs                     10 pcs           99%
No. of
operators        2.7                       2                25%
Output per
person           168                    217                 23%
Scrap            100  (indexed)           28  (indexed)  74%
Source: Suzaki (1987), p. 81



- 69 -
3.26        Under flow-line arrangements, a workstation is close to both the
source and destination of its product. The worker becomes part of a group
with responsibility to set and meet objectives (for which group members may be
rewarded) rather than just one of many workers performing repetitious tasks.
If problems arise that impede work flow within the group or between groups,
communication occurs quickly, followed by immediate efforts to solve the
problems--or else production soon stops. In effect, "centers of respon-
sibility" clear up problems between the shopfloor and other activities such as*
procurement, shipping, and design because all operations are physically or
operationally closer together.
3.27        Lot and inventory reduction.  Two other essential tools for
bringing the production process on-line are lot reduction and inventory
reduction. Although the ideal of lot sizes of one and "zero inventory" is
never really achieved, the aim is to move towards these via incremental
improvements. At each step, problems are solved as they are revealed so that
further reductions in lot size and inventory can take place. Both activities
have implications for action at the workstation, elsewhere in the production
process, and in other parts of the firm.
(a) Instituting lot size reduction is a straightforward manage-
ment decision. However, to achieve it demands a quick-change capability that
has many implications for workforce training, machine maintenance, and
engineering functions. At the same time, this flexibility and changeover
capability give product designers and marketing people new opportunities and
freedom of action.
(b) Reduction of workstation and buffer inventories is a manage-
ment decision that also requires comprehensive attention to quality, a major
change in relations with suppliers, including improvement in their delivery,
and a system for managing the in-house delivery of parts on demand. The most
widely discussed and probably least understood of the needed steps to facilit-
ate this is the so-called kanban system of inventory control.
3.28        Kanban is simply a method for parts ordering.  A card or chit (the
kanban) travels with a container of parts or sub-assemblies. When the
container is empty and the last part/sub-assembly is used, the card returns to
the part/sub-assembly supplier where it becomes a document authorizing the
production and release of another batch of parts/sub-assemblies. The objec-
tive in managing a kanban system is to minimize the number of containers/cards
in transit, those waiting to be worked on, and those waiting to be filled. /
3.29        However, var'.ous steps need to be taken before a company can
operate kanban effectively. These involve cutting lot sizes through reduced
109/ Kanban can be operated on a paper basis and does not need computers.
Toyota's Kamigo engine plant, which produces 250,000 units a year and is
widely reputed to be the most efficient plant in the world, operates
entirely on a paper-based kanban system. American plants, which are three
times larger, with six times more material inventory, use six times more
labor per engine than at Kamigo.



- 70 -
engineering set-up time, streamlining plant configuration, and introducing a
sound TQC program to minimize quality-related delays, stoppage, rework, and
lead times. Although kanban systems only work under JIT conditions, JIT can
work on the basis of other pull-through inventory control techniques. There
are many variations of kanban systems suitable for different types of firms
and products.11/ Most important, contrary to the common impression given in
the literature, kanban is only one aspect of the new practices. For most
companies it is the last part of the production process, put in place after
all other steps have been taken.
3.30        When combined with lot and inventory reduction, workflow reorgan-
ization generates numerous benefits. Based on experiences of small- to
medium-sized U.S. engineering firms, three of these appear potentially
significant from the perspective of developing countries.
3.31        Line reorganization often reduces the physical space required for
production and can generate substantial savings in unit fixed capital costs,
particularly when new plants use JIT principles.l!_/ For instance, Fort Motor
Company wanted to build a new power train plant in the early 1980s. It
received bids from American and Japanese companies for the same plant, same
product, same output level. The Japanese bid pegged construction costs at
US$100 million compared to US$300 million for the U.S. design, with total
space at 300,000 sq. ft. compared to 900,000 sq. ft. The differences were due
to the use of simpler, cheaper machines and the virtual absence of material
handling devices (carts, conveyor belts) and inventory storage. Savings on a
similar scale would be attractive for planners and financiers of investment
projects in developing countries.
1_0/  For instance,  some assembly lines can operate with kanban "squares."
These are squares marked out on the work tables at each assembly point.
After a line worker has finished his assembly task, he places the workpiece
in the kanban square at the adjacent station and, in turn, expects to find
another workpiece in his own square, just passed on from the previous
station. If the square is empty when he is ready to assemble, or if he
cannot keep the next kanban square filled as needed, this is a sign that
there are problems.
Iai   Inductoheat, a typical mid-sized job-shop producer of heating equipment,
decided to adopt some JIT practices in 1982.  Workplace reorganization
entailed five points: remove everything unnecessary, not just inventory;
create a place for everything; and pursue cleanliness, discipline, and
participation. One major aim was visibility, so that non-performance would
be obvious in the normal work pattern. Another was to increase frequency
of part delivery.   Simplifying work practices along these lines in the
first month eliminated five trailer loads of excess material, equipment,
tooling, and trash.  Sales in 1986 were 3.6 times those in 1981; return
on equity at 28% compared to -2% in 1982; inventory 50% less than five
years ago; workturns up 400%; floor space halved, and employment reduced
by 20% (Williams, 1986).



- 71 -
3.32        The second benefit of workflow reorganization relates to the
simplification of 2roduction Rlanning activities made possible by line
reorganization. In large multiproduct firms, the planning process for
clustered production can be complicated since many different parts, opera-
tions, and production routes are involved. This planning effort is informa-
tion-intensive, requiring either a lot of paper or sophisticated computer-
based systems, and a great many indirect personnel who monitor work flow in
the factory (schedulers, dispatchers, stock takers, expediters, inventory
clerks, and specialist data processors).
3.33        The new practices eliminate much of the need for this complicated
planning apparatus (and its supporting hardware, software, and staff)--
including elaborate and expensive computer-based plar.ning systems. Simplifi-
cation of the production planning process via reorganization could be particu-
larly important for developing country enterprises where X-inefficiencies
account for low productivity.
3.34        A final benefit significant for developing countries is that
continually reducing lot size and overcoming other obstacles to the smooth
flow of production results in considerable reductions in lead time. Lead
times are included in all aspects of production: production set-up, filling
orders, responding to customer requests for quotes and specification, produc-
ing designs and proto-types, and in getting new products to market.iH/
3.35        As a result of marketplace demands in recent years for quick
turnaround and rapid response to shifts in demand, lead-time factors have
become critical determinants of competitive success. Consequently, lead-time
reduction via the new practices is seen as one of the key dimensions of
competitiveness where Western manufactureis can effectively fight back against
both their Japanese and developing country competitors. This aspect has
direct implications for developing countries, which we take up in the last
chapter.
3.36        Reorganizing job-shop production along flow lines (and introducing
lot-size and inventory reductions) is a complicated process that can be done
only in stages and often over a long period. Achieving flow-line production
requires the gradual introduction of a variety of other practices that affect
the management of the workforce, relations with suppliers, quality of goods,
112/ Plants can reduce lead times only by solving problems that cause delays.
Such problems run the gamut of plant/firm operations: order-entry delays
and errors, wrong specifications, long setup times and large lots, high
defect counts, machines that break down, operators who are not well
trained, supervisors who do not coordinate schedules, suppliers that are
not dependable, long waits for inspectors or repair people, long transport
distances, multiple handling steps, and stock record inaccuracies. All
of these obstacles are likely to receive intense focus from the new
practices.



- 72 -
as well as operation of machinery. 2/ The scale of the changes required can
be daunting to managers of sophisticted multi-product, multi-stage plants in
the industrialized countries. Because of this, it is unexpected but not
surprising that practitioners universally acknowledge that reorganization is
perhaps suited to light assembly and job-shop engineering firms (both of which
are found in great numbers in developing countries) than to the mass produc-
tion of complicated final products such as automobiles. !L./
3.37        Eliminating variability via simole measurement and observation.
After line reorganization, the business of process management shifts to the
system itself and the identification and correction of problems of variabil-
ity. The aim is to move all systems toward constant performance within
acceptable standards. Identifying performance variations and quality defects
and tracking down their causes are not accomplished by ad hoc observation and
hunches. Hence, great emphasis is placed on the importance of rigorously
collecting, analyzing, and using data on the production process to identify
and solve quality problems.iH/
3.38        Various techniques for the collection and analysis of reliable,
accurate information on system performance have been developed and are
described in detail in a number of texts. Among widely used quality-control
techniques are a set of process control measures involving both quantitative
and qualitative data. The best known of these is Statistical Process Control
(SPC), which is a simple technique for measuring deviation from the mean for
different quality parameters. There are other techniques, shown graphically
in Figure 3.5 and described at length in all the practitioners' texts.11
.4 
113/ Practitioners propose a number of techniques for accommodating the
principles to the reality of firm, products, and layout-specific
conditions, and for making the transition. See Schonberger (1982).
114/ See Schonberger (1986), Chapter 6.
115/ From a training manual put out by Komatsu, Ltd. of Japan, a company that
nearly singlehandedly destroyed American tractor companies, such as
Caterpillar and is a major supplier of tractors to developing countries:
"The first step in quality control is to judge and act on the basis of
facts. Facts are data such as length, time, fraction defective and sales
amounts.  Views not backed by data are more likely to include personal
opinions, exaggeration and mistaken impressions. Data volume has nothing
to do with the accuracy of judgment. Data without context, or incorrect
data, are not only invalid but sometimes harmful...   It is necessary to
know the nature of that data and that proper data be picked as well."
(cited in Walton, 1986).
116,/  See also Crosby (1979); Feigenbaum (1961); Ishikawa (1972); and Shingo
(1985).



- 73 -
Figure 3.5 QualLty Control Charts Designed for Use on Shop Floor
-               ~~~~~~~Caus..4n4451.a
Seven
Helpful
Charts           _
Flow Chans     PMate Outs
_XLwL
Run (Tmuu ChAnsMssu 
I Cbt S_O  I
X~ it
Source: Walton (1986), p. 98
3.39        One significant characteristic of these techniques is that they
are intended primarily for use by the production workers who are responsible
for collecting the data and information needed to prepare them. There are
some, of course, such as run and control charts that require simple statis-
tical training to prepare. However, most of the expertise in using these
techniques to improve quality and performance lies with workers and line
supervisors.E/
3.40        These techniques, combined with the worker's understanding and
commitment to quality (for which he should also be paid), transform the
production workforce into quality-control experts. Enormous improvements in
quality and reductions in quality-related costs translate directly into
improved market performance and greater profitability. However, these
techniques are tools only, not the source of quality. They will be of little
use unless management creates conditions in which quality is seen by all as
the key to success.
fl7/ A common theme among practitioners is that quality control charts should
be on full display in the relevant work area. The objectives of making
shop performance data "public" are threefold: (a) everyone becomes aware
of the quality- control effort and what needs to be done to improve
performance; (b) valuable information is available so that everyone can
learn from it; and (c) customers may request to see or inspect the evidence
of a firm's quality performance.



- 74 -
Managing Supplier Relations
3.41        The new approach to organization calls for profound changes in the
relationship between suppliers and their customers.12/  In the old system,
suppliers and customers formed an arms-length or adversarial relationship.
The main criterion of supplier selection was price. Quality, delivery time,
and technology were secondary factors. Buyers multi-sourced component
purchases to provoke competition, which would drive down the price and reduce
their risks of supply disruption. Contracts were short-term and frequently
shifted from supplier to supplier because of fractional price reductions.
There was n6 exchange of design or production information--on the contrary,
outright hostility occurred if either party expressed the desire for such
information. Finally, suppliers' plants were located at a distance from
customers because suppliers were not tied to one customer.
3.42        Currently,  all aspects of the above situation are shanging.  In
an environment in which JIT organization ultimately will become paramount, and
with growing complexity in technology and rapid product change, the old buyer-
supplier relationship is becoming anachronistic. Plants have to be better
located to allow JIT delivery. Suppliers must be highly flexible. Design
secrets have to be shared, and buyers and suppliers must develop new modes of
cooperation. Most important of all, quality and reliability have become the
key determinants of supplier selection. Achieving these changes is not easy,
with much mistrust to be overcome on both sides.
3.43        Suppliers are having these new buyer requirements imposed upon
them while also having to introduce most of the intra-firm organizational
changes (discussed above) to meet the new performance criteria. As with
intra-firm practices, the new supplier relations are characterized by a number
of features. Four of these are particularly important:
3.44        (i)  Ouality.  The most significant change is the growing impor-
tance being given by buyers to component quality, i.e., its perfection and
consistency. Figure 3.6, based on a survey of the U.S. auto component
industry, shows clearly how quality has now become the single most important
determinant for supplier selection, while consideration of price has declined
considerably. The same phenomenon is occurring in other engineering product
groups.T
118/ Hoffman and Kaplinsky (1988).
119/ See Morgan (1986).



- 75 -
Figure 3.6 Trends in Perceived Assembler Criteria for
Selecting Suppliers
(1 - noc important; 5 - extremely Lmportant)
,rsm      _ ;bb            .SPis
_.           * _o*401
'S~~~~~Od
35
es. *~                  a .aea suema_ S
tem~~~~~~' -4
ae~ ~ ~~~~s ,         i i "  J   l ^
ft r   dUcU sme_arom
I ~ ~ ~ ~   ~  ~   II aAA see_________
Source: Hoffman and Kaplinsky (1988)
3.45        A key dimension of the shift to quality is that contract (i.e.,
entry) standards are now much higher: if componenc quality is not up to the
stricter buyer standards, suppliers are not allowed to bid on the contract.
Another dimension is that the quality of the product must not only remain
consistently high but also improve over time. In Japan, providing quality is
seen as the joint responsibility of the buyer and supplier. Buyers-typically
provide technical support to suppliers to upgrade their quality; this has
proved to be enormously important in he'ping suppliers make improvements, and
buyers frequently visit suppliers' plants to make sure quality procedures are
followed. This close interaction has been more difficult for U.S. and
European suppliers to accept but is now becoming common.
3.46        (ii)  New contractual relations.  The contrast between Japanese
and U.S./European buyer-supplier relations is probably strongest in the area
of contractual relations. In Japan, contracts are agreed to and implemented
in an atmosphere of mutual trust. In the U.S. and Europe, contractual
relations are habitually conflict laden, although fundamental change is
beginning to occur, principally in two areas: a shift to "single sourcing"
whereby a buyer uses only one supplier to provide a component; and the
replacement of single-year contracts by multi-year contracts. These contracts
are for much larger quantities than previously ordered, giving the winning
supplier the advantages of scale and stability that frequently have been
absent. The buyer gains in quality costs and reliability. The new contracts
require a whole new relationship between the two sides, featuring much closer
interaction on design and in production scheduling.



* 76 -
3.47         (iii)  Changes in the design relationshi2.  Another area where
change is occurring is the suppliers' role in the design and development
process. Traditionally, buyers provided the design work and specifications
for suppliers to bid and build on. This is now changing. First, with new
emphasis on the quality of the component, the quality of the specification
must increase considerably as well. Because variability in component quality
can no longer be tolerated, the precision of specifications is receiving a
great deal of attention.
3.48        Second, suppliers now are expected to take a much larger role in
the design process for compponents and to do much of the design work previously
carried out and provided by the buyers. Third, the nature of components is
changing. Instead of their being provided individually to the buyer for
assembly into sub-assemblies and final products, components are becoming part
of the modularization of component systems. This means, for instance, that
cars are designed in terms of systems--wheels and brakes, seats, the interior
"cockpit," door/window systems--which suppliers are expected to design, build,
and provide pre-assembled to the buyer. These modular components are ready to
be "plugged" into the car on an "as-needed' basis. Needless to say, new
technologies, particular.ly electronics and new materials, are important
here.L"/ The implications of these trends for developing countries as input
suppliers are profound and are discussed in the last chapter.
3.49        (iv)  JIT delivery and location.  The geographical demands of the
new system constitute perhaps the most visible change in supplier relations.
JIT delivery requires physical proximity between supplier and assembler.
Japanese performance in the area of JIT delivery is the ideal. / The JIT
delivery system requires organizational change first and is easily managed on
paper, as Toyota does, but once established, the use of computers is a logical
extension. The move to JIT delivery and proximate location of suppliers to
assemblers is not as far advanced in the U.S. and Europe, yet the trend is
unmistakable: as with quality, the ability of suppliers to deliver JIT is
becoming a key criterion for supplier selection. Table 3.3 shows that U.S.
auto component suppliers are moving towards JIT and proximate location.
Although less empirical data is available on developments in other industries,
the trend is expected to spread.1/
3.50        All four areas reviewed have important implications for future LDC
prospects. We turn to these in the final chapter.
120/ Hoffman and Kaplinsky (1987).
121i Typically in a Japanese plant, a large share of parts will be delivered
at least on a daily basis and sometimes more frequently.
122, Morgan (1986) and Hoffman and Kaplinsky (1987).



- 77 -
CHAPTER IV
CONCLWUSION
Summarv of the Issues
4.01        The technological and organizational transformation of the
engineering sector has been the primary focus of the preceding chapters. In
Chapter I, we argued that the focus of the innovation process was irrevocably
shifting from "islands" of automation towards the systemic integration of the
production, design and control spheres. Ultimately this means highly
integrated CIM facilities will become commonplace.
4.02        For the foreseeable future, the technological locus of the
competitive struggle in the engineering sector lies in the development and of
integrated, flexible systems within individual spheres and particularly within
the production sphere. The same movement towards greater integration is
occurring in the design and management spheres. The route towards their
integration within the firm is now well defined--although difficult to pursue.
4.03        Three aspects of this shift towards flexible integration are
important. First, flexible systems enjoy considerable economic advantages
over lines employing stand-alone automation technologies--a gap thac opens
even wider against production relying on conventional metalworking technology.
Second, the key to efficient use of flexible systems lies in a firm's own
flexibility in the management and organization of production. The effective
use of these systems demands the functional and vertical integration of
previously separate responsibilities, coupled with a sharp increase in the en-
gineering intensity of design and production. Simply put, integrated tech-
nologies work most effectively in integrated organizations featuring a highly
technically competent workforce.
. .
4.04        Third, and most important, a sizable share of the gains from
investment in flexible technology comes from organizational change. This
suggests that organizational innovation is in fact separable from technical
change--and frequently eliminates the need for more complex systems of
hardware. Calls for fundamental change in production organization and
managerial attitudes have become a central theme in much of the analytical and
prescriptive literature on industrial development in the advanced economies.
4.05        This growing emphasis on the importance of organizational innova-
tion separate and independent from the previously dominant preoccupation with
embodied technological change, is the most significant, albeit counterintui-
tive, finding of this research. The analysis and discussion in Chapters II
and III explored these developments further to illuminate the scale and
pattern of diffusion of organizational innovation at the sectoral level and
the nature of the changes occurring at firm level.



- 78 -
4.06        This review demonstrated that the new organizational pract4.ces
have moved outward from the motor vehicles sector in Japan artd across national
and sectoral bourndaries. After a period of-skepticism and misunderstanding,
management in the U.S. and Europe is increasingly aware that the introduction
of the new practices, independent of any technology-related activities, can
generate substantial and sustainable gains in productivity, quality, market
share and profitability.
4.07        Awareness of the need for change, and understanding of how to go
about it, are no longer confined to a few leading-edge fSrms, business school
professors, and messianic management consultants. The message is echoing in
the halls of government, in boardrooms, in the business schools and in the
media. Consequently, we feel the degree of attention paid to organizational
issues in this paper is justified, because oi their intrinsic importance and
because of the role they play as precursor and facilitator for the more widely
recognized changes coming from industrial application of information technol-
ogy.
4.08        A New Industrial Paradigm?  One of the broad hypotheses underlying
our analysis is tha.: flexible automation technology and organizational
innovations are coalescing into a new best-practice manufacturing system now
diffusing throughout industry. We would argue that in the course of this
diffusion process, the determinants of international competitiveness are being
changed and the rules governing the hierarchy of economies and industries are
being rewritten. The engineering sector, in its widest sense, is acting as
midwife to the birth and early development of what many other observers have
called a new industrial paradigm.
4.09        Obviously the "evidence" presented in Chapters I and II does not
constitute a sufficiently comprehensive and statistically robust sample to
"prove" the existence and diffusion of this new paradigm within the industria-
lized countries. Nor are the technological and organizational elements of
this new best-practice manufacturing system likely to be fixed or immutable.
Situation-specific conditions, the retarding weight of costly sunk investment
in existing technology, decades of confrontational labor-management and
supplier-producer relatiors, and the "tunnel vision" of U.S. and European
managers inevitably will interact with che principles of each element. The
end result will vary widely--from outright rejection to differing adaptations,
with different elements spreading waevenly and piecemeal across firms, sectors
and countries. .
Assessing the lmnlzcations for Developing Countries
4.10        These qualifications suggest that any attempt to predict the
speed, pattern and extent of adoption of this new best-practice system across
countries and sectors would be of little lasting value. Nevertheless, we
believe the documented processes of technological and organizational change
are powerfully indicative of the dominant trends now manifest in the global
manufacturing sector. These changes broadly define the parameters that will
1223_/  Piore and Sable (1984) argue this point well.



- 79 -
characterize the structure of manufacturing enterprises in engineering and
across many sectors. And since these changes will likely be major deter-
minants of international and national competitiveness, firms and policymakers
in the developing countries will have to confront the implications.
4.11        This is not an easy task since it is still difficult to define
precisely what the implications wi'll be. Although some aspects of IT-related
developments are affecting a small band of countries, the implications of
flexible mAanufacturing technologies and the new practices are at best only
generally discernible for most economies. This degree of uncertainty about
the future, coupled with an inadequate knowledge base, suggests rather
mundanely that the initial response of policymakers should be to become well-
informed about international trends and aware of the constraints inhibiting
their manoeuvering in particular situations.=/
4.12        Unfortunately, the available empirical literature concerned with
IT and developing countries is of limited use in evaluating the implications
of the new paradigm for the Third World and the actions open to policymakers,
for two reasons. First, country coverage in relation to the engineering
sector has been confined largely to trade-related questions of immediate
importance to a handful of countries. Second, the focus of this literature
has been, with one or two recent exceptions, almost entirely on developments
arising from the diffusion of stand-alone automation techniques and not on the
forms of technological and organizational change reviewed in this paper.
4.13          However, it is possible to consider the impact and implications
for the Third World at two levels: first, as issues arising from the manner
in which the technological and organizational developments may affect develop-
ing countries' capacity to respond both internationally and domestically.
Second, by focusing more narrowly on the issues that concerned us in Chapters
II and III--the problems and possibilities for the introduction of the new
organizational practices within the Third World.
ImRlications for the International and Nationil Competitive Context
4.14        The uneven diffusion of flexible manufacturing technology and
organizational change will remain for many years an important feature of
international trade and competition. A major feature of the diffusion process
will be the continued rapid spread of stand-alone automation technologies in
the industrialized countries and their less rapid spread in the developing
countries. As noted previously, there was an early (and continuing) warning
in the literature that differing rates of North-South diffusion would under-
mine the low-wage advantage of Third World manufacturer/exporters and generate
12W   Though lack of comprehensive up-to-date knowledge is a problem confronting
Third World policymakers in all areas, it is particularly problematic in
relation to those fields where information technology is having an impact-
-both because developments in the industrialized countries are moving
quickly and because so much information reaching developing countries on
these topics is not accurate.



- 80 -
unfavorable employment and welfare effects in developing countries. /  Given
the prior extensive treatment of these issues elsewhere, we wish to make only
three sets of points on this topic.
4.15        First, the most recent reviews of developing country use of stand-
alone automation suggest that at least until the mid-1980s, even the leading-
edge countries, the NICs, were still some way behind the advanced industrial-
ized countries. This comes out clearly in Jacobsson and Edquist's (1988)
comparison of relative stocks and rates of diffusion between the NICs and the
OECD countries.
4.16        In 1985 Korea had the most NC tools of all developing countries,
with 2,700 units, while Japan alone had 120,000; by 1985 the NICs had ap-
proximately 400 rviAots in use while there were more than 100,000 in the OECD
countries. In terms of CAD use, in 1985 the U.S., West Germany, Japan, the
U.K. and Sweden had more than 90,000 "seats" while the NICs had 2,100. By
introducing a scale-related factor into these comparisons (the density of use
per engineering sector employee), the figures show that the OECD countries had
8.5 times as many NCMTs as the NICs; 8.3 times as many CAD systems; and 43
times as many robots.
4.17        Comparing rates of diffusion is more difficult. It appears that as
of the mid-1980s the rate of diffusion of NCMTs in the NICs was not sufficient
to "catch up" with the industrialized countries, primarily because the share
of new investment allocated to these techniques in much lower in the NICs--7%-
23% as against 40%-62% for the OECD countries. Available information on the
rates of diffusion for robots and CAD systems is confusing: although most are
well below OECD levels, a few countries such as Singapore show high rates of
use, possibly on a par with developed countries.
4.18        Seg nd, despite this apparent widening of the technology gap, and
a good deal of well informed speculation L2±/ there is little hard evidence
that the use of stand-alone automation in the engineering sector in the
advanced countries has resulted in a substantial loss of export markets for
developing countries, or in the relocation of offshore facilities established
by Western producers. This may change as the level of technology integration
L2../ The literature unfortunately has failed to reconcile the rather contradic-
tory nature of these two positions. Presumably an economy could suffer
both the loss of competitive advantage due to non-adoption of new
technologies in one sector and the loss of jobs due to widespread adoption
in anothet.  However, both these arguments are so often stated in the
literature by different analysts in such uncompromisingly, all -encompassing
terms (reminiscent of the dependency arguments) that any other outcome is
ruled out!
126/ Jacobsson and Edquist (1988) make the most rigorous attempt in the
literature to assess this question by reviewing the relative rates of
penetration of automation techniques into sectors where NIC exports are
concentrated, and show that the threat to the NICs' competitive advantage,
though real, still lies in the distant future.



- 81 -
rises among Western users, but we suspect the initial fears were overstated
and based on an assumption that stand-alone automation techniques were much
more competitive against earlier electromecihanical versions and low wages in
developing countries than was actually the case.=/
4.19        Indeed, the most significant threat to competitive advantage in
the engineering sector for the less developed countries has come not from
process innovation but from product innovation in the machine tool sector.
The growing strength of the NICs as exporters of simple tools such as lathes
is being challenged severely by the growing design intensity of the product
(due primarily to use of CNC units), associated increases in minimum scale
economies, and the rise of Japan as a major competitor.=/  To maintain
export markets in developed, and increasingly in developing countries, a
switch will have to be made to the design and production of CNC tools. The
same logic probably applies to other machinery segments of the engineering
sector where the new product technology is making inroads, but this assumption
needs to be examined empirically.
4.20        For the many countries still far removed from the technological
frontier in machine building and exporting, the penetration of NC, CNC and
programmable controllers into machine design poses major new technological
obstacles to their (admittedly remote) prospects for entering the interna-
tional market. However, empirical research has shown that some NIC firms are
succeeding in making the transition to export of microelectronic-controlled
machines--and that many more firms are producing these for the domestic
market. .
4.21        The NIC experience suggests that government policy is extremely
important in supporting firms' learnirtg process and attaining minimum scale
economies while assimilating the new product technology with an eye towards
exporting.20/
4.22        More fundamentally, NIC firms exporting NC machines were already
competent machine builders before mastering the new product technology.
Without that prior experience, the task would have been inestimably more
difficult. This underlines the obvious, but often ignored, point that the
export aspirations of non-NIC countries in the machinery (and broader en-
127/ See Hoffman and Rush (1980), Jacobsson and Sigurdson (1983), Rada (1982)
and Kaplinsky (1982), for the early arguments; the articles and references
contained in Hoffman (1985) for a review of the first round of empirical
evidence and James (1987) for the most comprehensive listing of references
on this topic and a rather sanguine review of the literature.
128/ See Jacobsson (1985).
129/ Kim (1986) reports that 110 large and small firms are engaged in machine
tool production in Korea. Production of NC tools has increased from only
$200,000 in 1977 to over $38 million in 1985.
120/ Jacobsson (1985) and Franzman (1984, 1986).



- 82 -
gineering) sectors are constrained on the technical side not by their lack of
IT skills but because of the pervasive weakness in their basic mechanical
engineering and machine design and building_capabilities--coupled with
131
structural problems that are often policy-induced.=/  If such countries can
overcome these obstacles to the point where the export of conventional
machines is a viable proposition, then the success of NIC producers suggests
that moving into the microelectronics era may not be as great a hurdle for the
less advanced countries as most observers assume.
4.23        Our third point is a simple one and follows this line of argument
but in relation to the largely pessimistic view in the literature about the
prospects and impact of the use of stand-alone automation technologies in
developing countries. All the available evidence (relating primarily to
limited NIC use of NCMI) indicates that if developing country firms are
already efficient users of conventional techniques, they will face few
insurmountable technical or skill-related problems in using NCMTs or other
stand-alone automation techniques--though capital costs, scale requirements,
available skills, protected markets, and the availability of supplier support
will act as constraints. 12/
131l/ This is hardly a new argument in the technology and development literature,
but it does seem to have been largely forgotten by those concerned with
the impact of automation on the Third World.
132/  For the NICS and nearly-NICS, these problems are not insignificant, but
most are amenable to government intervention. A variety of mechanisms for
overcoming knowledge and skill constraints are available. In general, the
most important in the longer term are attempts to overcome the skill
constraint. Universities and technical institutions, as well as in-firm
training programs, must reorient their curricula away from conventional
skills and towards skills relevant to the use of the new techniques--NC
tool operation, maintenance and programming; and for engineers, the use
of CAD systems in basic and detailed design, etc.
The knowledge constraint can be attacked more directly by providing
potential information about the uses and benefits of the new technology
to potential users, by seminars, workshops, exhibitions, "showcase firms,"
national information centers, etc. At the same time, the local service
industry must be strengthened, particularly in software support, such as
maintenance and product upgrades, areas where foreign suppliers tend to
be weak. In relation to hardware, although import restrictions are a tool
for supporting development of local production capacity, these must be
eased in, to ensure that local users gain access to a wide array of
techniques--consistent with development of a competitive local supply
industry. See Jacobsson and Edquist (1988) for a more detailed discussion;
see lso Chudnovsky (1986) for a discussion of the experience of Argentina,
and James (1987) for a review of Third World use of the new technology.
See Chudnovsky (1986) for a discussion of the experience of Argentina, and
James (1987) for a review of Third World use of new technology.



- 83 -
4.24        The problem, of course, is that many Third World engineering firms
are not efficient users (or improvers) of conventional techniques. This
suggests that the macro and micro causes of- the inefficient use of existing
techniques must be tackled as a prerequisite to the introduction of stand-
alone automation. If this were to happen on a broad scale, the early concerns
about the use and impact of stand-alone automation take on a different
perspective: these techniques will be introduced into a dynamic context where
firm and sectoral productivity, output and employment are rising rather than
remaining static or falling, as in many countries today.
4.25        Arguments based on X-efficiency grounds often attribute perfor-
mance differences between North and South enterprises in machine-level
productivity (often surprisingly close) versus overall firm productivity
(almost always much lower in developing countries) to organizational deficien-
cies resulting from incompetent management. Recent literature on questions of
indigenous technical change and the mastery and absorption of imported
technology adopts a similar though less explicit perspective.
Problems and Possibilities for-Organizational Innovation in Developing
Countries
4.26        Before exploring these issues, we need to make one qualification
about the discussion that follows. Since there has been virtually no debate
about these issues in the literature, our observations on the prospects for
introducing organiza'. onal change in developing countries are offered as a
means of provoking -rther discussion rather than as definitive conclusions.
In this regard, we should make clear we do not intend to address the specific
policy implications of organizational change, as this would be far too
premature without comprehensive analysis and evidence.
The Problems Considered
4.27        In the course of reviewing the literature and prepar.ng this
paper, it was impossible not to compile a list of reasons why introduction of
the new organizational practices would face serious obstacles in the Third
World. Two sets of 'first-order" factors seem most problematic.
4.28        (i)  Skills constraints on the capacity for change.  The first set
ib related to the issue of skill level and the overall competence of manage-
ment, engineers and production workers in Third World firms. The successful
introduction of the new practices requires both a minimum number of well-
trained people (managers, staff and line workers) compared to the total
workforce, and a minimum degree of competence in their areas of specializa-
tion--including, of course, a willingness to change attitudes and embrace the
new practices.
4.29        For most Third World firms, critical shortfalls occur in manage-
ment and engineering. All the problems Western managers, engineers and
workers face in adapting to the new practices not only would exist in a Third
World context but couid be greatly compounded by lower levels of skill and
education and by cultural, religious, and racial prejudices that would inhibit
the establishment of working relations based on cooperation and mutual trust.



- 84 -
4.30        Long-standing management problems in the Third World are widely
acknowledged in the literature as a prime cause of poor industrial perfor-
mance.1 i Such analyses undoubtedly have merit, but they fail to see that
poor performance may result from the pursuit of totally inappropriate manage-
ment practices.=/ Yet the logic of the new practices suggests that Third
World firms will face problems not just because of poor managers but also
because of their fundamentally flawed management approach. T/
4.31        Much the same points can be made about the obstacles posed by the
engineering intensity of production under the new practices. The problem
caused by severe skills shortages will be compounded by poor deployment of
engineers and by attitudinal and cultural problems limiting their usefulness.
In most developing countries, engineers are white-collar and managerial,
whereas engineers in the new practices are constantly involved in solving
problems on the shopfloor in support of production workers. /
4.32        These same issues crop up if we consider the implications for the
introduction and use by developing countries of more flexible integrated
systems such as FMS. We can deal with this issue very briefly. There are
JJ.2/ See Kubr and Wallace (1984) for a review.
l14/ A typical example of this is Gershenberg (1987), a useful research report
concerned with the role of foreign versus local firms in the spread of
managerial know-how in Kenya. Unfortunately, the analysis is carried out
without once specifying ths type of management practices involved.
12I1 Thus, even if the managers have all been sent abroad to prestigious
business schools and got straight As, problems still lie ahead since to
be effective with the new practices they will have to dislearn everything
they learned in school! Schonberger (1982) sums this conundrum up well.
"Unfortunately, masses of (developing country) people (have been)
attending Western colleges so that they may return to their countries
and help build an industrial base.. .They may learn too much about
too little and go home to become experts in risk management or
automated storage retrieval, settling into lifetime careers as they
see their Western role models doing... In this scenario, the best
minds in the country will be wasted working out complicated solutions
to narrow problems.   One wonders how the plants they build and
operate will ever be competitive in world markets." (p. 210) (FTNT
Deming quote)
16/6 The problem of the underemployment and misdirected deployment of Third
World engineers in jobs that have nothing to do with production is well-
documented (see Teitel, 1987 and Hill, 1987). This problem is compounded
by the fact that those who do actually work inside enterprises are, like
their western counterparts, loathe to get their 'fingers dirty' by getting
too intimately involved with the solution of shopfloor problems.
Schonberger, 1982 cites some examples of this.



- 85 -
grounds for arguing that flexible systems, because of their lower scale
economies, might be well suited to the limited market conditions typical of
many developing countries. Equally, one could argue that the capital costs
and greater complexity of these systems work strongly against their
introduction in all but the most advanced segments of the NICs. These
alterative positions provide a speculative standoff because there are few, if
any, truly flexible systems in place in developing countries to demonstrate
empirical support.
4.33        Much more pertinent are the points made in Chapters II and III.
There is little sense in a firm's attempting to introduce flexible systems
unless its existing processes have been thoroughly restructured according to
the principles of "simplify, combine and eliminate." A firm must already have
taken steps to introduce greater functional and vertical integration as well.
For obvious reasons, the need to make these changes as a precursor to FMS
would be much greater in most third world enterprises. Add to this the
evidence that the successful introduction of the new organizational practices
yields commensurably greater gains than the use of new technology. Taken
together, these points comprise a powerful a priori argument that developing
country enterprises should be vigorously pursuing the new paths of
organizational innovation rather than making high cost and high risk invest-
ments in new technology--whether stand-alone or integrated.
4.34        The literature has always been concerned about the social and
private costs of low levels of engineering skill and limited learning oppor-
tunities.1!/  The Japanese experience (where engineering-intensive production
leads to consistent productivity increases) shows that the opportunity costs
of this situation are, if anything, much higher than previously thought.
4.35        Two additional factors compound the obvious problems posed by the
endemic shortage (and poor quality) of production skills--for draftsmen,
welders, fitters, foundry and heat treatment workers, machine tool operators,
tool and die makers, etc. First, educational levels for workers are very low,
thus increasing the difficulty of the retraining and upgrading task. Second,
in-firm training programs are non-existent or inadequate. Severe shortages of
skilled production workers are, in effect, a creation of underdevelopment
itself and of the lack of an industrial tradition in many Third World nations.
4.36        (ii)  Constraint.s on the scoge of change.  The second set of
constraining factors limits the scope ror firm-level change. Some of these
are internal to the firm. For instance, both financial and size limitations
will reduce a firm's ability to absorb the costs and dislocations of organiza
tional change. Resource constraints inhibit raw material purchases, leading
to low output. Typically, Third World engineering firms have a limited
capacity to develop new producrs, absorb new technology (particularly product
know-how) and open up new markets. This limits their product mix, constrains
137/ See Bell, Ross-Larson and Westphal (1985); Dahlman and Westphal (1983).



- 86 -
their growth potential, makes them vulnerable to market fluctuations, and
makes the internal environment less conduciye to change. /
4.37        External constraints are probably even more severe.  The most
obvious are those imposed by market factors, such as small size and fragmented
structure. In recent years, recessions have crippled firms' capacity to
transform themselves. Other, market-related problems also can distort or
inhibit responses--the low price of labor, poor consumer response to quality
and service, the price elasticity of demand, the import intensity of existing
producers, and the various factors (legislated and otherwise) that lead
customers to prefer imports over local products.T/
4.38        Other problems are imposed by industry structure, for example, the
distorting effect of foreign firms' dominance. The poorly developed network
of subcontractors and suppliers of specialized services forces many engineer-
ing firms to perform tasks in-house that might more efficiently be subcontrac-
ted out, or in the case of special services, to do without completely.
Another problem, greatly exacerbated by foreign exchange shortages, is the
lack of uniform and reliable raw material supplies.
4.39        Of course, government macro-economic and industrial policies often
play a major role in determining the nature and existence of many of these
external constraints. In one sense, the problems from government intervention
are similar to those that inhibit firm-level innovative behavior in general,
in that they effectively restrain the disciplinary character of the market and
its ability to reward performance with increased profits and market share.
4.40        The range of policy-induced problems is wide and well known--
excessive and incompetent state involvement in production; overvalued exchange
rates and artificially low producer prices; a grossly distorted price system;
excessive import protection; capacity licensing policies; and legislated
barriers to entry and exit that encourage suboptimal levels of Coocentration
and competition.
4.41        The debate over the merits of free-versus-controlled  markets and
over the appropriate forms of government intervention (posing a choice between
a free market versus state-dc:.inated economy) misrepresents the problem faced
by many developing countries. The choice is between efficiency and inefficien-
cy, competence and incompetence, and the costs of state monopoly versus the
benefits of dynamic and innovative performance by a firm. These conditions
can exist under open market conditions or in situations of state intervention.
4.42         The twin tasks of how to reform the market to provide incentives
for firms to be efficient and to invest in organizational change (and other
types of innovative effort) and how to devise benign interventions to assist
1U/  See UNCTAD (1985) for a review.
13I9 Similar distortions arise in the case of labor markets, where the low price
of labor would inhibit management from paying higher wages to attract more
skilled staff and production workers.



- 87 -
this process are difficult policy problems, aggravated by the present conten-
tious climate regarding the issue of state intervention in the economy. These
qualifications are a reminder of the need for caution in drawing cor.^lusions
about the issues raised in this paper. However, they also tell only one side
of the story, and we need to balance our assessment of the problems against
the very significant possibilities to which we now turn our attention.
The Possibilities Considered
4.43        Four aspects of the new practices provide strong a priori grounds
for our hypothesis concerning their applicability in developing countries.
First, as the discussion in Chapters II and III has brought out, most new
practices are neither scale, product, sector nor function-specific. The
pursuit of goals such as "eliminating fear from the workplace," breaking down
barriers among staff, and instilling a concern for quality are well within the
province of management anywhere.
4.44 The same is true with most of the techniques associated with preventive
maintenance, production line reorganization, management of machines, upgrading
of worker skill and responsibility, use of shopfloor quality control measures,
new forms of supplier relations, etc. /  They are all suitable for use in
most types and scale of discrete product manufacturing. The obstacles to the
introduction of the new practices in less developed countries may well lie
elsewhere, but they are not to be found in the realm of scale and product
140/ Of course, the most well-publicized elements of the new practices like
kanban inventory control are best suited for a mass production situation
since that is what they were designed for. Gains accruing strictly to
inventory reduction in Western firms carry much less weight in Third World
conditions of low capacity utilization.
However, this observation misses the point that the greatest gains from
inventory reduction come from the "reduction of waste" arising from the
exposure (and correction) of the variety of costly production problems
related to quality, treatment of machines, poorly trained workers, the way
production is organized, etc., reviewed in Chapter III.  Such problems
surely exist in all Third World engineering plants and considerable gain
would accrue from their elimination. If there are no inventories to reduce
(or if the scale of operations is smaller), this makes the task of problem
exposure and solution and production "control" easier not more difficult!



- 88 -
characteristics or in the type of tasks carried out. ./ The problems tackled
by organizational change are faced by all developing country engineering firms
(indeed, in most less developed countries' manufacturing and service enterpr-
ises), regardless of their technological competence, size, product mix, or
market orientation.
4.45        Second, there is no mystery behind how these practices work.
Indeed there are numerous "how-to" books that describe practically how firms
should go about introducing the new practices. Moreover, most management
consulting firms provide specific advice on the practices (e.g. through
seminars). In short, the new practices are codifiable and accessible.
4.46        Third, though we cannot provide precise figures, it seems logical
that the cap!.tal cost (and foreign exchange cost) involved in introducing the
new practices will be relatively low--certainly much lower than the costs
involved in introducing new technology. Low cost is accompanied by low risk--
particularly considering that it is possible and, indeed, advisable to
introduce these practices on a step-by-step basis, only moving as fast as the
absorptive capacity of the people and sector involved.
4.47        Finally, although many types of firms face difficulties in
introducing the new practices, the heterogenous nature of the engineering
sector means that many firms in that sector stand a good chance for successful
organizational change.
4.48        Conduits for change.  Among the best candidates would be subsi-
diaries of foreign firms (Western and Japanese) that have already introduced,
or are pursuing, organizational change in their home locations. Many of these
firms are well represented in particular segments of the engineering sector in
developing countries--and would have sufficient technological, financial, and
managerial resources to pursue organizational change. A few have started to
introduce these practices in their overseas subsidiaries (such as some auto
assemblers and component firms in Brazil, India and Mexico).i2 
141/ Indeed we can draw attention here to a related point made by the
practitioners and mentioned earlier, that techniques such as production
line reorganization are in fact much better suited to lower-scale, light
assembly and job-shop operations typical of LDC plants. Schonberger (1982)
agrees:
"The JIT/TQC approach seems particularly fitting in underdeveloped
countries. They have the advantage of very low labor costs and no
problem about too much specialization and complicated solutions.
JIT/TQC is natural for the kinds of labor-intensive manufacturing
that generally begin in these countries." (p. 210)
Schonberger obviously has in mind the manufacturing sector in general, and
assembly industries in particular, in making his assessment of LDC
applicability.
2JZ  See Robinson (1988) for examples of success in the introduction of JIT
practices in U.S. auto subsidiaries in Mexico.



- 89 -
4.49        Obviously the same points, only stronger, can be made in relation
to the subsidiaries of Japanese corporations that have a strong presence
throughout Asia and increasingly in Latin America. One would expect these
firms to have tried to introduce the same management methods they use at
home.=/=/
4.50        The policy objective in relation to these firms would be to
convince them that it was in their best interest to transfer the new organiza-
tional practices to their subsidiaries abroad, along with management and
worker retraining and restructuring of supplier relations. However, to assess
whether Japanese, U.S., and European subsidiaries are viable conduits for the
transfer of the new practices, much more research is necessary.
4.51        Domestic participants in change.  Various types of domestic firms
in less developed countries also are potential candidates for organizational
change, particularly those involved in assembly-intensive activity in sectors
AV   This doesn't always hold true as we know that until two years ago, Nissan's
Mexican plant was operated very much like a traditional U.S. plant--
presumably because it was still very profitable to do this behind Mexico's
tariff walls.
I4./ The Maruti joint venture involving Suzuki in the production of cars in
India is one of these and is qualitatively discussed by Khera and Namoto
(1986). Among the problems cited are that the Japanese like to "give small
doses of technology over a prolonged period of time, whereas the Indians
are people in a hurry"; Japanese insistence on "cleanliness and order-
liness" as a key to productivity does not come naturally to the workers;
the Japanese believe in technology transfer through hands-on experience
while the Indian engineers are used to learning through study and don't
like to get their hands dirty; and finally where the Japanese insist on
time consuming and tortuous tests of component quality carried out "in the
field,  the Indian government insists blindly on meeting a strict local
content timetable without regard to quality or the capabilities of the
suppliers.
We have learned of complaints from the government, suppliers and in the
Indian press over the slow pace of technology transfer in this project and
over the "exploitative" nature of Suzuki's insistence on quality which
excludes Indian suppliers. Such fears are no doubt based on the experience
of having worked with Western TNCs who have made a fine art of maintaining
recipient dependency by spuriously enforcing quality clauses in the
technology transfer contract. Our guess is that this is not what is at
work here, but rather a lack of understanding on the part of the Indians
of the role of quality. American suppliers and the press made exactly the
same complaints about Japanese assemblers in the U.S. until the Americans
learned just how far behind they were in qualityl Clearly, there is much
to learn from further study of the Japanese technology transfer experience.



90 -
such as motor vehicles, domestic appliances, electric motors, pumps, etc.,,,/
4.52        In the NICs (and nearly NICs)--loaving aside their significanc
involvement in the motor vehicle industry--engineering and, more specifically,
capital goods firms are usually among the most competent and most resilient
enterprises in the manufacturing sector.=/  In the engineering industry,
export competitiveness is increasingly a crucial determinant of success. This
brings firms right up against the fact that flexibility, quality, and respon-
siveness are key competitive elements in the international context and
suggests that firms may not have a choice whether to introduce the new
practices--they may simply not be able to survive unless they do.
4.53        In the most poorly developed countries, the number of domestic
privately-owned firms that exhibit similar strengths to NIC firms drops off
precipitously. But some firms with acceptable preconditions for organ-
izational change can be found. UNTCTAD surveyed 37 firms with more than 100
employees each and a reasonable degree of technological and managerial
competence in a variety of engineering products in Tunisia, Thailand, Tanzania
and Peru. Similar firms operate in both the agricultural and industrial
segments of the engineering sector in many other countries.if/
4.54        Such a sample can include public sector enterprises operating in
the engineering sector (frequently in capital goods production in the poorer
countries) or purchasing significant inputs from smaller firms. These firms
operate on a large scale, have a semblance of managerial and engineering
JAI/ For instance, in more than 20 countries, the assembly of cars, commercial
vehicles and four-wheel tractors is on a significant scale, with 30% or
more local content (implying some degree of subcontractor network).
J14/ For instance, a recent study by UNCTAD (1985) on capital goods production
in developing countries, surveyed 87 domestically owned firms in India,
South Korea, brazil and China capable of producing "complex" capital goods
in the machine tools, process equipment and electrical equipment subsec-
tors. All of these exhibited characteristics that would make them suitable
participants in change efforts.   they were large (number of employees
ranged from 476 to 10,000), dominant in their sector (firms surveyed
accounted for 70% of total output in respective subsectors), well
established (their 'age' ranged from 13 to 42 years), professionally
managed and technologically competent (% of engineers/R&D staff to total
workforce ranged from a low of 10% in China through 25% _n electrical
equipment producers  in India).   Franzman (1982), Amsden (1985),  and
Jacobsson (1985) surveyed similar firms in Hong Kong, Taiwan Province and
Argentina.
147/ Child and Kaneda (1975), although dated, describe an extremely robust group
of small, medium and large sized diesel engine and component manufacturers
in Pakistan who collectively exhibited many of the preconditions for
organizational change. Thoburn (1973) and Bell and Scott-Kemmis (1980)
do the same for Malaysia and Thailand, and more references could be cited.



v 91 -
competence, exhibit a degree of sSecialization, and exercise some influence
over their network of suppliers.  J
4.55        Finally, the argument is easily made that cultural factors would
all but exclude some countries from adopting the new practices. However,
other countries might be favorable candidates precisely because of cultural
affinity for some aspects of the new practices. These countries include the
NICs and other countries in Asia, particularly China, where many firms with
the sort of favorable enterprise characteristics mentioned above are likely to
be found--but where extensive state control of the market poses many problems.
Certainly, during our own visits to Hyundai and other South Korean auto firms,
we could see clear evidence that the new practices were making an important
contribution.T/
4.56        Some anecdotal evidence on the application of the new practices in
developing countries. Fortunately, however, our case for applicability is not
entirely based on speculation and a priori reasoning since evidence is
beginning to accumulate that the new practices are being transferred to
developing countries, not just in the engineering sector but in other sectors
as well.
4.57        The information available is very anecdotal at this stage but
demonstrates the possibilities for change among different types of firms and
sectors. / For example, in Singapore, a Japanese subsidiary producing
machine tools introduced JIT and TQC in the early 1980s. As a result, worker
productivity rose by 72%. The quality of the output was sufficiently high so
that 95% of its annual sales (30 million Singapore dollars) are exported to
Japan, the U.S. and Europe. Another example from the same country relates to
a U.S. subsidiarv producing disk drives and cartridges for computers. It
introduced JIT and TQC practices in 1985. Lead-time (per day) was cut by 50%,
inventory reduced by over 50%, and product variety increased from four to 45
types.
14/  Public enterprise performance in all of these areas obviously varies.
Brazil, Mexico, India and China boast some very successful and very dynamic
enterprises (and some disasters) that could both introduce change
internally and be a progressive force for restructuring and reforming the
supplier network. In the poorer countries, there are fewer success stories
in the public sector--so few that many may be dismantled or privatized.
But where government remains committed to their survival, the pressure is
now on for them to be more efficient, and, hence, public sector managers
may be willing to pursue some of the new practices. The fact that many
of the changes could be introduced with very low levels of investment is
an additional and very important attraction that, of course, has relevance
right across the types of firms we are discussing.
149/ Kerbs (1987) argues that the same is true in U.S. subsidiaries of Korean
firms.
150Q/ Unless indicated, these examples come from unpublished research, including
our ongoing work in this area funded by the Rockefeller Foundation.



- 92 -
4.58        It is worthwhile noting that two.of the best documented examples
of the successful introduction of the new practices come from Latin America--a
region exhibiting significant cultural differences from Asia. The first
example is a large (20,000 employees/20 factories) clothing producer in
Brazil. Senior management, drawing on its knowledge of Japanese techniques
and group technology concepts, has reorganized production in each plant.
Groups of workers (14-22 members) cooperatively plan and carry out related
sets of operations on a given output level per day. Group workers master a
range of skills (with higher pay for more skills) and receive bonuses based on
group, not individual, performance. Work moves through the cutting, assembly
and finishing rooms on a strict JIT basis, and keeps station or buffer
inventories small.
4.59        Under the system, work-in-progress inventory was eliminated
completely after accounting for 30%-40% of product costs; the cycle time was
reduced from weeks to hours; productivity was increased by between 200% and
400% for different tasks; and space requirements were reduced from 80 sq. ft
to 30 sq. ft per workstation. / This is a textbook JIT operation. The
example is intriguing not just because of its Latin American origin and the
results, bus because few Western apparel producers have introduced
organizational change on this scale.
4.60        The second example !2/ concerns a Venezuelan engineering firm
that began introducing orga.izational change in the early 1980s. Productivity
has increased by 25%, along with improvements in other areas (shown below),
all achieved with virtually no capital investment.
Table 4.1: PRODUCTIVITY INCREASES IN
AN ENGINEERING FIRM
1981        1986
Scrap                          0.43%       0.23%
Customer returns
(returns/hundred units sold)  0.46%        0.11%
Labor productivity
(standard units produced
per man hour)                  6.44         8.32
Source: Hoffman (1988).
151! Gaetan (1986).
1S2/ Given by Bessant and Rush (1987).



- 93 -
4.61        There is also evidence that institutional mechanisms are being put
into place in some countries to facilitate the diffusion of knowledge of the
new practices. In Brazil, a small, private, non-profit training institute,
INAM (the National Institute of Management) now specializes in the provision
of consultancy and training services for the new practices. In Venezuela, a
program backed by the Andean Pact has been set up to provide similar services
for Venezuelan firms; it includes consultancy advice on automation tech-
nologies. Not surprisingly, the Japanese government, through its development
aid schemes, is supporting a large number of training programs aimed at
upgrading the skills of managers in developing country firms. The precise
content of these programs is not clear from the information available to us,
bnit it seems logical that Japanese m..anagement practices would form the model
for any training provided.
4.62        One cannot generalize from the above examples, of course--
particularly since there are also indications from literature and ongoing
research that problems have been encountered by foreign firms trying to
introduce the practices to their subsidiaries.!2/  On top of thls is the
reality that the development experience is littered with examples of North-
South "technology" transfer--involving both embodied and disembodied
technology--which has failed either because of the vast cultural differences
or because sufficient care was not taken to adapt new ideas to the very
different conditions that exist. Nevertheless, the fact that successful
introduction has been accomplished in some cases suggests that at the very
least the question of applicability remains an open issue.
4.63        Structural constraints revisited.  In these last comments, we
return to some points about the existence of severe internal and external
structural constraints on the introduction of organizational change in the
LDCs. Some Latin American firms have adapted to the mass-production-oriented
international division of labor in a way that suggests they are well-placed to
introduce the new organizational practices. /
I15/  Fukuda (1988)  indicates that Japanese service sector firms have had
particular problems transferring their management culture to their Asian
subsidiaries. However, the analysis is qualitative and heavily influenced
by the comparative management literature and does not yield general
conclusions.
15/  Sabel (1986).  In developing his analysis--where the main concern is in
fact with the prospects of Latin American firms introducing flexible
automation--Sabel draws in an original way on the work of Katz and others
in the well-known BID/CEPAL project.   The results of this project are
reviewed in Katz (1980 and 1982).  Related references can be found in
Dahlman and Westphal (1982); Bell, (1982); Nelson (1979); and Teitel (1979
and 1981).



- 94 -
4.64        We expand this argument to include three components.
(i)   Latin American firms originally set up along mass-production lines
have been forced to adapt to a variety of market constraints in a
way that means they are now "systematically different from
comparative firms in the advanced countries. Whereas the latter
organize operations sequentially on the model of the assembly
line...the former are organized as a collection of semi-autonomous
shops under one roof. Each shop specializes in a particular
manufacturing operation or the production of a certain family of
parts; similar machines or clusters of machines are thuz grouped
together rather than dispersed in sequences defined by the steps
required to build a particular product."155
(ii)  Although the subcontractor network in Latin America and elsewhere
is poorly developed, possibilities exist, as happened in Singapore
(and, we would add, South Korea and Taiwan), where Japanese and
American machine-tool builders created an efficient supplier
network almost from nothing.=/ New assembler-supplier relations
suggest a better way of organizing supplier relations in the less
developed countries where supplier quality is low. Final
assemblers could, in principle, take an active role in developing
and upgrading supplier capability, particularly for quality and
reliability. This means foregoing for awhile the well known
benefits of purc..asing abroad, but the returns in terms of an
upgrading of the supplier network could be considerable.
(iii)  The severe crisis in demand experienced by the Latin American
engineering sector in the early 1980s forced many of these firms
to experiment with different forms of production organization and,
1.5/ Sabel, p. 46. Thus, one sees here firms that may already have laid the
basis for organizational change and who we know from the additional work
of Dahlman, Teitel and others, have accumulated an impressive set of
technical skills suited to the sort of continuous incremental innovation
that is an essential component of the new practices.
1.56/  See Franzman (1984) for the positive side and Amsden (1977, 1985) for the
difficulties. We would add that the informal metal working sector in many
countries exhibits characteristics of flexibility resilience and great
ingenuity that are necessary attributes of supplier networks that have been
created in Japan to serve large firms in the auto and other sectors. In
developing countries, such firms, and their links to formal sector
customers, could be strengthened considerably if large firms and
governments actively sought to assist and encourage their development a
la the Japanese model instead of competing against them and trying to tax
and legislate them out of existence.



- 95 -
in short, to become more flexible in their deployment of
resources. _0
4.65        At this stage, it is of course impossible to confirm or deny
Sable's speculations and our extension of them. However, the questions he
raises prompt us to consider anew the nature of some of the constraints
mentioned earlier.
4.66        Looking again at the skill and competence issue.  Among the most
problematic of these are the low skill and competence levels of management and
workforce. It takes time for an economy and an education system to create
competent managers, technically proficient engineers, and skilled production
workers. Currently there are massive training efforts underway throughout the
developing world and in industrialized countries.1/  This is positive and
will help close the skills gap.T/
4.67        We can make a related point about the skill differences between
workers operating under the old approach to production and those under the new
practices. The latter workers may appear much more highly skilled than
production workers elsewhere, particularly those in developing countries, but
this is a misreading of the nature of their skills. The main difference is
that they are multi-skilled in typical production line jobs and are thus able
to perform a larger variety of tasks than are workers operating under Fordist
relations of production.
4.68        Workers under the new conditions are trained, paid and allowed to
do jobs that line workers elsewhere are not permitted to do. These restric-
tions on the latter have nothing to do with their skill level or ability to
learn but with management's choice not to give them the necessary training and
responsibility. Inevitably, the number-of entrants into the labor pool who
have the basic skills and education to become candidates for multi-skill
training is very limited--and this will remain a severe constraint for a long
while. But in the particular case of the engineering sector, existing workers
are, by definition, from the pool of competent people who in fact do have the
ability to learn. They may not have been given the training or the opportun-
ity to fulfill their productive potential--the critical policy problem is how
to get management to invest in their upgrading.
15./ This process has been well documented in Porteous (1987) who looked at the
efforts of Brazilian machine tool producers to survive the drastic 1980-
84 period where demand for their products dropped by nearly 70% overall,
and yet many firms survived and have emerged even stronger. Sabel suggests
on the basis of the Brazilian experience, that such "crisis-tested" firms
would make good candidates to adopt the new practices.
158/ See Kubr and Wallace (1984) for a review.
.15/ However, it should be clear that the substance of what is taught in the
courses is as important as the number of people being trained. Management
trainees will be ill-served if they are not introduced to best practice
management, as discussed in this paper.



- 96 -
4.69        Other constraints reexamined.  Scarcities of raw materials and
imported inputs, while acting as a constraint on output, should also alert
managers to the financial value of economizing on their use, for reducing
waste in general, and for eliminating defects in particular. Many of the new
practices, even introduced on a piecemeal basis, directly attack these
problems and should be attractive to developing country managers and
entrepreneurs.
4.70        Similarly, the predominance of old, low-volume machines in many
developing country engineering firms is often cited in the literature as a
constraint on output and productivity growth. So too is the fact that small
size--often mandated by government policy--has prevented firms from pursuing
both product specialization and internal specialization of tasks, such as
quality control. This has forced managers, engineers and production workers
to perform work for which they haven't been trained. The same negative
perspective comes out in more recent concerns that the low cost of labor will
inhibit developing country firms from investing in expensive automation
technology. It is possible to reinterpret each one of these "constraints" and
find potentially positive reasons for the introduction of the new practices.
4.71        "Filtering" conventional views about the problems of prod-action in
developing countries through an analytical framework informed by an under-
standing of the new principles and practices results in surprising insights.
We would venture the hypothesis that one of the reasons why some Western firms
have had difficulties with the new practices is precisely because of scale
factors, because their own production organization is hopelessly overcompli-
cated, and because they have large sunk investments in elaborate, high cost
and complex machinery. In Third World plants where scale is lower, the
process more labor intensive, and product flows much less complex, principles
that emphasize simplicity as the key to productivity improvement might even be
easier to implement than in the advanced industrial countries.  ..
4.72        The same logic applies to developing an entirely new set of
production relations between management and workers in a context where neither
group is as well educated or trained as their counterparts in advanced
countries. In the absence of a mass production tradition, and with only
limited Fordist relations of production, might it not be easier to introduce
the new relations into what are essentially "greenf'eld" sites?
4.73        The questions raised above, having to do with the possibilities
and the problems associated with introduction of the new approaches to
production organization and management in developing countries, have wide-
ranging implications. For instance, much conventional wisdom underpinning
mainstream industrialization literature on technology and industrial develop-



- 97 -
ment needs to be reconsidered.1/  On the research slde, the efforts of
domestic firms ln developing countries to apply these ideas, and the effolts
of Western and Japanese TNCs to transfer theim to their subsidiaries, need to
be examined.
4.74        In the area of policy, the evidence and arguments given above
indicate many aspects that need to be reexamined--for example, the scale of
resources devoted to management training and the substance of that training;
the nature of existing incentive programs designed to encourage new investment
in fixed capital; direct foreign investment and technology transfer policies;
local content policies; the underlying assumptions about attainable
productivity and output levels, and export prospects, under liberalization.
4.75        Thus the agenda for thought, research and action is extensive and
provocative. It is hoped that this paper will stimulate further exploration--
by a wider audience--of the issues raised by the emergence of the new
practices and their possible introduction in developing countries.
16/  Since this field is the author's main concern, the issue is worthy of some
provocative speculation.   It seems to us that many concepts that are
conventlonal wisdom in the technology and development literature and from
which we derive policy implications and proposals need to be thoroughly
reexamined through a new lens. We include here such cherished ideas as
the nature of "learning" within Third World firms, the importance of
incremental technical change and "capacity stretching", distinctions
between  'technological"   capacity  and  "productive"  capacity,   the
constituents of technological "mastery", the importance of organizing the
technology transfer process to maximize access to core technological
knowhow, etc.
Some of these concepts may have to be discarded or completely redefined
--such as the distinction between productive and technological capacity
and the notions of mastery, which pay lip service to the importance of
operational skills but really see the main source of productivity growth
as a firm's technical change capacity as distinct from its operational
skills. One of the implications of the process and organizational change
reviewed in this paper is that these distinctions no longer many any sense
in theory or in practice and that firms, particularly, those in the Third
World, should concentrate on developing the right form of production
organization and operational skills so crucial to the technical change
process.    Likewise,  the  failure of  the technology  and development
literature to grasp fully the importance of management skills per s is
now shown to be grossly at variance with the reality of some of the most
important examples of industrial development.   The literature has to
grapple fully with the implications of the Japanese experience, and with
the fact that the Japanese "model" is now defining industrial best-
practice.   It is clear that the new organizational paradigm has some
fundamental and challenging implications for the way problems of industrial
production, technology transfer and competitiveness in the Third World are
characterized and for the policy implications derived from that analysis.



D 99 -
BIBLIOGRAP
Abegglen, James and G. Stalk (1985) Kaisha: The JaRanese Corporation (New
York: Basic Books).
Abernathy, N. (1978) The Productivity Dilemma (Baltimore: John Hopkins
University Press).
Abernathy, N., K. Clark, and A. Kantrow (1983) Industrial Renaissance (New
York: Basic Books).
Altshuler, A. et al. (1984) The Future of the Automobile (Cambridge, Mass.:
MIT Press)
Amsden, Alice (1985) "The Division of Labor is Limited by the Rate of Growth
of the Market," Cambridge Journal of Economics.
Amsden, Alice (1977) "The Division of Labor is Limited by the Type of Market:
The Case of the Taiwanese Machine Tool Industry," World Development, 5,
No. 3, pp. 217-233.
Arbose, Jules (1985). "Manufacturing: Management's New Competitive Weapon,"
International Management, March 1985.
Ayres, R. and S. Miller (1984).  "Survey of Robotics and Flexible Manufactur-
ing Systems," Mimeo. Carnegie-Mellon University.
Bell, Martin (1982). "Learning and the Accumulation of Industrial Technologi-
cal Capacity in Developing Countries," Paper prepared for the Interna-
tional Workshop on Facilitating Indigenous Technological Capability,
Edinburg University, May.
Bell, Martin (1986). The Acguisition of Imported Technology for Industrial
Development: Problems of Strategy and Management in the Arab Region,
U.S. ECWA, Baghdad.
Bell, Martin, B. Ross-Larson and L. Westphal (1985) Technological Change in
Infant Industries: A Review of Empirical Evidence. World Bank Staff
Working Paper, Washington, D.C.
Bell, Martin and D. Scott-Kemmis (1980) "Learning and Technical Change in the
Development of Manufacturing Industry: A Case Study of a Permanently
Infant Industry" in F. Stewart and J. James (ed) The Economics of New
Technology in DeveloRing Countries, (London: Francis Pinter Ltd.).
Benchmark Research Ltd. (1987) "Surveys of UK Computers in Engineering, 1985,
1986, 1987."  Reported in Engineering Computers, Findlay publications,
Horton-Kirby, Kent.
Bernstein, Aaron and W. Zellner (1987) "Detroit vs. the UAW: At Odds Over
Teamwork," Business Week, August 24, 1987, pp. 54-55.



- 100 .
Bessant, J. (1985) "Flexible Manufacturing Systems - An Overview."
UNIDO/IS.539, UNIIDO, Vienna.
Bessant, J. (1986). "From Scientific Soldiers to General Jumbo: Employment
Trends in Computer-Integrated Manufacturing," Mimeo, Innovation Research
Group, Department of Business Management, Brighton Polytechnic.
Bessant, J. and B. Haywood (1986). "Flexibility in Manufacturing Systems,"
OMEGA: International Journal of Management Science, Vol. 14, No. 6.
Bessant, J. and Rush, H. (1987). "Integrated Manufacturing," Report submitted
to UNIDO, Center for Business Research, Brighton Business School.
Boddy, D. and Buchanan, D. Managing New Technology. (Oxford: Blackwell Pubs.
Ltd.).
Brandt, Richard (1986). "Retool or Die: Jobs Shops get a fix on the future,"
Business Week, June 16, 1986.
Brodner, P. (1983). 'Skill-based production - the superior concept," in H. J.
Bullinger and H. Warnecke, (eds) Towards the Factory of the Future.
(Stuttgart: Fraunhofer--IAO).
Carson, Clay (1986) "At Sanyo's Arkansas Plant the Magic Isn't Workirg,"
Business Week, July 14, 1986, pp. 51-52.
Chandler, A. P. (1962) Strategy and Structure (Cambridge, Mass.: MIT Press).
Chandler, A. P. (1977) The Visible Hand (Cambridge, Mass.: Harvard University
Press).
Child, F. C. and H. Kaneda, "Links to the Green Revolution: A Study of Small-
Scale Agriculturally-Related Industry in the Pakistan Punjab," Economic
Develogment and Cultural Change, Vol. 23, No. 2, January.
Chudnovsky, Daniel (1987), "The Diffusion and Production of Numerically
Controlled Machine Tools with Special Reference to the Argentine
Experience,' Mimeo, Center for International Economics, Buenos Aires.
Clark, R., The Japanese Comgany (1979). Manufacturing Matters: The Myth of
the Post-Induggtrial Economy, (New York: Basic Books).
Cohen, Stephen S. and J. Zysman (1987). Manufacturing Matters: The Myth of
the Post-Industrial Economy, (New York: Basic Books).
Cook, James. (1986). "Back to Simplicity," Forbes, August 25, 1986,
pp. 32-33.
Crosby, Philip B. (1979). Ouality is Free: The Art of Making Ouality
Certain. (New York: McGraw-Hill).



- 101 -
Dahlma:. Carl and L. Westphal (1982) "Technological Effort in Industrial
Development--An Interpretive Survey of Recent Research," in Stewart and
James, (op. cit.).
Deming, W. Edward (1986). Out of the Crisis (Cambridge, Mass.: Massachusetts
Institute of Technology).
Deming, W. Edward (1982). Ouality. Productivity and Competitive Position
Cambridge, Mass.: Massachusetts Institute of Technology).
Dempsey, P. (1983). "New Corporate Perspectives in FMS," in K. Rathmill
(ed.), Proceedings of the 2nd International Conference on Flexible
Manufacturing Systems," (Bedford, U.K.: IFS Publications).
Dickson, K. and .J. March (1978). The Microelectronics Revolution: A Brief
Assessment of the Industrial Impact with a Selected Bibliography.
(Birmingham: Technology Policy Unit, University of Aston).
Dosi, Giovanni (1982). "Technological Paradigms and Technological Trajec-
tories," Research Policy, Vol. 11, No. 3, June.
Economic Commission for Europe (1986). Recent Trends in Flexible Manufac-
turing, UN. ECE, New York.
Economist. The (1987). "Factory of the Future," The Economist, May 30, 1987.
Electronics Week (1984) July 21, 1984.
European Trade Union Institute (1983) The ImRact of Microelectronics on
Employmer- in Western EuroRe in the 1980s (Brussels: ETUI).
Feder, B. J. (1987). "Hewlett Accepts Automation," in New York Times,
September 10, 1987.
Feigenbaum, A. V. (1961). Total Ouality Control: Engineering and Management.
(New York: McGraw-Hill).
Fleck, J. (1982). "Employment Implications of Robots," in T. Lupton, (ed.)
Proceedings of First International Conference on Human Factors in
Manufacturing, (Kempston U.K.: IFS Publications).
Franzman, Martin (1986). "International Competitiveness, Technical Change and
the State: The Machine Tool Industry in Taiwan and Japan," World
Development, Vol. 14, No. 12.
Franzman, Martin (1982). Learning and the Capital Goods Sector Under Free
Trade: The Case of Hong Kong, Working Paper No. 1, Center for African
Studies, University of Edinburgh.
Franzman, Martin (1984). "Promoting Technological Capability in the Capital
Goods Sector: The Case of Singapore," Research Policy, 13:33-54.



- 102 -
Fraser, Douglas (1986). "Management Discovers the Human Side of Automation,"
Business Week, September 29, 1986, pp. 70-75.
Freeman, C., J. Clark and L. Soete (1984). Unemployment and Technical
Innovation:  A Study of Lon, waves and Economic Develo2ment.  (London:
Francis Pinter Ltd.).
Frobel, F., J. Heinrichs and 0. Kreye (1980). The New International Division
of Labor. (Cam';-t.dge University Press).
Fukuda, John (1988). Japanese-Style Management Transferred: The Exgerience
of East Asia. (London: Routledge).
Fukuda, John (1986). "What We Can Really Learn from Japanese Management,"
Journal of General Management, Vol. 11, No. 3, Spring, pp. 16-26.
Gaetan, Manuel (1986). "An Alternative to UPS," Bobbin, February.
Garvin, David A. (1983). "Quality on the Line," Harvard Business Review,
September-October, pp. 65-75.
Garvin, David A. (1988).  Managing Oualitv:  The Strategic and Competitive
Kdgj. (New York: Free Press).
Gershenberg, Irving (1987). "The Training and Spread of Managerial Know-how:
A Comparative Analysis of Multinational and Other Firms in Kenya," Wo7Ld
Develogment, Vol. 15, No. 7, July.
Gerwin, D. (1982). "Do's and Don't's of Computerized Manufacturing," Harvard
Business P.eview, March-April.
Hall, Robert H. (ed.) (1981). "Driving the Productivity Machine: Production
Planning and Control in Japan," (Falls Church, VA: American Production
and Inventory Control Society).
Hall, Robert W. (1983). Zero Inventories. (New York: Dow Jones Irwin).
Hampton, William J. (1987a). "Why Image Counts: A Tale of Two Industries,"
in Business Week, June 8, 1987.
Hampton, William J. (1987b). "Will Saturn Ever Leave the Launchpad," Business
Week, March 16, 1987.
Handke, G. (1983). "Design and Use of Flexible Automated Manufacturing
Systems," in K. Rathmill, op. cit.
Hayes, Robert H. and Steven C. Wheelwright (1984). Restoring Our Competitive
Edge: Competing Through Manufacturing. (New York: John Wiley & Sons).
Hayes, Robert H. (1981). "Why Japanese Factories Work," Harvard Business
Review, July-August, pp. 57-66.



- 103 -
Haywood, B. and J. Bessant (1986). "Cad-Cam Integration: Flexible Manufac-
turing Systems in the U.K., "Mimeo. Department of Business Management,
Brighton Polytechnic.
Helm, L. and K. Buell (1987). "Why Kodak is starting to Click Again,"
Business Week, February 23, 1987.
Hill, Stephen (1987). "Basic Design Principles for National RIesearch in
Developing Countries," Technology in Society, Vol. 9.
Hines, C. and G. Searle (1979).  Automatic UnemRloyment.  (London:  ERR
Hirschman, Albert 0. (1981). "The Rise and Decline of Development Economics,"
in Hirschman, A. Essays in Tresgassing. (Cambridge: Cambridge Univer-
sity Press).
Hoerr, John (1986).  "'A._-:ment Discovers the Human Side of Automation,"
Business Week, September 29, 1986.
Hoffman, Kurt (1986). Managing Technological Change in Developing Countries:
The Impact and Policy Implications of Microelectronics. (London:
Commonwealth Secretariat).
Hoffman, Kurt (ed.) (1985). Microelectronics. International Competition and
DeveloRment Strategies: The Unavoidable Issues, Special Issue of World
Development, March 1985.
Hoffman, Kurt and Norman Girvin (1989). Managing International Technologv
Transfer:  A Strategic Aggroach for Developing Countries.  International
Development Research Center, Ottawa, Canada.
Hoffman, Kurt and Howard Rush (1980). "Microelectronics, Industry and the
Third World," Futures, August 1980.
Hoffman, Kurt and Raphael Kaplinsky (1988). Driving Force: The Global
Restructuring of Technologv. Labor and Investment in the Automobile and
Components Industry. (Boulder, Colorado: Westview Press).
Ishikawa, Kaoru (1972). Guide to Ouality Control. (Tokyo: Asian Produc-
tivity Organization).
Ishikawa, Kaoru (1985). What is Total Ouality Control? JaDanese Way.
(Englewood Cliffs, J.J.: Prentice-Hall INc.).
Jacobsson, S., "Technical Change and Industrial Policy: The Case of Computer
Numerically Controlled Lathes in Argentina, Korea and Taiwan," in
Hoffman (1985) op. cit., pp. 353-370.
Jacobsson, S. and J. Sigurdson (eds.) (1983). Technical Trends and Challenges
in Electronics. Research Policy Institute, University of Lund, Lund,
Sweden.



- 104
Jaikumar, Ramchandrar. (1986). "Postindi;strial Manufacturing," Harvard
Business Review, November-December, pp. 69-76.
James, Jeffrey (1987). "Microelectronics and the Third World: An Integrative
Survey of the Literature," United Nations University, Maastricht,
Holland.
Jelinek, M. and J. D. Goldhar (1984). "Strategic Implications of the Factory
of the Future," Sloan Management Review, Summer 1984.
Johansson, Henry J. (1988). "Simplify the Shop Floor," New York Times.
December 1, 1988.
Jones, Bryn and P. Scott (1987). "Flexible Manufacturing Systems in Britain
and the U.S.A.," New Technology. Work and Employment, Vol. 1, No. 2,
March, pp. 27-36.
Jones, Daniel and J. Womack (1985). "Developing Countries and the Future of
the Automobile Industry,' in Hoffman (1985) op. cit., pp. 393-408.
Juran, J. M., ed. (1951). Oualitv Control Handbook. (New York: McGraw-
Hill).
Kaplinsky, Raphael (1984'   Automation in a Changing World, (New York:
Longmans Publishing Cv.
Kaplinsky, Raphael (1986). "Changes in Economy of Scale: The Implications
for Appropriate Technology,' Mimeo. Institute of Development Studies,
University of Sussex.
Kaplinsky, Raphael (1982). "Comparative Advantage in an Automating World."
IDS Bulletin, Vol. 13, No. 2, March.
Kaplinsky, Raphael (1985). "Electronics-Based Automation and the Onset of
Systemofacture: Implications for Third World Industrialization," in
Hoffman (1985) op. cit., pp. 423-440.
Kaplinsky, Raphael (1984b). "The International Context for Industrialization
in the Coming Decade," in R. Kapl5vsky (ed.).  Third World Industriali-
zation in the 1980s. (London: F-ank Cass).
Katz, Jorge (1980). Domestic Technology Generation in LDCs: A Review of
Research Findings, Working Paper No. 35, BID/CEPAL Research Programme on
Scientific and Technological Development in Latin America, Buenos Aires.
Katz, Jorge (1982). "Technology and Economic Development: An Overview of
Research Findings," in M. Syrquin and S. Teitel (eds.). Trade.
Stability. Technology and Eauity in Latin America, (New York: Academic
Press), pp. 285-315.
Kim Linsu (1987). "New Technologies and Their Economic Effects: A Feasibi-
lity Study in Korea," Mimeo. Korea University.



- 105 -
Kerbs, Ken (1987). "Korea's Newest Export: Management Style," Business Week,
January 19, 1987, p. 66.
Khera, K. and I. Namoto (1986). "Total Technological Trauisfer Through
Collaboration," Mimeo.
Koibayashi, Koji (1986). "Quality Management of NEC Corporation," Oualitv
Progress, April 1986.
Kubr, M. and J. Wallace (1984).  Successes  .d Failures in Meeting the Manag:
ment Challenge, World Bank Staff Working Paper No. 585.
Washington, D.C.
Kono, T. (1982). "Japanese Management Philosophy: Can it be Exported?" Long
Range Planning, June, pp. 90-102.
Lamming, R. (1986). "For Better or Worse: Buyer/Supplier Relationships in
the UK Vehicle Industry," International Journal of Operations Manage.
ment, December.
Landes, David J. (1972).  Prometheus Unbound.  (Cambridge:  Cambridge Univer-
sity Press).
Lawler, Edward E. and Susan A. Mohrma (1985). "Quality Cir1.ces After the
Fad," Harvard Business Review, January-February, pp. 65-71.
Liebenstein, Harvey (1966). "Allocative Efficiency vs. X-Efficiency," The
American Economic Review, Vol. 56, pp. 392-415.
Liebenstein, Harvey (1977). "X-Efficiency, Technical Efficiency and Incom-
plete Information Use: A Comment," Economic DeveloRment and Cultural
Change, Vol. 25, No. 2, January, pp. 311-316.
Magaziner, I. and M. Hout (1980). Minding America's Business. (New York:
Harcourt, Brace and Jovanovich).
Matsumoto, Koji (1982). "The Secret of Japanese Productivity," Journal of
Jaganese Trade and Industry, No.1
Microelectronics Monitor (1987). I sie 21, Fall 1987. UNIDO, Technology
Division, Vienna, Austria.
Mitchell, R. (1986). "Detroit Stumbles its Way to the Future," Business Week,
June 16, 1986.
Mitchell, R. (1986).  "How Ford Hit the Bull's Eye with Taurus," Business
Week, June 30, 1986.
Morgan. I. P. (1986) "The Purchasing Revolution, " IMEDE, August 1986.



- 106 .
Murakama, Y. (1982).  "Toward a Socio-Cultural Expitnation of Japan's Economic
Performance," in K. Yamamura (ed.) Policy and Trade Issues of the
Ja2anese Econ3My.:  American and JagAnese Perspectives.  (Seattle:
University of Washington Press).
Nasar, S. (1987).  "Competitiveness:  Getting it Back," fortMne, April 27,
1987.
National Economic Development Office (1984). The Impact of New Technology on
Engineering Batch Production. (London: NEDO).
National Machine Tool Builders Association (1986). The Economic Handbook of
the Machine Tool Industrv. 1986-87, NMTBA, Washington, D.C.
National Research Council/National Academy of Engineering (1986). Toward a
New Era in U.S. Manufacturing: The Need for a National Vision.
National Academy Press, Washington, D.C.
Nelson, Richard (1979). Innovation and Economic Develogment: Theoretical
Retrosbect and Prosgect, Working Paper No. 31, BID/CEPA1 Research
Programme in Science and Technology in Latin America, Buenos Aires.
New, C. (1986). Managing Manufacturinf Operations in the U.K. (Cranfield,
Surrey: Bvitish Institute of Management).
Northcott, J. et. al. (1986). Microelectronics in Industry. (London: Policy
Studies Institute).
Office of Technology Assessment (1984). Computerized Manufacturing Automa-
tign. OTA, Washington, D.C.
Pack, Howard (1980). Fostering the Capital Goods Sector in LDCs: A Survey of
Evidence and Reguirements, World Bank Staff Working Paper, No. 376,
Washington, D.C.
Perez, Carlotta (1985). "Microelectronics, Long Wages and World Structural
Change: New Perspectives for Developing Countries," in Hoffman (1985)
op. cit., pp. 441-463.
Piore, Michael J. and Charles F. Sabel (1984). The Second Industrial Divide:
Possibilities for Prosoerity. (New York: Basic Books).
Piper, Allan (1986). "Is Traditional Management Dead?" International
Mgnazement, January, pp. 22-26.
Port, Otis (1987). "Making Brawn Work With Brains," Business Week, April 20,
1987.
Port, Otis (1986).  'High Tech to the Rescue," Business Week, Business Week,
June 16, 1986.
Port, Otis (1987).  "The Push for Quality," Business Week, Ju.ne 8, 1987.



- 107 -
Porteous, Michael (1987) Learrning Under Crisis: The Machine Tool Industry in
Brazil, Ph.D. thesis, Science Policy Research Unit, University of
Sussex.
Rada, Juan (1982).  Microelectronics:  Its Impact and Policy Imnlications,
UNIZO ID/WG. 372/5, Vienna.
Radford, G.S. (1922).  The Control of Ouality in Manufacturing.  (New York:
Ronald Press).
Rohter, Larry (1987). "Plants in Mexico Help Japan Sell to U.S., New York
Times, May 26, 1987."
Robinson, A. E. (1988). "A Comparative Study of the Economic Effects of
External and Internal Linkages Achieved Through Compensatory-Type
Investments: The Mexican Automotive industry." Ph.D. thesis (unpub-
lished), School of Government and Business Administration, George
Washington UniverUity, Washington, D.C.
Rosenberg, N. and C. Frischtak (1984). "Technological Innovation and Long
Waves," American Economic Review: Papers and Proceedings, May.
Sabel, Charles F. (1986). "Changing Models of Economic Efficiency and Their
Implications for Industrialization in the Third World," in Development.
Democracy and the Art of Trespassing, A. Foxley (ed.) (University of
Notre Dame Press).
Sabel, Charles F. and J. Zeitlin (1985). "Historical Alternatives to Mass
Production," Past and Present, No. 108, August, pp. 133-76.
Saporito, Bill (1986). "IBM's No-Hands Assembly Line," Fortune, September 15,
1986, pp. 105-107.                                             -
Schandler, Phillip (1981). "U.S. Workers Feel that Reporting Waste Doesn't Do
Any Good," Washington Star, April 15, 1981, p. A-5.
Schiller, Z. (1987). "Appliance Repairmen are Getting Lonelier," Business
Week, June 8, 1987.
Schonberger, Richard J. (1982). JaDanese Manufacturing Techniques: Nine
Hidden Lessons in Sim2licitv. (New York: Free Press).
Schonberger, Richard J. (1986). World Class Manufacturing: The Lessons of
Simplicitv ADDlied. (New York: Free Press).
Senker, Peter and Arnold, Erik (1982). Designing the Future: The Skills
Implications of Interactive CAD. Occasional Paper No. 9, Engineering
Industry Training Board, Watford, U.K.
Senker, Peter and Beesley, Mark (1988). "The Need for Skills in the Factory
of the Future," New Technology. Work and Em2loyment. Spring, 1988.



- 108 -
Shaiken, Harley (1982). "Technological Change and Shop-Floor Response."
Mimeo, Program in Science, Technology and Society, Massachusetts
Institute of Technology.
Shewart, W. A. (1931). Economic Control of Ouality of Manufactured Product.
(New York: Van Nostrand and Company).
Shingo, Shigeo (1985).  A Revolution in Manufacturina:  The SMED System.
(Stamford, Conn.: Productivity Press).
Shingo, Shigeo (1981). Studv of Toyota Production Svstem from Industrial
Engineering ViewDoint. (Tokyo: Japan Management Association).
Simon, Ruth and A. Tanzer (1987). "Fiasco at Jujitec," Fobes, March 23, 1987,
pp. 60-61.
Soete, Luc (1985). "International Diffusion of Technology, Industrial
Development and Technological Leapfrogging," in Hoffman (1985) op. cit.,
pp. 409-422.
Soete, Luc and G. Dosi (1983). Technological Trends and Emplovment:
Electronics and Communications.  (London:  Cower Publishing Co.).
Stavro, Barry (1986). "Digging Out," Forbes, November 3, 1986.
Suzaki, Kiyoshi (1987). The New Manufacturing Challenge: Techniaues for
Continuous Improvement. (New York: Free Press).
Takamiya, M. (1981). "Japanese Multinationals in Europe: Internal Operations
and Their Policy Implications," Columbia Journal of World Business,
Summer, 1981.
Taylor, Frederick Winslow (1919). Shop Management. (New York: Harper &
Bros.).
Teitel, Simon (1981). "Creation of Technology Within Latin America," The
Annals of the American Academy of Political and Social Science, Vol.
458, November, pp. 136-150.
Teitel, Simon (1987). "Science and Technology, Country Size and Economic
Development: An International Comparison," World Development, September
1987.
Teitel, Simon (1979). Notes on Technical Change Induced Under Conditions of
Protection, Distortions and Rationing, Working Paper No. 34, BID/CEPAL
Research Programme on Science and Technology in Lacin America, Buenos
Aires.
Thoburn, J. T. (1973). "Exports and the Malaysian Engineering Industry: A
Case Study of Backward Linkages," Oxford Bulletin of Economics and
Statistics, Vol. 35, No. 2, May, pp. 91-117.



- 109 -
UNCTAD (1985).  The Capital Goods Sector in Developing Countries.  UNCTAD
Technology Division, Geneva.
Voss, C. (1986). "A Manufacturing Strategy Perspective," in C. Voss (ed.),
Managing Advanced Manufacturing Technology. (Kempston, U.K.: IFS
Publications).
Waterlow, G. and J. Monniot (1986). The State of the Art of CAPM in the U.K.
Report to the ACME Directorate, Science and Engineering Research
Council, Swindon, U.K.
Wall, T. et al (1986). Presentation made at Manpower Services Commission
Workshop on Skills and New Technology, Aston University, March 1986.
Walton, May (1986).  The Deming Management Method.  (New York:  Dodd, Mead &
Company).
Weiss, Andrew (1984). "Simple Truths of Japanese Manufacturing," Harvard
Business Review, july-August, pp. 119-125.
Wheelwright, Steven C. (1981). "Japan--Where Operations Really are
Strategic," Harvard 3usiness Review, July-August, pp. 67-74.
Williams,. Douglas (1986). "JIT vs. MRP: The Push-Pullers Call a Truce,"
Automotive Industries, July, pp. 30-31.
Womack, James (1987). "The Development of the Chinese Motor Vehicle Industry:
Strategic Alternatives and the Role of Foreign Firms," Mimeo. Mas-
sachusetts Institute of Technology. Paper presented at the Pacific
Forum, Makaha, Oahu Hawaii, December 1, 1987.
Womack, James (1986). "Prospects for the U.S.-Mexican Relationship in the
Motor Vehicle Sector," Mimeo, M.I.T.
Work, C. P. et al. (1987). "How to Beat the Japanese," U.S. News and World
Report, August 24, 1987.
Wyatt, Sally (1987). The Changing Technological Scene: The Case of the OECD
Countries. UNIDO IPCT. 43. Technology Trends Series No. 7, Technology
Division, Vienna.
Yang, C. T. (1977). "Management Styles: American vis-a-vis Japanese,"
Columbia Journal of World Business, Fall.



- 110 -
INDUSTRY SERIES PAPERS
No.  1    Japanese Direct Foreign Investment:  Patterns and Implications
for Developing Countries, February 1989.
No.  2    Emerging Patterns of International Competition in Selected
Industrial Product Groups, February 1989.
No.  3    Changing Firm Boundaries:  Analysis of Technology-Sharing
Alliances, February 1989.
No.  4    Technological Advance and Organizational Innovation in the
Engineering Industry, March 1989.
No.  5    Export Catalyst in Low-Income Couttries, November 1989.
No.  6    Overview of Japanese Industrial Technology Development,
March 1989
No.  7    Reform of Ownership and Control Mechanisms in Hungary and China
April 1989.
No.  8    The Computer Industry in Industrialized Economies:  Lessons for
the Newly Industrializing, February 1989.
No.  9    Institutions and Dynamic Comparative Advantage Electronics
Industry in South Korea and Taiwan, June 1989.
No. 10    New Environment for Intellectual Property, June 1989.
No. 11    Managing Entry Into International Markets;  Lessons From the
East Asian Experience, June 1989.
No. 12    Impact of Technological Change on Industrial Prospects for the
LDCs, June 1989.
No. 13    The Protection of Intellectual Property Rights and Industrial
Technology Development in Brazil, September 1989.
No. 14    Regional Integration and Economic Development, November 1989.
No. 15    Specialization , Technical Change and Competitiveness in the
Brazilian Electronics Industry, November 1989.
No. 16    Small Trading Companies and a Successful Export Response: Lessons
From Hong Kong, December 1989.
(next page)



: )1 STR2ES SERIES PAPI IS CON'T.
- 111 -
No. 17    Flowers:  Global Subsector Study, December 1989.
No. 18    The Shrimp Industry:  Global Subsector Study, December 1989.
No. 19    Garments:  Global Subsector Study, January 1990
No. 20    World Bank Lending for Small and Medium Enterprises:  Fifteen
Years of Experience, Dec. 89.
No. 21     Reputation in Manufactured Goods Trade, Dec. 89.
No. 22   Foreign Direct Investment From the Newly Industrialized Economies,
December 1989.
No. 23     Buyer-Seller Links for Export Development, March 1990.
No. 24    Technology Strategy & Policy for International Competitiveness: A
Case Study of Thailand, February 1990.
Note:     For extra copies of these papers please contact Miss. Wendy Young
on extension 33618, Room S-4101.



- 112 -
mCY SERIES PAPERS
No. 1     Energy Issues in the Developing WorLd, February 1988.
No. 2     Review of World Bank Lending for Electric Power, March 1988.
No. 3     Some Considerations in Collecting Data on Household Energy
Consumption, March 1988.
No. 4     Improving Power System Efficiency in the Developing Countries
through Performance Contracting, May 1988.
No. 5     Impact of Lower Oil Prices on Renewable Energy Technologies, May
1988.
No. 6     A Comparison of Lamps for Domestic Lighting in Developing Countries,
June 1988.
No. 7     Recent World Bank Activities in Energy (Revised September 1988).
No. 8     A Visual Overview of the World Oil Markets, July 1988.
No. 9     Current International Gas Trades and Prices, November 1988.
No. 10    Promoting Investment for Natural Gas Exploration and Production in
Developing Countries, January 1989.
No. 11    Technology Survey Report on Electric Power Systems, February 1989.
No. 12    Recent Developments in the U.S. Power Sector and Their Relevance for
the Developing Countries, February 1989.
No. 13    Domestic Energy Pricing Policies, April 1989.
No. 14    Financing of the Energy Sector in Developing Countries, April 1989.
No. 15    The Future Role of Hydropower in Developing Countries, April 1989.
No. 16    Fuelwood Stumpage:  Considerations for Developing Country Energy
Planning, June 1989.
No. 17    Incorporating Risk and Uncertainty in Power System Planning, June
1989.
No. 18    Review and Evaluation of Historic Electricity Forecasting
Ezperience, (1960-1985), June 1989
No. 19    Woodfuel Supply and Environmental Management, July 1989



- 113 -
ERCY SERIES PAPERS   cont'd
No. 20    The Malawi Charcoal Project - Experience and Lessons, January 199u.
No. 21    Capital Expenditures for Electric Power in the Developing Countries
in the 1990s, February, 1990.
No. 22    A Review of Regulation of the Power Sectors in Developing Countries
No. 23    Summary Data Sheets of 1987 Power and Commercial Energy Statistics
for 100 Developing Countries, March 1990
No. 24    A Review of the Treatment of Environmental Aspects of Bank ENergy
Projects, March 1990
Note:     For extra copies of these papers please call Ms. Mary Fernandez on
extension 33637.