Building Research & Information

ISSN: 0961-3218 (Print) 1466-4321 (Online) Journal homepage: http://www.tandfonline.com/loi/rbri20

Zephyrs of creative destruction: understanding the

management of innovation in construction

Graham Winch

To cite this article: Graham Winch (1998) Zephyrs of creative destruction: understanding the
management of innovation in construction, Building Research & Information, 26:5, 268-279, DOI:
10.1080/096132198369751

To link to this article: https://doi.org/10.1080/096132198369751

Published online: 14 Oct 2010.

Submit your article to this journal

Article views: 4334

View related articles

Citing articles: 193 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rbri20
0961–3218 # 1998 E & FN Spon

Zephyrs of creative destruction: understanding the

management of innovation in construction

Graham Winch

Bartlett School Of Graduate Studies , University College London , Gower Street , London WC1E 6BT ,
UK
E -mail : g . winch@ ucl. ac. uk

The aim of this paper is to propose a comprehensive framework for the management of innovation in
construction , addressing the construction innovation problem in two distinctive ways at the institutional
and Žrm levels. First, an institutional perspective derived from research on complex systems industries is
developed which provides an alternative to the volume production model for construction innovation
research. The roles of the innovation infrastructure, innovation superstructure and systems integrator are
all identiŽed and applied to construction. The paper then moves on to the Žrm level where the two key
innovation dynamics ­ the top-down adoption=implementation dynamic and the bottom up problem
solving=learning dynamic are identiŽed. The paper ends by calling for more case studies of the trajectories
of construction innovations.

L’ objet de cet article est de proposer un cadre global ouÁ ge rer l’ innovation dans le secteur de la construction;
l’ auteur aborde la question de l’ innovation sous deux angles diffe rents, au niveau des institutions et celui des
industriels . En un premier temps, on de veloppe une perspective institutionelle de rive e de la recherche sur les
systeÁ mes complexes; on de bouche alors sur une alternative au modeÁ le de volume de production applique aÁ la
recherche en matieÁ re d’ innovation dans la construction. Les roà les de l’ infrastructure et de la superstructure de
l’ innovation et celcui de l’ inte grateur de systeÁ mes sont tous de Žnis et applique s aÁ la construction. L’ auteur
passe ensuite au niveau de l’ industriel et de Žnit les deux axes principaux de l’ innovation, la dynamique
descendante d’ adoption=mise en ú uvre, d’ une part et, d’ autre part, la dynamique ascendante de re solution
des probleÁ mes et d’ enseignment aÁ en tirer, L’ auteur demande, pour conclure, que soient pre sente s davantage
de cas d’ e tude portant sur les itine raires suivis par des innovations dans le secteur de la construction.

Keywords: construction innovation, systems integrator, complex product system, adoption=implementation ,

problem solving=learning

One problem after another of the supply of passed construction by. Ever since the emergence
commodities to the masses has been success- of volume production methods in the late 19th
fully solved by being brought within the century, there have been repeated attempts to
reach of the methods of capitalist produc- apply them to the construction industry, ranging
tion . The most important one of those that from Gilbreth’ s work study of bricklayers in the
remain, housing, is approaching solution by 1900s, the mass production of housing in Dessau
means of the prefabricated house. in the 1920s, to the industrialized housing of the
1960s. Only recently, there have again been calls
Introduction for the Henry Ford of housing (Miles 1996).

As with most of his prognoses for the future of Such attempts have repeatedly failed, with the
capitalism , Schumpeter, writing in the late 1930s result that the relative cost of housing compared
(1976 p. 68), was wrong about construction ­ his to other goods and services has been rising
’ gale of creative destruction’ (1976, p. 84) has inexorably . This is a problem common to all

268 Bu i l d in g Re s e a r ch & In f o r m a t io n (1998) 26(4), 268–279

 ZEPHYRS OF CREATIVE DESTRUCTION

advanced nations, as a recent international but, I would submit, it has a wider relevance to
seminar demonstrated.1 Therefore, it is worth all construction industries.
reecting upon why the construction of housing
and other built products has been so resistant to
the virtuous cycle of simultaneous cost reduction Construction as a complex systems
and quality improvement that has beneŽted most industry
other industries over the last century. The
tenacity and pervasion of the problem suggests The burgeoning literature on innovation, like
that the volume manufacturing model might not research on most management topics, tends to
be the appropriate one for construction. assume that all industries follow a product
development type model. In this model Žrms
The problem is not that there has been no follow market signals to develop new products
innovation ­ indeed bibliometric data suggest that are produced on a volume basis and sold into
that the industry is a lively source of new ideas.2 a mass market. Such products have distinctive life
The problem is that the rate of innovation lags cycles where the innovative opportunity shifts
behind most other sectors, and appears to be from the product itself in the early stages of the
falling further and further behind. Certainly, life cycle to the process by which it is produced in
rates of productivity growth in the EU and US the later stages. The new technologies that are
industries are low compared to other sectors, or incorporated in these new products are, in the
even negative (Quigley, 1982; Allen, 1985; most sophisticated industries, the result of re-
Margirier, 1988). Moreover, innovation efforts in search and development programmes, and in
the industry are disproportionately orientated such cases innovation rates may be usefully
towards product enhancement rather than pro- measured by patent registration. In this model,
cess improvement (Gann et al., 1992). This paper innovation is largely an in-Žrm problem, although
explores some of the distinctive structural fea- suppliers may have to be involved to ensure that
tures of the construction industry with the aim of the new product can be made, and so increasing
analysing some of the principal processes of attention has been paid of late to the processes of
construction innovation, and thereby providing a innovation within networks of Žrms.
comprehensive conceptual model for the manage-
ment of innovation in construction. It will be This Schumpeterian model has tended to be
suggested that the reasons for the relatively low reected in recent thinking on innovation in
rate of innovation lie with these structural construction , but new research has identiŽed
features, and that unless innovation programmes another innovation model speciŽc to what have
fully comprehend these features, they are likely been called complex product systems (Miller et al.,
to prove unsuccessful. 1995) which may be more appropriate. Complex
product systems are distinguished by the follow-
The exploration will be at two levels. The Žrst is ing characteristics:
the structural features of the industry as a whole,
· many interconnected and customized ele-
adapting work on complex systems industries to the
ments organized in an hierarchical way;
construction case, and thereby providing an
analysis of the institutional context in which · nonlinear and continuously emerging prop-
Žrms manage their innovation processes. The erties where small changes to one element of
second is an analysis of innovation processes at the system can lead to large changes else-
the level of individual Žrms and projects, where in the system;
capturing the many different types of process
· a high degree of user involvement in the
which have to be managed for successful
innovation process.
innovation. For the purposes of this exploration,
the management of innovation is deŽned, follow - The industries which create such complex product
ing Van de Ven (1986), as the ’ management of systems are characterized as complex systems
new ideas into good currency’, which elegantly industries, and share many of the complex and
restates the Schumpeterian distinction between emergent features of their products.
invention (generating new ideas) and innovation
(applying new ideas). The analysis is based It has become increasingly common to analyse
mainly on the experience of the UK industry, innovation in terms of actor ­ system networks

269
WINCH

(e.g. Law and Callon, 1992), an approach applied integrators who supply complete ight simulation
to construction by Shove and her colleagues systems to air carriers for the training of their
(Connaughton et al., 1995; Rasmussen and Shove, aircrew .
nd). However, to date, these applications have
been to speciŽc innovations (cladding and ready- Examination of the analysis of the principal
mixed concrete respectively), and need to be features of the constructed product by Nam and
taken further to provide a generic model. Gann Tatum (1988) strongly supports the contention
(Gann et al., 1992; Gann, 1997) has developed a that the constructed product is a complex product
model of the construction system as a supply system, and that construction is, therefore, a
chain, showing the activities of the different complex systems industry. However, it has a
actors in the system as the inventions deriving number of distinctive features which also set it
from research and development (R&D) pro- apart from the model of the complex systems
grammes are diffused and implemented on industry developed by Miller and his colleagues.
speciŽc projects. However, formal R&D, while
very important, is not the only source of First, the systems integrator role is shared between
innovation in construction or any other industry, the principal architect=engineer and the principal
and a broader perspective that captures all contractor. Thus construction typically has two
modes of innovation is required. The strength separate systems integrators ­ one at the design
of the complex systems industries model is that stage and one at the construction stage. Secondly,
of providing a clear articulation of the institu- the fragmentation of the professional bodies in
tional context in which Žrms innovate, categoriz- construction has weakened their ability to act as
ing the different types of actor in the innovation honest brokers of innovations as they typically
network, and identifying their distinctive roles in threaten the interests of one or other amongst
the innovation process. This allows the develop- them. In Žlling this gap, government funded
ment of a generic model of the structural context national construction research organizations ­
of the management of innovation in construction. such as the Building Research Establishment in
Developing their model in the context of the the UK (cf, Courtney, 1997) ­ have played a very
ight simulation industry, Miller et al. (1995) important role in brokering innovations through
distinguish between the innovation superstructure the regulatory system. Thirdly, trade contractors
of air carriers (clients ), regulators, and profes- (specialized suppliers) are rarely given full techni-
sional bodies, and the innovation infrastructure of cal authority, and are often subject to separate
specialized suppliers and aircraft builders. The specialist consultants. These two also need to be
interface between the two is the role of systems distinguished from component suppliers. Fig. 1

innovation superstructure

professional
clients regulators
institutions

SYSTEMS INTEGRATORS
principal principal
architect/ contractor
engineer

trade specialist component

contractors consultants suppliers

innovation infrastructure

Fig. 1. Construction as a complex systems industry (source: adapted from Miller et al., 1995, Fig. 2).

270
 ZEPHYRS OF CREATIVE DESTRUCTION

illustrates the complex systems industry concept either by an exploitation trap where the system is
applied to construction. institutionally locked into particular technologies
as in Sweden, and the exploration trap where
While recent research has emphasized the multi- technologies are continually re-invented in a
ple sources of innovation in construction, and the circular rather than progressive manner as in
embeddedness of actors in innovation networks, Denmark. Extending this model, it might be
two of the key roles ­ that of the regulators, and suggested that the British system is a victim of
that of the professional institutions and research the exploration trap, where everything tends to
establishments as brokers has received less atten- be designed from Žrst principles for every
tion. The importance of the regulatory role has project. This explains the paradox of how the
been demonstrated by the research of Bazin British system can apparently have simulta-
(1993), who analysed the regulatory regime, and neously too much and too little innovation ­
hence space for innovation without regulatory there are plenty of new ideas, but they tend not
change, in the leading European nations, an to achieve good currency.
emphasis supported by Quigley’ s (1982) analysis
of US housebuilding. By regulatory regime is here
meant the technical regulations aimed at assuring The dynamics of innovation in a
the integrity and performance of the constructed complex systems industry
product, not those socio-economic regulations
essentially aimed at controlling what products Innovation is inherently a process through time,
are built where. Bazin characterizes the different and in order to illustrate the dynamics of
national regimes along two dimensions ­ the innovation in a speciŽc complex systems industry,
degree of responsibility of the state for the and the interactions between the innovation
enforcement of technical regulations compared superstructures and infrastructures, it will be
to that of the actors in the system, and the degree useful to take one relatively well-researched
to which the technical regulations are prescription example ­ the shift from load-bearing masonry
or performance oriented, as is shown in Fig. 2. to structural framing in the UK around 1900
(Bowley, 1966; Cusack, 1986; Cusack, 1987;
Although this basic complex systems industrial Lawrence, 1990). Two innovations were available
structure is shared by all construction industries, ­ reinforced concrete and steel ­ both imported
institutional differences between national con- from abroad (France and the United States
tracting systems mean that the brake on innova- respectively ), but they had very different trajec-
tion can occur in different ways. Work by Bonke3 tories while meeting similar kinds of problems,
suggests that innovation dynamics can be stied not untainted by xenophobia.

responsibility responsibility
of the state of the actors

performance
France regulation

Great Britain
Spain
The Netherlands

prescriptive
Germany regulation

Fig. 2. Construction technology regulatory systems (source: adapted from Bazin, 1993).

271
WINCH

The arrival of steel framing in the UK was inspired cular client needs within the innovation super-
by the rebuilding of Chicago after the Žre of 1871 structure, but it was systems integrators in the
which resulted in the Žrst completely framed shape of principal contractors who championed
building in 1885. The Žrst steel frame building in the innovation, particularly in the case of rein-
the UK was the Ritz (1905), but because the forced concrete. Resistance to change came from
London building regulations speciŽed load- elsewhere within the innovation superstructure.
bearing masonry, its walls remained of load- The design systems integrators displayed vir-
bearing thickness. It was Selfridge’ s (1909) that tually no interest in the new techniques, nor did
Žrst fully exploited steel framing in London. the potential suppliers of components, the steel
Inspired by the American model, Selfridge, a and cement industries. The regulatory system is
Chicago store owner, proposed to develop a new often cited as a major source of problems, but, at
type of department store with large sales oors, least outside London, these can be exaggerated,
and extensive display windows, thus steel framing and they did start to respond after 1909 to
was a product enhancing innovation. Selfridge demands from clients for new building types.
commissioned the Chicago architectural practice Crucially, as Fig. 2 shows, it is the character of
of Burnham to design his proposed store in the regulations that matters, and this is a func-
London’ s Oxford Street. The structure was en- tion of the purposes which they are intended to
gineered by the chief engineer of the principal serve .
contractor, Waring White, which had built the
Ritz . He was Bylander, a Swede who had learnt It is the way in which the professional institu-
his trade in Germany and America. A major task tions carried out their brokering role that slowed
of the British architect appointed to adapt Burn- the innovation process, and it was this brokering
ham’ s designs to the British context was to lobby that provided the basis for the new regulations.
the authorities for waivers to the building regula- The Royal Institute of British Architects’ (RIBA)
tions to allow the store to be built. As a result of strategy with respect to both technologies was to
these efforts, and those of other Oxford Street control the activities of principal and trade
stores keen to compete with Selfridge’ s, the contractors by developing detailed prescriptive
regulations started the slow process of change in speciŽcations, although it possessed little compe-
1909. tence in either of the new technologies. This was
done without liaison with the Institution of Civil
Reinforced concrete arrived by a rather different Engineers (ICE), who dealt with the new tech-
route as a contractor promoted innovation that nologies separately and reluctantly.4 Most nota-
met needs for large, heavily loaded industrial bly, neither included in their deliberations the
buildings where Žre was a particular hazard and principal contractors that knew most about the
aesthetic considerations were minimal. It appears new technologies, Mouchel and Waring White.5
to have been a process improving innovation, In particular, the RIBA resisted the use of
allowing the more efŽcient construction of an patented reinforced concrete systems because it
existing building type. Reinforced concrete tech- meant that they lost control over the technology.
nology differed from steel framing in that it was It was the recommendations of the RIBA which
developed as competing patent systems, the formed the basis of the new regulations from
leading one being that of Hennebique. As part 1909 onwards. Thus the dynamics of the British
of its strategy of overseas expansion, Hennebique institutional structure in construction resulted in
secured the contract for the Žrst fully framed the shift from load-bearing masonry to structural
reinforced concrete building in the UK ­ a framing lagging that in other industrial countries
provender mill in Swansea (1897) ­ and ap- by 20 years or more.
pointed Mouchel as their UK agent. Mouchel and
Partners licensed the system to trained contrac-
tors who paid royalties in return for the working
drawings for a building. Although there were Construction innovation processes
1073 Hennebique system buildings in the UK by
1911, none of these were in London during this The previous sections have suggested a way of
period due to the building regulations. understanding the structural features of the
construction industry that provide the context of
In both cases, innovation was driven by parti- the innovation process, but it is at the Žrm level

272
 ZEPHYRS OF CREATIVE DESTRUCTION

that decisions to adopt new ideas take place, However, unlike many other industries, innova-
thereby moving them into good currency. The tions in construction are, typically , not imple-
term ’ Žrm’ here encompasses both contracting mented within the Žrm itself, but on the projects
organizations and professional practices. The upon which the Žrm is engaged ­ adoption
literature on innovation processes is both exten- decisions by Žrms have to be implemented on
sive and inconclusive (Wolfe, 1994); this section projects. These projects are collaborative engage-
will start from Žrst principles, and propose a ments with other Žrms within the project
model of innovation processes in construction, coalition , and so almost all innovations in con-
placing the deŽning feature of construction, in struction have to be negotiated with one or more
common with other complex systems industries, actors within the project coalition. An individual
as a project-orientated industry, at its heart. This Žrm’ s ability to do this will be strongly inu-
is presented in Fig. 3, which shows the four enced by its role within the industry as deŽned
processes to be managed for successful innovation in Fig. 1.
in construction.
The projects upon which Žrms are engaged offer
The innovation literature typically identiŽes two another, internal, source of new ideas ­ problem-
basic processes ­ diffusion and implementation solving on projects (Slaughter, 1993). Groa k
(e.g. Tornatzky and Fleischer, 1990, part III). (1992) argues that this process is a fully blown
Rogers (1995) is the standard work on the former , alternative to formal R&D through the activities
while Winch (1994) provides an extensive review of ’ researcher-practitioners ’. Construction projects
of the latter. The interface between the two involve considerable problem-solving as the
processes is the decision to adopt a new idea, general repertoire of technologies and techniques
usually following the perception of a perform- is adapted and applied to meet the speciŽc
ance gap in relation to competitors. Indeed, client’ s needs in interaction with the constraints
measures of the rate of diffusion of a new idea of the site. For problem-solving to become
are, in essence, measures of the number of innovation , the solutions reached for the parti-
adoption decisions within a deŽned population cular problem faced on the project must be
over a particular period of time. Once adopted, learned, codiŽed, and applied to future projects
the innovation has to be installed and commis- ­ knowledge that remains tacit is difŽcult to
sioned so that it achieves technical success (i.e. manage into good currency. Thus the model of
works as speciŽed), and consolidated within the construction innovation proposed here has two
organization so that it yields performance beneŽts distinctive moments ­ a top-down moment of
to the business (i.e. justiŽes the investment) adoption=implementation , and a bottom-up mo-
(Winch, 1994, chap. 10). The new ideas which ment of problem solving=learning which, a
are diffused and implemented may be the out- contingency approach would suggest, need to
come of formal R&D processes , transferred from be managed in different ways. New ideas can
abroad or other sectors, or copied from leading either be adopted by Žrms and implemented on
innovators in the sector ­ whatever the source, it projects, or result from problem-solving on
is , by deŽnition, external to the innovating Žrm, projects and be learned by Žrms. Both are, a
and the ow of new ideas modelled by Gann et priori, as important as each other in the construc-
al. (1992; Gann, 1997) applies. tion innovation process.

adoption
environment firm

implementation learning

project environment
problem
solving

Fig. 3. A model of construction innovation processes.

273
WINCH

The management of innovation in tended to stie innovation (Thompson, 1968,

construction chap. 13). Similarly, innovative design solutions
which save construction costs on site reduce the
In developing his general management perspec- returns to designers paid on a fee basis rather
tive on innovation, Van de Ven (1986) identiŽed than increasing them ­ as Bowley put it, ’ the
four central problems ­ the management of architect is in the position of a commission sales-
attention; the process problem of managing ideas man whose interest is to get as big a value of
into good currency; the structural problem of turnover as possible’ (1966, p. 356). The lack of
managing part ­ whole relationships; and the such incentives in British construction has led to
strategic problem of institutional leadership. It most innovations being what Bowley (1966)
will be useful to discuss some of the implications called ersatz ­ innovations that only took place
of the ideas sketched above under these headings. because preferred solutions were no longer avail-
able due to external factors.

Incentives for innovation in construction cannot

The management of attention
be improved without the development of a gain-
Business life is complex and dynamic; humans sharing approach, where rewards are split be-
have limited cognitive capability. They tend to tween clients and the actors in the project
focus upon those things that solve their most coalition . The shift from competitive tendering
immediate problems and the only way to en- to partnering provides one of the most important
courage innovation is to give it sufŽcient salience. opportunities for moving towards such an ap-
Exhortation is not enough; the incentives which proach. Those in a position to innovate need to
motivate actors in particular directions need to be rewarded for taking such risks. If they are so
favour innovation. Thus the route to the success- rewarded, they will have incentives both to adopt
ful management of attention is through the new ideas from outside the Žrm, and to capture
incentive structures that inform decision-making; the learning from problem solving to propose
the corollary is that if incentive structures do not better ways of doing things to the client.
favour innovation, then innovation is unlikely to
take place . As Ive (1996) argues, construction
innovation depends upon the coincidence of the
The managing ideas into good currency
means, motive and opportunity to innovate. Thus
the way in which the RIBA brokered innovations Perhaps the most consistent Žnding of research on
in structural framing reduced the means of innovation is that innovations need champions ­
principal contractors to innovate through pre- the importance of technically competent cham-
scriptive regulations; stied their motives because pions in construction is well demonstrated by
they could not appropriate the returns to their Nam and Tatum (1997). Ideas are carried by
innovations through patenting; and restricted the people, and ideas are the rallying point around
opportunity to industrial building types which which collective action mobilizes. However, while
had no architectural input. such champions in complex systems industries
can come from any part of the industry, and in
The essence of incentive structures that favour construction typically come from component sup-
innovation is the appropriation of the rewards of pliers (Pries and Janszen, 1995), the case studies of
innovation by those that take the risks of Nam and Tatum (1997) show that the role of the
innovation ­ this is the logic behind patents for principal architect=engineer and principal con-
instance. Most importantly, this means that any tractor is central in all innovations. Unless the
innovation where the client appropriates all the systems integrator is convinced of the merits of
returns is unlikely to happen. Many principal the new idea, and has the skills to incorporate it
quantity surveyors consider it good professional into the system as a whole, change is likely to be
practice to ask all tenderers to price variations slow . As shown in Fig. 1, the systems integrator is
proposed by one contractor at tender stage in the at the interface between the innovation super-
hope of obtaining an even better price.6 The structure and the innovation substructure ­ new
disincentive effect is clear, and even the Royal ideas are proposed within the latter and accepted
Institution of Chartered Surveyors’ ofŽcial history within the former, mediated by the systems
argues that quantity surveying practice has integrator.

274
 ZEPHYRS OF CREATIVE DESTRUCTION

According to Miller and his colleagues (1995, subsystem optimization at the cost of whole
p. 381), the systems integrator has four main system suboptimization are quashed.
functions:
The founding purpose of the Institution of Civil
· the skills to integrate interdependent compo- Engineers (ICE) ­ the international model for the
nents into a coherent whole; professional organization of technical expertise ­
· detailed knowledge of client requirements; was ’ facilitating the acquirement of knowledge
requisite in their profession and for promoting
· knowledge of the rules and regulations Mechanical Philosophy’ (cited Watson, 1988, p.
governing the industry; 12). This model was replicated in industry after
· the competence to use and develop the industry and country after country during the
mathematical models of aircraft performance 19th century. In nonprofessionalized systems
obtained from the aircraft builders. such as the French, the corps acted as both
broker and champion of new ideas, and the Ecole
While the last of these may be regarded as sector des Ponts reached the zenith of its international
speciŽc , the other three would appear to apply in inuence in construction technology during the
all complex systems industries. Žrst half of the 19th century (Picon , 1992, p. 465).
Campagnac and Winch (1997) provide a more
From this perspective, it can be suggested that detailed comparison of these two modes of
one of the major factors reducing the rate of organizing technical expertise.
innovation in construction is that the systems
integrator role is split between two very different The problem, at least in British construction, is
actors ­ the principal contractor and the princi- that this broker role is itself spread amongst a
pal architect=engineer. This means, a priori, that number of professional bodies. Abbott’ s (1988)
the mediating and championing roles essential to analysis of the ’ system of professions’ as a
successful innovations are less likely to be carried competition for jurisdictional advantage can be
out effectively. A second problem is that while well applied to the British construction profes-
principal architect=engineers typically display sions . The ICE aggressively defended its position
competence in the regulatory framework and against competition from the breakaway institu-
client requirements, they often do not have the tions of mechanical and electrical engineers
skills to integrate the subsystems into a total throughout the 19th century, successfully pre-
system. In this, they are not well complemented venting them from obtaining royal charters until
by the principal contractor whose integration the 1920s (Buchanan, 1989, chap. 5). The cur-
capabilities are typically restricted to the manage- rently competing claims from different profes-
rial rather than technical level. sions to be the leader of the construction team
from architects, quantity surveyors and chartered
builders are part of the same process. It remains
to be seen whether the Construction Industry
The management of part-whole relationships
Council, founded in 1988 with a leading objective
Innovations on complex product systems are to ’ develop, encourage and co-ordinate . . . the
inherently interactive with the rest of the system industry’ s research programme’ (cited Watts,
­ innovating within the parts while losing sight of 1997), can develop enough authority to take over
the whole is inherently dysfunctional ­ ’ impec- the brokering role itself.
cable micrologic often creates macrononsense and
vice versa’ (Van de Ven, 1986, p. 598). One of the One of the distinctive features of the construction
great strengths of the analysis of Miller and his industry in many countries is the role of
colleagues is the emphasis placed upon the role of government funded national construction re-
independent brokers in the innovation process. In search organizations who also play a vital
the ight simulation industry this is played by the brokering role, and, it can be suggested, the
(British) Royal Aeronautical Society at an interna- (British) Building Research Establishment has
tional level. It is the function of these actors in the done much to Žll the vacuum left by the
innovation superstructure to broker new ideas so competing professional institutions. Many of
that those that improve the parts and complement these establishments around the world are
the whole go forward and those that favour currently being privatized (Seaden, 1997), and it

275
WINCH

remains to be seen whether they will be able to of the innovation superstructure ­ the regulatory
retain this vital brokering role once they are environment on the one hand, and the profes-
obliged to compete directly with other actors in sional bodies, research establishments and uni-
the system. In many of the most dynamic and versities on the other.
innovative industries, the universities have
played the brokering role by providing a space While demanding clients and other elements in
where the merits of competing technologies can the innovation superstructure play a vital role in
be evaluated and disseminated to enable innova- stimulating the search for new ideas, it is the
tion, as well as their more traditional role as the innovation infrastructure that has the responsi-
source of many new ideas. Indeed, the evidence bility of managing them into good currency.
from the information technology and biotechnol- Here, the capacity for organizational learning is
ogy industries suggests that intensive interaction critical (CIOB, 1995), both in the top-down and
with world class research universities is a pre- bottom-up modes. In the top-down mode, the
requisite for a technologically dynamic industry. processes of adoption and implementation are
essentially iterative learning cycles as the features
of the new idea and the existing organizational
Institutional leadership and the innovation
context are mutually adjusted (Winch, 1994,
context
chap. 10). In the bottom-up mode, new ideas
The particular features of the construction inno- generated through problem-solving need to be
vation context were identiŽed in the previous learned by the organization so that they can be
section , and more attention needs to be given to managed into good currency on future projects.
’ creating these intra- and extra-organizational The capacity of the Žrm to learn is, arguably, the
infrastructures in which innovation can ourish’ most important determinant of its ability to
(Van de Ven, 1986, p. 601). One of the strongest innovate on projects.
themes running through the recent innovation
research is the role of tough customers in good One the most notable features of the contempor-
designs (Gardiner and Rothwell, 1985); customers ary organization of the UK industry is the
are one of the most important sources of innova- relatively low ability of Žrms to learn (CIOB,
tions (von Hippel, 1988). 1995). This is particularly true of the principal
contractor, which plays a key systems integration
Current policy in the UK identiŽes the experi- role , yet stresses managerial rather than technical
enced client as the main institutional leader in integration capabilities. In Japanese Žrms such as
stimulating construction innovation, yet there Kajima, on the other hand, the engineering func-
must be doubts regarding clients’ ability to play tion acts as the focus of organizational learning
this role. Nam and Tatum (1997) show that the as it codiŽes the learning from problem-solving
client needs to be technically competent in order on projects and implements these innovations
to understand innovative proposals from systems on future projects, while passing up problems
integrators , and hence to take the risk of that cannot be solved at the project level to
innovating. From this perspective, a particular the R&D functions for further work. A compari-
weakness of the British system is that the single son of a British management contractor with a
most important client ­ the state ­ has no Japanese general contractor (Ota, 1990) found
equivalent of the US Army Corps of Engineers, that the principal difference between the two
the Staatsbauamt, or the Corps des Ponts et was that learning from problem solving was
Chausse es to assess innovative proposals, while captured by the Japanese Žrm, but remained tacit
many local authorities, private sector clients, and and rested with individuals in the British one.7
privatized utilities have been outsourcing their This creates a highly uid market for ’ star’
architectural and engineering functions. The construction managers between indistinguishable
ability of the client to assess and even propose Žrms, rather than generating distinctive compe-
engineering solutions to the principal archi- tencies within Žrms for which client will pay
tect=engineer and principal contractor played an premia.
important role in the successful Glaxo project
(Winch et al., 1998). The analysis of complex The current situation in British construction
systems industries also suggests that more atten- whereby the downstream systems integrator does
tion needs to be given to the two other elements no actual site work seriously disrupts the

276
 ZEPHYRS OF CREATIVE DESTRUCTION

problem solving=organisational learning dy- tion superstructure and infrastructure in the

namic. In many industries, operatives are one industry moves closer to enabling the dynamic
of the main sources of incremental innovation, summarized in the epigraph to this concluding
often captured through suggestions schemes and section . Schumpeter was wrong about the pro-
quality circles, yet these groups are completely mise of volume production for the construction
externalized from the Žrm in construction industry, and Miller and his colleagues provide a
through labour-only subcontracting and casual trenchant critique of the Schumpeterian model of
employment. Whatever problem solving goes on innovation to explain why he was wrong.
is not learned by the Žrm. Similarly trade
contractors selected on a competitive tendering Secondly this paper has presented a two-moment
basis have absolutely no incentive to share model of innovation processes at the Žrm level as
learning with the systems integrator. For partner- a conceptual framework deŽning the manage-
ing to fully realize its potential in construction it ment problem in construction innovation. The
must be extended to the principal contractor=- distinctive project organization of the industry
trade contractor relationship. means that innovation consists of both an
adoption=implementation dynamic and a pro-
blem solving=learning dynamic. Both of them
Some concluding thoughts rely on the capacity for organizational learning
by the Žrm, and some reasons were given why
The need to co-ordinate innovation in complex this may be relatively low in construction. While
systems industries such as ight simulation differences do exist between principal and
requires a complex institutional superstructure: specialist consultants on the one hand, and
principal and trade contractors on the other, the
New technology proposals are channelled model is intended to apply to both types of Žrm
through professional bodies such as the equally.
Royal Aeronautical Society. Acceptance test
guides are established by regulators who The poor state of knowledge generally in
then specify approval requirements and innovation process research (Wolfe, 1994) favours
validate tests during and after the develop- a case study approach which will allow theory
ment of a [system]. After contracting, trust building (Yin, 1989). Here, most authorities
and reciprocity are necessary between appear to agree, the object of inquiry is most
buyers and sellers. Because many uncertain- appropriately the innovation itself; thus we need
ties have to be resolved during the process case studies of the trajectory of particular
of innovation [systems] cannot be purchased innovations , identifying who generates new ideas
as arm’ s length market transactions in the and how they are managed into good currency,
standard model. Instead, intense relational collected within a consistent conceptual
transactions develop, allowing for constant framework.8 This would allow the development
information exchange and regular inter- of concepts and measures which would allow a
action between industry participants. Con- more quantitative approach at a later stage This
tinuity of relationships is valued, and paper has attempted to sketch out what such a
respected and helps deŽne the competence conceptual framework might look like at both the
of the partners. Innovation . . . unfolds with- structural (Fig. 1) and process (Fig. 3) levels .
in a set of governing institutions where . . . However, much more work is required before we
co-operation and competition co-exist can really get a grip on the sources and
(Miller et al., 1995, p. 384). applications of new ideas in the construction
industry. I hope that this article will stimulate
The argument of this paper has had two main that work.
thrusts. First, that constructed products are com-
plex product systems, and that the construction
industry is a complex systems industry. It follows
that the characterization of the innovation process
References
in ight simulation is also applicable to construc- Abbott, A. (1988) The System of Professions, University of
tion. It follows that the rate of innovation in Chicago Press, Chicago.
construction will only improve when the innova- Allen, S.G. (1985) Why construction productivity is

277
WINCH

declining, The Review of Economics and Statistics, 67, of constructed products and resulting limitations of
661­ 69. construction technology, Construction Management
Bazin, M. (1993) Analyse strate gique en science et and Economics, 6, 133 ­ 48.
technologie, Cahiers du CSTB, 2643. Nam, C.H. and Tatum, C.B. (1997) Leaders and
Bowley, M. (1966) The British Building Industry, Cam- champions for construction innovation, Construction
bridge University Press, London. Management and Economics, 15(3), 259­ 70.
Buchanan, R.A. (1989) The Engineers: a History of the Ota, K. (1990) Appraisals of management contractors in
Engineering Profession in Britain 1750 ­ 1914, Jessica the British construction industry MSc thesis ­ UCL,
Kingsley Publishers, London . Building Economics and Management, London.
Campagnac , E. and Winch, G. (1997) The social regula- Picon , A. (1992) L’Invention de l’Inge nieur Moderne,
tion of technical expertise: the corps and profession, Presses Ponts et Chausse es, Paris.
in France and Great Britain. In: R. Whitley and P.H. Pries, F. and Janszen, F. (1995) Innovation in the
Kristensen (eds): Governance at Work, Oxford Uni- construction industry: the dominant role of the
versity Press, Oxford, pp. 86­ 103. environment, Construction Management and Econom-
Chartered Institute of Building (1995) Time for Real ics, 13(1), 43­ 51.
Improvement, Chartered Institute of Building, Ascot. Quigley, J. (1982) Residential construction, in R.R.
Connaughton, J., Jarrett, N. and Shove, E. (1995) Nelson (ed.) Government and Technical Progress,
Innovation in the Cladding Industry, Department of Pergamon Press, New York, pp. 361­ 410.
the Environment, London. Rasmussen, M. and Shove, E. (nd) Concrete conclusions:
Courtney, R. (1997) Building Research Establishment ­ report of pilot study of innovation and change in
past, present and future, UK, Building Research and the in-situ concrete industry, mimeo, Lancaster
Information, 25(5), 285­ 91. University.
Cusack, P. (1986) Architects and the reinforced concrete Rogers, E.M. (1995) The Diffusion of Innovations (4th edn)
specialist on Britain 1905 ­ 08, Architectural History, Free Press, New York.
29, 183­ 96. Schumpeter , J.A (1976) Capitalism, Socialism and Democ-
Cusack, P. (1987) Agents of change: Hennebique, racy (5th edn) George Allen and Unwin, London.
Mouchel and ferro-concrete in Britain 1897 ­ 1908, Seaden, G. (1997) The future of national construction
Construction History, 3, 61­ 74. research organizations, Building Research and Infor-
Gann, D. (1997) Should governments fund construction mation, 25(5), 250­ 56.
research? Building Research and Information, 25(5), Slaughter, S. (1993) Innovation and learning during
257­ 67. implementation: a comparison of user and manu-
Gann, D., Matthews, M., Patel, P. and Simmonds, P. facturer innovations, Research Policy, 22, 81­ 95.
(1992) Construction R & D: Analysis Of Private And Thompson, F.M.L. (1968) Chartered Surveyors: The Growth
Public Sector Funding of Research and Development in of a Profession, Routledge and Kegan Paul, London.
the UK Construction Sector, Department of the Tornatzky, L.G. and Fleischer M. (1990) The Processes
Environment, London. of Technological Innovation, Lexington Books,
Gardiner, P. and Rothwell, R. (1985) Tough customers: Lexington.
good designs, Design Studies, 6. Van de Ven, A.H. (1986) Central problems in the
Groa k, S. (1992) The Idea of Building, Spon, London. management of innovation, Management Science,
Ive, G. (1996) The client and the construction process: 32(5), 570­ 607.
the Latham report in context, in S. Gruneberg (ed.) Van de Ven, A.H., Angle, H.L. and Poole , M.S. (eds)
Responding to Latham: the Views of the Construction (1989) Research on the Management of Innovation,
Team, Chartered Institute of Building, Ascot, pp. Ballinger, Grand Rapids.
37­ 43. von Hippel, E. (1988) The Sources of Innovation, Oxford
Law, J. and Callon, M. (1992) The life and death of an University Press, New York.
aircraft: a network analysis of technical change, in: Watson, G. (1988) The Civils, Thomas Telford, London.
W.E. Bijker and J. Law (eds) Shaping Technology= Watts, G. (1997) The National Centre for Construction in
Building Society, MIT Press, Cambridge, pp. 21­ 52. the UK, Building Research and Information, 25(5),
Lawrence, J.C. (1990) Steel frame architecture versus the 279­ 84.
London Building Regulations: Selfridges, the Ritz, Winch, G.M. (1994) Managing Production: Engineering
and American technology, Construction History, 6, Change and Stability, Oxford University Press,
23­ 46. Oxford.
Margirier, G. (1988) Le secteur du baà timent et des Winch, G.M. and Campagnac, E. (1995) The organization
travaux publics dans la crise: comparison France, of building projects: an Anglo-French comparison,
RFA, Italie, Royaume-Uni, Actes du Colloque: Europe Construction Management and Economics, 13(1), 3­ 14.
et Chantiers, Plan Construction et Architecture, Winch, G.M., Usmani, A. and Edkins, A. (1998) Towards
Paris. total project quality: a gap analysis approach,
Miles, J. (1996) Where is the Henry Ford of Future Housing Construction Management and Economics, 16(2),
Systems? Royal Academy of Engineering, London. 193­ 207.
Miller, R., Hobday, M., Leroux-Demers, T. and Olleros, Wolfe, R.A. (1994) Organizational innovation: review,
X. (1995) Innovation in complex systems industries: critique and suggested research directions, Journal
the case of ight simulation, Industrial and Corporate of Management Studies, 31(3), 405-31.
Change, 4(2), 363­ 400. Yin, R.K. (1989) Case Study Research (2nd edn), Sage,
Nam, C.H. and Tatum, C.B. (1988) Major characteristics Beverly Hills.

278
 ZEPHYRS OF CREATIVE DESTRUCTION

Endnotes ICE was believed to be ignoring the potential of the

new framing technologies (Buchanan, 1989).
1 5
The proceedings of this UK Engineering and Physical The neophyte Concrete Society was more open, and
Sciences Research Council funded seminar on Hu- both Bylander and principals from Mouchel were
man Resources for Construction Innovation are active in its governance.
6
available as Construction Productivity Network This information was obtained during an interview (29
Workshop Report No. S196, 1997. Many of the October 1992) with an employer’ s agent as part of
ideas in this article were Žrst introduced at that Želdwork comparing the organization of social
meeting, and I am very grateful to all the housing projects in France and Great Britain ­ see
participants in the meeting for the stimulating Winch and Campagnac (1995) for the results.
7
quality of debate over the two days. A similar point was made in a rather different manner
2
I am grateful to David Gann for this information. More by a speaker at a Construction Productivity Seminar
generally, this article has been much enhanced by Construction Innovation; Learning from Abroad (Report
the constructive criticism of David Gann, although No 811L 1998) who presented the results of
he, of course, bears no responsibility for the Žnal benchmarking his construction Žrm against 1000
shape of the argument. European manufacturing Žrms, noting that it rated
3
Presentation at le Groupe Bagnolet seminar, Steinbach relatively high on the skills of its people, but
1997. relatively low on its management processes.
4 8
This is well illustrated by the establishment of the This is the strategy deployed by the Minnesota
Concrete Society in 1908, which became the Institu- Innovation Research Program (Van de Ven, et al.,
tion of Structural Engineers in 1922, because the 1989) with impressive results.

279