strategy, governance and industrial organization: historical perspectives on … · 2007-07-12 ·...

43
Strategy, Governance and Industrial Organization: Historical Perspectives on the Dynamics of Industrial Clustering in England. ANDREW POPP Lecturer in Business History School of Management Royal Holloway University of London Egham Surrey TW20 0EX [email protected] STEVE TOMS Professor of Accounting and Business History University of Nottingham Business School Jubilee Campus Wollaton Rd Nottingham NG8 1BB Tel: + 44-115-951-5276 Fax: + 44-115-956-6667 [email protected] JOHN WILSON Reader in Business History University of Nottingham Business School Jubilee Campus Wollaton Rd Nottingham NG8 1BB [email protected]

Upload: others

Post on 03-Apr-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Strategy, Governance and Industrial Organization: Historical Perspectives on the Dynamics of Industrial Clustering in England.

ANDREW POPP Lecturer in Business History

School of Management Royal Holloway

University of London Egham Surrey

TW20 0EX [email protected]

STEVE TOMS Professor of Accounting and Business History

University of Nottingham Business School Jubilee Campus

Wollaton Rd Nottingham NG8 1BB

Tel: + 44-115-951-5276 Fax: + 44-115-956-6667

[email protected]

JOHN WILSON Reader in Business History

University of Nottingham Business School Jubilee Campus

Wollaton Rd Nottingham NG8 1BB

[email protected]

Page 2: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Strategy, governance and industrial organization: Historical perspectives on the

dynamics of industrial clustering in England

Abstract. The paper integrates two important areas of literature, the Chandler model of

corporate hierarchy and the discrete alternative models of small firm flexible

specialization, in order to explore long-run processes of structural and strategic

change. To achieve synthesis, resource based view (RBV) and resource dependency

theories are combined to explain the evolution of different industry structures. A

typology is developed delimited by the opposite cases of classic industrial district and

managerial hierarchies, as well as hybrid alternatives to each. Empirical cases are

offered, showing historical examples of conditions favouring each and also illustrate

the sufficient and necessary conditions underpinning transitions from one mode to

another. The paper has important implications for conventional interpretations and for

the main areas of theory hitherto considered separately.

1

Page 3: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

1. Introduction

What forces shape the processes of structural change in firms and industries? How do

we account for the simultaneous presence of giant corporations and dynamic clusters

of smaller firms? What are the processes that mediate co-operation, competition and

accountability within these differing methods of industrial organization and transition

between them?

The ambition of this paper is to answer these questions with reference to the

industrial history of the United Kingdom from the industrial revolution to the present

day, drawing together two important strands of literature that have had wide influence

and addressing important gaps in each, in order to construct an integrated explanation

of industrial and corporate change. That such an ambitious agenda should be

addressed is justified by the discrete evolution of these literatures. The first and

perhaps most dominant area is the Chandler (1962, 1977, 1990) model of strategy,

structure and business evolution. This first paradigm of scale-driven, hierarchical

corporate structures is diametrically opposed by the second, namely, studies of

industrial clusters and districts.

Theoretically and empirically these literatures apparently represent two distinct

schools of inquiry with regard to economic growth and development. Each draws tight

correlations between corporate structures and forms and alternative models of

capitalistic accumulation; namely, mass-production and flexible specialization.

Nonetheless, both, whether explicitly or not, portray these models as dichotomous and

mutually exclusive. Sabel and Zeitlin admit as much in the title of their seminal article

of 1985, ‘historical alternatives to mass production’. More recently, Scranton cast his

Endless Novelty as being about an ‘“other side” of the Second Industrial Revolution’

that created ‘technological and organizational transformations distinct from, but

2

Page 4: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

comparably significant to, the creation of rountinized assembly, bureaucratic

management, and oligopolistic competition’ (1997, 3).1 Thus, the Chandlerian model,

driven by the internalization of scale and scope economies and embodied in the large

multidivisional firms of managerial capitalism, is most commonly associated with the

new industries of the Second Industrial Revolution and their characteristics of

standardization and throughput. On the other hand, clustering is equated with a small

firm economy of flexibility, reliance on sharing of resources external to the individual

firm and more personal regimes of ownership and management. Trajectories, in

structural and strategic terms, are assumed to be quite different.

Between these competing paradigms, there is a third area of literature; that which

promotes the notion of heterarchy as an alternative to the hierarchical model of

corporate organization. Both the heterarchy concept and the clustering literature

represent a challenge to the dominant Chandlerian paradigm. However, unlike the

proponents of the small firm economy of flexible production, the heterarchical

challenge was motivated by a growing awareness that the hierarchical structure was

not best suited to the problems faced by the multinational corporation (MNC). In the

heterarchy model, MNC subsidiaries are recognised as repositories of local

knowledge and resources, which through lateral linkages can become sources of

competitive advantage (Bartlett and Ghoshal, 1989, Hedlund, 1986, Pralahad and

Doz, 1981). Broadening the perspective on heterarchies, Rugman and Fenton (2000)

have conceptualized MNCs as ‘flagship firms’, offering a model of regional business

development that links MNCs, suppliers, customers, competitors and non-business

organizations such as government and regulatory bodies. These authorities claim that

the advantages accruing to heterarchical organizations are quantifiable in transaction

cost terms, but with the corollary that strategic advantage is further promoted by

3

Page 5: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

geographic dispersal, at least relative to hierarchical organizations (Hedlund, 1986,

1994). However, some of the advantages in heterarchies are similar to those also

present in industrial clusters, with the important exception of geographical location.

Heterarchies are geographically dispersed, but with concentrated ownership, whereas

clusters are geographically concentrated but with dispersed ownership. Both are

organized, however, on the basis of lateral linkages. The important difference is

therefore the role of ownership, and in particular governance structure.

More recently, Pettigrew and Fenton (2000) have suggested ways in which the

apparently conflicting nature of hierarchical and heterarchical forms and strategies

may be reconciled. Based on an extensive survey of European business, they

concluded that while outsourcing and the formation of long-term strategic alliances

had increased over the course of the 1990s, internal reorganization had also proceeded

apace. Thus, the hierarchical and heterarchical approaches may be regarded as

complementary, given the imperative need to obviate the risks associated with

collaborating with competitors.

At the same time, it is also clear that governance and accountability, also neglected

by Chandler (Toms and Wilson, 2003), may have had an important role to play in

fashioning internal reorganizations, in that external pressure by shareholders on

managements has mounted to such an extent that those managements have been

obliged to introduce much more rigorous control mechanisms. The heterarchy concept

has been highlighted in order to demonstrate its potential to impinge on both the

Chandlerian and clustering/flexible specialization paradigms, thus reinforcing the

argument that there is both a need and the potential to reconcile two paradigms that

have largely been construed as mutually exclusive until now.

4

Page 6: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Such a reconciliation would mesh with empirical observation of how, in the past,

‘ideal-typical’ Chandlerian firms have emerged from ‘classic’ Marshallian industrial

districts and how, more recently, the Chandlerian firm has begun to ‘splinter’. As the

brief review above suggests, it is perhaps through consideration of issues of

governance and accountability that such a reconciliation may be achieved. The

objective of this paper is, then, to use the mechanisms of governance and

accountability, rooted in resource dependency theory, to explain contrasting outcomes

in hierarchies, clusters and hybrid organizational forms and processes of industrial

transformation.

Furthermore, developments in corporate hierarchies, industrial clusters and MNCs

are also related by historical process. At the empirical level, analysis should promote

explanations of divergence in models of industrial organization in different periods,

different countries and different industries. In order to do this, the paper develops a

theoretical model to explain the industry structure and dynamics according to industry

and firm resource characteristics and the extent of resource dependency. Both the

resource based view (RBV) of the firm and the resource dependency perspectives are

well established in the strategic management literature. However, there have been

relatively few attempts to synthesize them, least of all in a dynamic historical

perspective (Toms and Filatotchev, 2003). Any such synthesis will therefore have an

important potential contribution to the literature explaining organizational

developments and strategic decision-making.

A further contribution arises because any model that integrates the notion of

flexible specialization must also address the important empirical and theoretical gaps

in much research into industrial clusters. It is acknowledged that the literature on

clustering is emergent and that there is an extensive research agenda (Breschi and

5

Page 7: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Malerba, 2001, 830). Institutional research when applied to networks and clusters has

tended to reinforce the separate consideration of such groupings, for example by

concentrating on the degree of mimetic isomorphism within industries or regional

clusters (Beckman and Haunschild 2002; Westphal et al. 2001). Such processes stress

the degree of path dependency also proposed by other researchers, including those

developing life-cycle models of clustering (Best, 2001, Swann et al, 1998). These

approaches not only allow for the significance of structural factors, but also recognise

that clustering may generate vicious as well as virtuous forces. Clusters become

‘saturated’ or ‘congested’ as the level of firm entries grows, such that ‘beyond a

certain point congestion limits the attractiveness of an existing cluster for entry’

(Swann et al, 1998: 60). As attractiveness declines and negative externalities begin to

outweigh the positive, the innovatory dynamic is lost.

However, in privileging technological drivers, both of these models neglect other

equally, perhaps more, important factors in the dynamics of clustering. Best (2001)

argues that the technological dynamism of the ‘entrepreneurial firm’ is translated into

technological dynamism at the level of the cluster or region via the ‘collective

entrepreneurial firm’, defined as a ‘self-organizing agent for change composed of

networked groups of mutually adjusting enterprises’ (Best, 2001: 83). In other words,

cluster dynamics are also dependent on governance forms and arrangements.

Externalization, the most distinctive characteristic of clustering, is thus dependent on

the co-ordination of interdependent but institutionally dispersed resources. In terms of

theory, the clustering literature’s greatest weakness perhaps relates to the governance

arrangements developed to meet this challenge.

In contrast, Porter (1990) argues that the principal task in building national

competitiveness is the organization of ‘clusters’ or strong ‘diamonds’ based on

6

Page 8: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

appropriate factor conditions, demand conditions, related industries and firm level

strategy, structure and rivalry. Whilst it is widely agreed that trade patterns are the

outcome of scale economies, Porter’s model is concerned only with scale economies

that can be internalized by the firm, (Davies and Ellis, 2000: 1191, 1199) and not with

the external economies that underpin much of the research into industrial clusters.

Furthermore, where firms can locate production according to the availability of scale

economies overseas, it is not clear what firm or which nation gains in terms of

competitive advantage (Davies and Ellis, 2000, Reich, 1990, 1991).2 Assessing the

performance impact of hierarchies, heterarchies and clusters is therefore highly

problematic and in any case neglects how the benefits of each model are distributed.

In summary, the principal outstanding task for the clustering research agenda is to

consider resource location and the sources of internal and external economies of scale

and scope and to incorporate governance and accountability perspectives. This is

consistent with the broader ambition of the paper, which is to synthesize the RBV

with resource dependency to explain the historical development of different industry

structures. In so doing, the paper also aims to fulfil its fundamental aim of linking the

hierarchical Chandlerian paradigm with the alternatives of MNC heterarchy and small

firm flexible specialization, reconceptualizing these not as mutually exclusive but as

existing on complex continua. If successful, there are important implications for the

Chandlerian view, which has retained much of its influence as a dominant paradigm

in both management theory and business history (Whittington et al 1999; Whittington

and Mayer, 2000), as well as for the institutional and neo-classical views of industrial

organization.

The remainder of the paper is structured as follows. Section 2 examines the

literatures on the RBV and resource dependency and articulates a synthesis between

7

Page 9: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

them in order to explain industry organization characteristics. The synthesis leads to

an analytical model, presented in section 3, which identifies four possible ‘pure’ cases

of industrial organization and identifies the processes of transition within the

typology. Section 4 introduces empirical evidence to illustrate the cases outlined in

the model. Whilst these are drawn from the personal research agendas of the authors,

it is argued that they have a general relevance that substantiates our overriding claims.

Section 5 offers discussion and conclusions.

2. Resource distribution, resource dependency and industrial organization

This section aims to synthesize two literatures that assist our understanding of

industry organization. Industry organization in this context includes both structural

characteristics and governance arrangements. Governance arrangements refer to the

mechanisms used, whereby organization and network members are held accountable

to each other and to external resource providers. These definitions are used so that the

two main areas of theory to be employed, the RBV and resource dependency theory,

can be accommodated into a single model.

The RBV concentrates on difficult-to-replicate, firm-specific assets that promote

competitive advantage. Such resources might include specialized production facilities,

trade secrets and engineering experience (Teece et al 1997). They might also include

firm-specific idiosyncratic knowledge assets (Castanias and Helfat 2001). Such firm-

specific factors are traditionally considered as the major drivers of strategic change,

according to the RBV (Barney 1997). According to this view, managerial and

entrepreneurial resources drive growth and diversification (Whittington and Mayer

2000). The RBV has been extended to the competence-based theory of the firm in

8

Page 10: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

which the firm constructs capabilities through internal learning processes (Teece et al

1997; Chandler et al, 1998).

A weakness of the RBV is that it ignores resources that are non firm-specific.

However, as the literatures on clustering and networking suggest, resource sharing

arrangements may also promote competitive advantage (Arthur, 1990). As Mathews

notes, citing Silicon Valley as an example, in ‘the real economy’ firms are often

‘placed in positions of mutual dependence’, because ‘resource configurations within

the economy usually span firms’ (2003: 128, 133). As such, sharing arrangements

within clusters frequently involve the transmission of uncodified tacit knowledge,

while the configuration of non-firm specific assets can have a major impact on

strategy and structure (Swann et al.: 1998) An important reason is that innovation

‘spills over’ from the originator to benefit third party firms (Jaffe, 1989, Feldman,

1994). Dynamic external economies of scale occur where there are accumulations of

local knowledge as a result of repeated interactions along established channels

(Glaeser et al 1992).3 According to this view, clustering of firms in industrial districts,

trade associations and other networked organizations may be promoted through

sharing trade secrets and drawing on local pools of experience and skilled labour

(Amin and Thrift: 1994). These shared resources form the basis of agglomeration-

based external economies of scale (Kamien et al 1992). In summary, the RBV

requires extension to consider the location of resources, inside and outside the firm, in

order to explain the full range of possibilities for industrial organization and

associated governance arrangements.

An alternative perspective allowing such an extension is the suggestion that mass

production and flexible specialisation lie at opposite ends of a continuum. In industrial

organisation terms, this is the equivalent of the trade-off between scale economies and

9

Page 11: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

allocative efficiency (Oughton and Whittam, 1997, 6). Internalisation of production in

large-scale units create scale-related, cost-reduction benefits on the one hand, whilst

on the other there are the incentive and price reduction benefits associated with the

absence of market power. It might be added that in the former case there are few

benefits from clustering, since the necessary resources are internalised within the firm

through integration, whilst diversification requires product and market dispersion.

Only in relatively competitive and geographically proximate industries are there likely

to be external economies of scale benefits from clustering.

However, these performance benefits have been analysed without reference to the

governance of clusters, which is another important dimension potentially explaining

their dynamic development. Resource dependence theory assumes that firms respond

to external pressures, but their power relative to these pressures may be contingent on

the configuration of the resource-user/resource-provider relationship (Pfeffer 1992;

1997; Frooman 1999). Resource dependency is therefore important because it

influences the degree to which the cluster is externally monitored. Where the cluster is

relatively self-sufficient, there is little incentive to subject the cluster to the scrutiny of

external monitors. In turn, this dependency reflects the growth rate of the industry. In

rapidly expanding or evolving industries, it is entirely probable that the cluster will

require outside resources in order to finance and produce the required asset base.

Conversely, in contracting industries the reduction of dependency will have the effect

of making clusters more difficult to police from outside, whilst the cluster members

may use their internal networking arrangements to promote capacity sharing, output

and price restrictions and other related strategies.

These and similar strategic actions formulated by the cluster members are likely to

be a result of the dynamic interaction of the resource availability and resource

10

Page 12: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

dependency characteristics of the cluster. Where external stakeholders provide

resources, it is necessary to put in place arrangements for co-ordination and

monitoring to mitigate opportunistic behaviour by network members (Gulati et al

2000; Jones et al 1997). Such accountability processes emphasize information flow,

highlighting processes of socialization and offering a broader perspective than the

Chandlerian view of technology as the exploitation of associated scale and scope

economies (Casson 1997; Hamilton and Feenstra 1995; Langlois and Robertson

1995).

For those associated with what Granovetter has termed the ‘strong embeddedness’

perspective on regional business networks (1992: 5), including Piore and Sabel

(1984), Sabel and Zeitlin (1985) and Staber et al. (1996), powerful, cohesive, district-

wide governance is derived from deep implantation of actors in robust regional and

sub-regional socio-cultural structures. From this perspective, the principal outcome of

the distinctive embedded governance systems of the ‘ideal-typical industrial district’

is a creative balance between co-operation and competition, maintained by the sway

of widely-held, consensual normative value systems.

However, harmonious district governance is often seen as achieved through

networks that are formulated as manifestations of a highly reified concept of

‘community’ (Piore 1992). Piore, for example, argues that clusters ‘survive and

prosper only if the economies that are external to particular productive units are

internalized as parameters in the decisions of some higher-level organization unit’

(1992: 437). However, as he is certain only that this higher-level function cannot be

fulfilled by the market, Piore suggests that the ‘relationships … organized by

contracts’ recorded by Marshall meant that he ‘did not observe industrial districts’ in

the sense understood by contemporary network scholars (1992: 437). The repeated

11

Page 13: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

elision of governance issues in cluster studies remains a problem. As Mathews

contends, the issue remains how the shifts in resource configurations characteristic of

successful clusters ‘are accomplished, and whether they call for specific institutional

interventions, or are accomplished by the strategic actions of the actors themselves’

(2003: 135).

In clusters, then, as in networks, openness and secrecy depend on existing patterns

of control through agency and delegation (White 1992: 93). At the same time,

evolutionary economics suggest that the interaction of governance arrangements and

resource patterns may be intricately linked with and central to the development of new

and existing capabilities. Metcalfe, for example, argues that increasing emphasis is

now being placed on ‘patterns of institutedness, the nature of the rules, practices and

procedures that maintain and modify institutionalized relationships. It is these rules or

habits that give distributed innovation processes their stability. They … provide the

frameworks for generating and combining knowledge’ (2001: 577). It is through such

processes that there emerge ‘fresh conjectures’ (Metcalfe, 2001). As the added

emphases suggest, these arguments are as applicable at the level of the cluster as they

are at that of the firm. Such arguments also inject further dynamism into the cluster

by, at least partially, making the rate of growth a function of technological change and

a property or outcome of the cluster as system of resources and attendant capabilities.

The importance of the interaction of governance arrangements and resource

distribution might be most acute in relation to intangible assets, particularly those

referred to by Storper as ‘untraded interdependencies’ and described as attaching to

‘the process of economic and organizational learning and coordination’ (Storper,

1997: 21). These seem to bear some likeness to Marshall’s ‘mysteries of trade’ that

are to be found somehow ‘in the air’, and which according to the RBV form the

12

Page 14: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

source of competitive advantage only for the individual firm. Amongst groups of

firms, the key point is that the impossibility of contracting for the exchange of such

resources means that their circulation and distribution has to be achieved via other

means. From this perspective, ‘relational assets’, such as networks, become vital. But

the natural limits to trust ensure that networks may also work to restrict access to such

scarce resources (Cookson, 2003; Toms and Filatotchev, 2003). As Breschi and

Lissoni argue, if ‘epitemic communities’ will not disclose their ‘common codebooks’,

then they have the power to act in highly ‘exclusionary’ ways (2001: 989). In the

intricately and densely structured relations characteristic of clustered industries, where

proximity remains a powerful force, such outcomes are, it may be suggested, as likely

as any ever-refreshed pattern of access. Together, these arguments stress the need for

a more structured and nuanced view of governance arrangements in clusters.

In sum, casual awareness of the empirical record must throw elements of the

Marshallian model, and its more recent manifestations, into doubt. In particular, the

literature has too little to say about the impact of the heterogeneity of the positions

occupied by different actors in the business structures of clustered industries. The size

of firms and their structural positions, in terms of both vertical and horizontal linkages

and their relative power, all influence access to and utilization of resources, and hence

the challenges of allocation and co-ordination facing entrepreneurs and managers.

These challenges in turn shape priorities, interests, attitudes and behaviours within

and beyond firm ‘boundaries’. The above arguments point, then, to the need for a

framework that captures the co-determining relationship between governance and

resource issues in clusters and, indeed, the Chandlerian firm. Having noted both the

gaps and areas of fit between these two core literatures we will now turn our attention

to exploring the dimensions of a framework able to capture and reconcile them. For

13

Page 15: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

the purposes of concision, we will in next sections focus largely on clusters and the

processes that facilitate (or constrain) transition away from that organizational form.

3. A proposed synthesis

We will now attempt to synthesize the above arguments in order to offer an integrated

perspective on resources, governance and industrial organization that can offer a

structured analysis of the empirical evidence. Arguing that Chandlerian models of

corporate and industrial change marginalize the importance of the externalization of

scale and scope economies, Toms and Wilson (2003) contend that structural

transformations may best be analysed by reference to two interacting continua. The

first describes accountability (from high to low) and the second the balance between

externalization and internalization, thereby forming an analytical two-by-two matrix.

Focusing on a ‘series of theoretically sound hypotheses for explaining the process of

change’ as firms move from one quadrant of the matrix, Toms and Wilson do not

concentrate on specifying in detail the forces shaping the balance between

externalization and internalization. Clustering clearly has a significant role to play in

this respect and, thus, we attempt here to refine that element of the Toms and Wilson

model and to link it, through the concept of entrepreneurial scope, to the patterns of

structural change emerging from cluster dynamics.

According to Toms and Filatotchev, ownership and governance structures, allied to

perspectives on the strategic resource content of business activities, in which

‘managerial and entrepreneurial resources drive growth and diversification,’ form ‘an

important context that moderates strategic response’ in periods of both growth and

crisis (2003: 70, 69). Conceptualizing governance in terms of accountability (from

14

Page 16: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

transparent to opaque) and the resource base of the firm as either narrow or extensive,

Toms and Filatotchev capture the characteristics of networks with a further matrix. In

this model ‘the degree of transparency will be a function of the degree of dependency

on external stakeholders for resources’ (2003: 71). Here, resource dependency is

linked to the industry growth rate as functions of technological change and the

regulatory environment, and thereby ‘impacts on the social construction of networks’

as a governance mechanism. Technology and location of productive resources also

impacts on the necessity for and ability of firms to internalize resources, or,

conversely, how successful they are in constructing and maintaining networks with an

effective basis in trust.

Here, we contend that the key to understanding cluster dynamics, and thus the

potential of clusters to generate structural transformations, lies in exploring the

interdependency of governance structures and scale and scope economies in clusters,

emphasizing the multi-level nature of both governance arrangements and patterns of

resource distribution and dependency. In other words, governance structures exist

(and overlap) at the level of the firm and the cluster, whilst resource bases and

dependency at the level of the firm must be situated in relation to resource distribution

and dependency at the level of the cluster. Moreover, it is the cluster-level governance

arrangements that integrate firm-level resource bases and cluster-level resource

distribution and dependency. The balance between internalization and externalization

of scale and scope economies is an expression of the interdependency of governance

and resource issues for both firms and clusters. Incorporation of the resource base

with resource dependency facilitates the simultaneous consideration of transaction

costs and agency costs as ‘information costs’ in the context of organizational

evolution (Casson, 1997).

15

Page 17: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

To summarise, if the characteristics of clusters are to be contrasted across industry

and through history, then the degree of internalization of resource and the degree of

resource dependency and transparency to external monitoring are likely to be

important determinants of cluster characteristics. An external resource base which is

shared through clustering arrangements will promote a flat, lateral structure, whilst

internalisation promotes hierarchy. In the former case, the relatively equal distribution

of resources across the cluster membership creates mutual resource-sharing

arrangements. This structure might be described as a ‘heterarchy’ hub-style structure

with the lateral linkages referred to by Hedlund (1986), but allowing the ownership

and accountability arrangements to vary. Heterarchy here is used in contrast to the

hierarchy that is promoted by resource internalisation within a single firm/ownership

structure in Chandler’s model. At the same time, dependence on resource providers

external to the cluster promotes centralisation, as the cluster members need to create a

conduit for securing new resources. If the cluster members are resource self-sufficient

and resource independent, this promotes equality within the cluster through

arrangements to share existing resources, and such clusters may be characteristically

decentralised. The proposed general relationships are set out in Figure 1, while Figure

2 offers some contrasting examples.

16

Page 18: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Resource dependency

High Low

Internal Hierarchical,

centralised

Hierarchical,

decentralised

Resource base

External Heterarchical,

centralised

Heterarchical,

decentralised

Figure 1 Resources, governance and industry characteristics

Resource dependency

High Low

Extensive Conglomerates, M-

forms etc

TNCs

Family firms, loose

holding companies

Resource base

Narrow Hub-based

industrial district,

putting out systems

etc

Associations of

firms, price fixing

cartels

Figure 2 Resources, governance and industry characteristics: Empirical examples

17

Page 19: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

In Figures 1 and 2, the resource base refers to the resource bases of the individual

firms that comprise the cluster. As the resource base is centralised, a single firm takes

over an increasing number of functions, for example by buying up other members of

the cluster, thereby internalizing the resource bases of the constituent firm(s).

Movement may also occur in the opposite direction through spin-offs, spin-outs and

demergers. Resource dependency meanwhile refers to the degree of dependence of all

cluster members in the aggregate upon external resource providers. These resources

include finance, labour, raw materials, etc.

As the figures suggest, the characteristics of clusters, and other forms of industry

structure, can be ascertained with reference to two characteristics. First, there is the

existing or ex ante resource base, geographical location and their tangibility or

intangibility. Examples of tangible resources include production capacity or local raw

material sources, whereas intangible refers to R&D expertise and local pools of

knowledge assets. Second, one must consider the growth rate, which is linked

historically to the developmental and innovation propensities of the resource base. If

the asset/resource base of the cluster needs to be modified for, say, reasons of new

discovery and technical change, the degree of resource dependence will be altered.

The precise characteristics of the cluster will depend upon which of these forces are

the strongest.

For example, the forces of innovation/ technical change may be strong, creating

high dependence on financial resource providers, but which may be inadequate (for

example, the capital markets of the industrial revolution. See Wilson 1995). Of

course, the converse can also be true; weak forces of change in the asset base and low

technical discovery coexisting with powerful governance agents that demand

exit/restructuring. Successful clusters might be expected in cases where both aspects

18

Page 20: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

work positively as sufficient and necessary conditions. It follows that clusters might

be less successful not just where neither operates positively, but also in cases where

one of the two fails to operate.

The dependent variables of the model in Figures 1 and 2 are the cluster

characteristics. These include components such as institutional arrangements, market

micro-structure, characteristics of entrepreneurs, labour relations, competitive

advantage, internal accounting arrangements, and so on. Figure 2 sets out some

example scenarios. In quadrant 1, where each firm in the cluster commands an

individually narrow resource base, but dependency on external resources is high, the

cluster members must identify an appropriate conduit by which the resources are to be

secured. The likely effect is to empower the section of the cluster, perhaps a particular

firm, which is delegated responsibility for securing these resources. This may create

moral hazard and adverse selection problems within the network, where a single firm

acts as agent for the remaining firms (Zaheer and Venkatraman 1995), in addition to

the principal-agent relationship between the delegated firm and the resource provider.

Accounting and accountability structures must therefore be created which operate in

two important directions. These problems might be solved by internalising the

relationships, for example, where one firm takes over the other members of the cluster.

In this situation, central management attempts to solve the moral hazard problem by

resort to internal planning and management accounting controls that might be

characterised by M-form structures. In this model the principal responsibility of the

corporate centre is raising resources and possibly personnel functions, and distributing

them as rationally as possible to the product divisions.

Where there is less dependence on outside resources, for example in family

capitalism, the accountability structure of the cluster alters again, this time, for

19

Page 21: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

example, through the use of interlocking directorships within a controlling elite that

manages a looser hierarchy, perhaps through a holding company or federal structure.

Finally, where the resource base for individual firms is narrow, but there is no external

resource dependence, the centralisation effect that arises from the use of a delegate

firm disappears. However, there are still moral hazard and free-riding problems

amongst the relatively equal participant firms and cluster norms may be enforced

through industry associations and agreements on issues such as price fixing. In these

conditions, the participants have incentives to share accounting information so that

costs of production are known and prices can be maintained at levels above production

cost.

Can the processes underlying transitions be formalized more clearly, however?

First, there is a propensity for different dynamics to induce movements in particular

directions. Thus, the dynamic on the vertical axis is primarily technological change,

though changes in market configurations are also likely to impact on this plane.

Operating on the horizontal axis we are more likely to find forces that impact on the

social ownership of firms, such as company law and rules governing financial

disclosure and other aspects of corporate governance. As noted above, however,

changes in resource bases, in whatever direction they occur, will alter the degree of

resource dependency at the level of the cluster. Appropriate arrangements with regard

to resource dependence are about ensuring access to the resources required for current

and future strategies. Finding this new position is at the core of the challenge of

handling change effectively.

Understanding the specific outcomes arising from particular conjunctions of

industry organization and forces for change requires a view on how resource issues

(bases and dependencies) interact with governance systems – how and to what extent

20

Page 22: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

appropriate and effective structures of resource dependence can be built. Here we

contend that resource bases and dependencies, on the one hand, and governance

systems, on the other, change in fundamentally different ways and, perhaps most

importantly, at different rates.

Forces for change powerful enough to induce transitions across our matrix,

whether they originate in the technological realm, in the market, or in other

environmental factors, are likely very often to be relatively radical and discontinuous.

Governance in clusters, in contrast, rooted very often in networks and other related

structures (kinship, religion or ethnicity, for example), is, as we note above, socially

constructed. Effective monitoring and accountability are highly dependent on

information, and information costs are heavily influenced by relational structures and

properties, including trust. Even as dispersed ownership and efficient markets for

shares emerge, governance long remains heavily socially constructed. (One need only

note policies for the recruitment and succession to top management positions to

reinforce this argument). On the governance side, a practical example might include

the recovery and maintenance of invested capital to meet outstanding financial claims.

Where there are significant surpluses or deficits, resulting from cycle effects, financial

crises or other causes, the reordering of such financial claims may be slowed down by

legal requirements and vested or conflicting interests. As a socially constructed

phenomenon, governance, especially in complex, systemic clusters, is much less

likely to experience rapid, radical and discontinuous change. Thus, in periods of

structural stress in clusters misalignment between resource and governance

arrangements is likely to develop. Transition towards finding a new position with

regard to resource bases and dependencies will only occur to the extent, and in the

direction allowed by, more slowly evolving governance systems. Thus, whilst we

21

Page 23: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

present here arguments about the mechanisms underlying transitions within the

model, emphasising misalignment between resources and governance to explain

discontinuous development in business structures, exploring specific outcomes in

different clusters is thus a contingent, time-dependent and empirical question.

Moreover, further ‘drag’ can originate in the resource sphere itself. In particular, it

can be very difficult to ‘unlock’ asset specificities located in the interstices of the

complex, organizationally dispersed production networks of many clusters.

In sum, we recast the balance between internalisation and externalisation, and thus

industry structure, not as simply determined by either technology or markets, and only

rarely in the gift of the strategizing firm, but also as a function of the interaction of

governance arrangements and the concepts of resource distribution and dependency,

examined here for the purposes of exposition through the case of clustering. The

dynamics that flow from these arguments are non-deterministic and multi-directional.

Thus, if, at the level of the firm, governance follows resource dependency, as

suggested by Toms and Wright (2002), then governance at the cluster level may

determine access to resources for current and future structures and strategy for both

firm and cluster. In other words, governance shapes the viability and direction of

restructuring. Explaining similar linkages in the context of developments in the UK

corporate sector between 1950 and 2000, Toms and Wright (2002) argue that

‘structure and performance are outcomes of decisions where information costs are

driven by ex ante strategy and institutional arrangements’. The key point, once again,

is the non-deterministic nature of the processes; our case studies will detail clusters

where conditions did lead to adjustment through the creation of a collective

entrepreneurial scope and others where they did not.

22

Page 24: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

4. Clusters in pottery, chemicals, and aerospace

We will now explore the utility of the framework outlined above in the context of

three British industrial clusters; the North Staffordshire Potteries, the Northwest

chemical industry, and the Northern Ireland aerospace industry.

The North Staffordshire Potteries are amongst the most-well known of England’s

‘classic’ industrial districts. How might they be positioned with regard to the

framework outlined above? By the mid-nineteenth century, this cluster had, following

a process of slow development, reached a state of considerable maturity (Weatherill,

1971; Popp, 2001).

Individually, the district’s many firms, varing widely in size, had extensive

resource bases and were highly vertically integrated. At the same time, the cluster, as

a whole was highly resource self-sufficient – dependence on external stakeholders for

critical resources was low. The district was dependent on external sources for

important physical inputs, most noticeably clays from Cornwall in Southwest

England, but the relatively simple and undifferentiated nature of these resources

meant that trade in them could be handled through unmediated, market-like

relationships, obviating the need for the establishment of complex monitoring

mechanisms. Other critical resources, in particular finance, technology and skills,

where their qualities and characteristics demanded that more rigorous monitoring be

in place, were all overwhelmingly generated within the cluster, and typically within

the individual firm. Thus, finance for both working and fixed capital was typically

secured within a highly personalized regime; family, unincorporated partnerships and

retained profits predominated. In relation to technology, the artisanal background of

many entrepreneurs, in conjunction with a strong empiricist tradition and weak forces

23

Page 25: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

for technological change in the resource base, led to the internalization within the

individual firm of both process and product innovation. These forces meshed well

with the extensive productive resources of firms. Skills and skilled labour, in

apparently classic Marshallian fashion, were to a greater extent a property of the

cluster as a whole, but here again personalized regimes of recruitment (Dupree, 1995)

and on-the-job training served further to extend the command of the firm over an

extensive resource base.

However, for one important resource, market information, the cluster was highly

dependent on external stakeholders. The characteristics of market information as a

resource, and in particular its asymmetry, intangibility and tacitness, meant effective

monitoring by both providers and users was vital in an industry serving highly

differentiated and dispersed consumer markets. But firms varied widely in their

success in solving this problem. A small cohort of larger firms was able to build links

with both domestic and foreign markets in which trust and reciprocity went some way

to guaranteeing the quality of market information (McKendrick, 1960; Ewins, 1997).

These firms were also usually able to use these arrangements to facilitate the building

and maintenance of a further powerful, and highly firm-specific, resource, namely,

reputation. However, in this area there was little resource sharing and the vast

majority of firms remained condemned to reliance on highly asymmetric relationships

with external resource providers (Popp, 2002).

What were the governance arrangements, at the level of both the firm and the

cluster, within which were situated these patterns of resource distribution, access and

dependence? First, in common with other English districts and clusters (Wilson and

Popp, 2003) we would stress the impact of the highly contingent nature of the

processes through which the Potteries clusters was formed, strengthened and

24

Page 26: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

deepened in the highly uncertain and risky economic and institutional environment of

the eighteenth and early nineteenth centuries (Bresnahan et al, 2001; Feldman, 2001).

In the absence of effective means for forming joint stock companies, firms remained

small and personally managed and owned. Where technologies permitted in other

industries, these conditions encouraged narrow resource bases and extensive resource

dependency at a firm level, particularly with regard to knowledge and information.

But given a highly risky business environment, entrepreneurs sought wherever

possible to situate their resource dependencies in governance frameworks that

guaranteed some stability by drawing on pre-existing ‘organic’ ties, whether of

family, religion or community (Cookson, 1997; Caunce, 2003). In the Potteries, this

tendency was compounded by the vertical integration and extensive resource bases

consequent upon the technology of pottery production. However, despite emphasis on

the way in which entrepreneurs drew on the embedded relationships of family and

wider social networks, some caution is required, for these arrangements in no way

precluded fierce competition within the cluster or highly individualistic attitudes at the

level of the firm (Wilson and Popp, 2003b). Much network-like behaviour did take

place, of course, but this tended to remain relatively self-contained. We may think of

these networks as ‘capsule networks’, with relatively defined and impermeable

boundaries. The district as a whole cannot be conceived of as a network organization.

Clusters were thus often important vehicles for entrepreneurship, as predicted by

theory (Casson, 2003; Feldman, 2001).

Indeed, in North Staffordshire, extensive firm-level resource bases, in combination

with low cluster-level resource dependence, had shackled the emergence of effective

cluster-level governance mechanisms. Interests and priorities were divergent, resource

sharing rare and insignificant, monitoring mechanisms weak and networks opaque and

25

Page 27: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

exclusionary. In particular, leadership, predicted as significant to successful cluster

adaptation processes, was absent (Casson, 2003; Carnevali, 2003). Firms were

hierarchical and the cluster decentralized, missing the lateral heterarchy-like linkages

that emerge to handle the demands of resource sharing. With reference to Figure 1, we

can situate the Potteries in the top right-hand quadrant; a classic Marshallian industrial

district displaying a decentralized and hierarchical structure.

From the 1870s onwards, the pottery industry began to face increasingly severe

external challenges. Low-cost foreign competition and protectionist regimes overseas

impacted on domestic and export markets. Price-oriented competition in increasingly

integrated mass markets appeared to threaten with redundancy many of North

Staffordshire’s resources, particularly intangible assets such as worker skill,

reputation and privileged access to market information (Popp, 2000b). It is important

to note that this new competition was largely driven by low-cost labour; at this point

in time there were only weak forces for change in the asset bases of pottery industries

in all countries.

At the same time, developments within the cluster posed further challenges as the

‘thickness’ of localized resources generated increasing instability in the population of

firms, waves of start-ups and failures seeming to increase in magnitude with every

cycle (Popp, 2001). Here it is important to distinguish resource circulation between

exiting and entering firms from resource-sharing amongst incumbent firms, where

interdependencies fed into industrial and technological upgrading (Saxenian and Hsu,

2001).

The response to this period of crisis was weak. Structural change, in terms of both

the number and size of firms and of patterns of integration, was limited (Popp, 2001).

The cluster remained highly competitive within itself and the already fragmented

26

Page 28: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

cluster level governance mechanisms of the district disintegrated further, leading for

example to the decline of features such as repeat trading and notions of reciprocity.

This is not to say that some governance agents did not attempt to demand and promote

restructuring, for example through associational activities such as the foundation of

the North Staffordshire Exchange (Popp, 2000a), but the extensive resources bases of

firms, the thinness and weakness of resource sharing, and the opaqueness of

networks, all conspired fatally to shackle their attempts to build more centralized

cluster-level governance arrangements. The firm-level independence consequent upon

existing patterns of resource distribution and access, buttressed by powerful cultural

value systems (Popp, 2003b), blocked organizational and industrial upgrading and

transformation. The cluster remained locked in same quadrant of Figure 1.

The cluster of chemical industries in Northwest England4 developed much more

rapidly than the Potteries. Both the technology of alkali production using the Leblanc

process and the rapid mode of development had significant implications for the

resources bases of firms and the resource dependence of the cluster as a whole. In

particular, the highly emergent and constantly evolving nature of the Leblanc process,

and the need to assemble large, complex and expensive resource sets, promoted

relatively narrow resource bases at the level of the individual firm. A natural corollary

of these conditions was considerable resource sharing at the level of the cluster. The

situation was captured in a letter written by Ludwig Mond to his parents in 1872:

[h]ere where we have fifty factories in a few square miles, are many factories,

some of which have started in a very small way and which deal with only a very

part of soda manufacturing. The buy the by-products from the other factories of

sell their products to other factories for working up … We have many examples

of small factories which have hit on the right branch (Hardie, 1950: 64).

27

Page 29: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

This pattern of resource distribution bound together the entrepreneurs of the cluster in

a complex web of interdependencies and exchange relationships that rippled out

beyond the core productive processes in the manufacture of soda and alkali to

encompass an ever more refined utilization of the industry’s numerous waste and by-

products, the provision of important inputs, including both raw materials and

engineering know-how and plant, and the distribution and exploitation of finished

products in a wide range of industries. Paradoxically, the lack of social embeddedness

displayed by the industry, Widnes being effectively created by the industry, promoted

the formation of the lateral, heterarchical linkages and an openness to dependency on

external providers that was largely absent from the Potteries. Thus, the development

of the industry unrolled through a process of spin-offs and spillovers led by the kind

of ‘seed-corn’ firms identified in studies of other English clusters (Popp, 2003a;

Lloyd-Jones and Lewis, 2000; Lloyd-Jones and Lewis, 2003). Moreover, many key

entrepreneurs had pursued highly peripatetic careers prior to settling in Widnes and

thus brought with them a wide range of extra-regional linkages that were often to

prove vital in securing a wide range of resources, including technical knowledge,

finance and market linkages. Though individual firms were typically founded on a

regime of personal capitalism, the cluster was much less self-sufficient than was the

Potteries in the same period. In sum, resource bases were relatively narrow and

resource sharing within the cluster high. These conditions were matched by relatively

high dependence on external resource providers. These resource conditions combined

with hierarchical governance arrangements at the level of the firm, but with

heterarchical, centralized, networked arrangements at the level of the cluster and

beyond.

28

Page 30: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

As the cluster and industry began to mature, new challenges arose. With both

technologies and the firm population beginning to stabilize from the early 1870s,

attention turned to reinforcing and defending the cluster’s locational advantages

(Warren, 1983). Thus, though pressure for change in the resource base remained

weak, change was demanded in governance arrangements at the level of the cluster

and beyond. Leadership had long played a role in the development of the cluster, for

example in the provision of infrastructure with public good like characteristics, but

this feature became further accentuated. Where the leading governance agents in the

Potteries had proved to have insufficient authority to either promote or demand

reorientation, leaders in Widnes, building on the interdependencies consequent on

widespread resource sharing and dependence, were far more successful in this

direction. The Widnes Traders’ Association, in particular, provided from the mid-

1870s onwards a valuable vehicle for strengthening both intra- and inter-cluster

linkages, whilst at the same time facilitating greater centralization on key governance

agents that acted as increasingly powerful conduits for the distribution of valuable

resources, especially information and political leverage (Popp, 2003a). Referring to

Figure 1, the cluster moved further towards to the bottom left quadrant as progressive

centralization occurred, supplementing but not, as yet, supplanting the district’s web of

heterarchical linkages.

However, these challenges were to be dwarfed by those that arose in the 1880s,

when pressure for change in the resource base of both firm and cluster was heightened

dramatically by the introduction of the Solvay process by Brunner, Mond and Co.

(Reader, 1970; Chandler, 1990). Though it continued to be refined, and still had its

adherents, the Leblanc process was doomed by the cost advantages Brunner, Mond

enjoyed from their exclusive licence to work the Solvay process in the UK. It became

29

Page 31: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

increasingly apparent that a radical reorientation of the resource bases of both firms

and cluster would be imperative, but that this would prove extremely difficult in the

context of continued ‘de-verticalization’ and narrow individual resource bases. Thus,

the members of this heterarchical and centralized hub-based industrial district were

faced with the choice of either competing each other into the ground or restructuring.

Building on the centralized, cluster-level governance arrangements, they chose the

latter course, internalizing resources in 1890 through the formation of the United

Alkali Company. Referring again to Figure 1, the alkali manufacturers of the Widnes

district,5 moved through merger from bottom left to top left, with internalized

hierarchical linkages replacing externalized heterarchical linkages. The bridge

between two radically different governance configurations pre- and post-merger was

the progressive centralization that had been characteristic of the cluster from the

outset, and which now formed the basis of what was, for the time, a relatively

advanced organizational structure within the UAC (Chandler, 1990). In turn, in 1926,

the UAC was an important component in the newly-created ICI, completing the

transition from ‘classic’ industrial district to modern-corporation.

While never as large as those in Seattle or Toulouse, the Northern Ireland

aerospace cluster has since the mid-1930s been one of the most important

manufacturing groups in that region. Composed of one large firm, Shorts Brothers

(since 1989, Bombardier-Shorts) and a dozen medium-sized specialist engineering

ventures, most of its resources were situated in the locale. First, it was able to access

the rich source of skilled labour and sub-contracting firms that was linked to the

engineering and shipbuilding industries of Northern Ireland – denoting links to other

well-established regional resources. Second, Queen’s University Belfast has provided

30

Page 32: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

both graduates and design input through its world-renowned aerospace engineering

department.

Most importantly, though, as the principal firm, Shorts Brothers (since 1989,

Bombardier-Shorts), was in public ownership between 1943 and 1989, this ensured a

plentiful supply of capital, whether in the form of direct subsidies, development grants

or debt write-offs. Not only was this money provided because of the severe problems

besetting Northern Ireland’s traditional staple industries (linen and shipbuilding), but

also one must remember that 95 per cent of the 8,000 workforce at Shorts Brothers’

six main sites in and around Belfast were Protestant. It is sufficient to note that since

the seventeenth-century Protestants had dominated the region politically and

economically. With sectarian violence on the increase, especially after 1969, it was

regarded as politically acceptable to continue to subsidise Shorts Brothers, most

notably through the auspices of the Northern Ireland Office, the British government

department with specific responsibility for that region. Between 1976 and 1989, for

example, Shorts Brothers received £186 million from the British taxpayer, as well as

substantial government contracts for its military products. Thus, Shorts Brothers were

significantly dependent on the British Government as an external provider of

resources.

With ample supplies of skilled labour and design input, the cluster was largely self-

sufficient, the principal firm operating in a hierarchical and centralised manner, except

with regard to financial resources. This places the cluster in the top-left-hand quadrant

of Figure 1. The governance arrangements were also inherently centralised, given the

role of the state as both financier and principal customer. At the same time, as what

was Northern Ireland’s largest manufacturing employer, Shorts Brothers came to be

regarded as a central feature of the Protestant majority’s political importance,

31

Page 33: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

cementing the network along sectarian lines. This cohesion was one of the cluster’s

strongest features, providing clear leadership over a period dominated by the collapse

of Northern Ireland’s staple industries and the escalation of religious conflict.

However, the state’s willingness to subsidize Shorts Brothers, and consequently the

rest of the cluster, failed to provide the discipline required to improve

competitiveness. High dependence on the British Government as an external resources

provider was not matched by adequate monitoring and accountability. In particular, it

is vital to note that substantial debts were accumulated, amounting to £750 million by

1989. This indicates how even though £100 million of public money was spent in the

1970s developing the SD360 transport aircraft, profitable sales were slow to appear in

the accounts. Furthermore, not only was the product range regarded as uncompetitive,

‘fundamental weaknesses in the company’s accounting records’ were discovered

during the 1980s. Cash funding forecasts also proved woefully inadequate, largely

because Shorts were still using a financial control system that failed to accommodate

all the information generated by its various activities (Northern Ireland Audit Office,

1991: 13). It was consequently no surprise when the Conservative government

concluded ‘that Shorts’ best prospects for survival lay in the private sector’ (Northern

Ireland Audit Office, 1991: 14).

The catalyst for change in this context was the Thatcherite government’s desire to

privatize as much of the public sector as possible, especially the loss-making ventures.

In the case of Shorts Brothers, in 1989 a Canadian transport group, Bombardier Inc.,

was chosen from amongst several dozen bidders to secure the future of this loss-

making firm. It is important to add that as a means of improving the attractiveness of

Shorts Brothers in 1989 the government not only provided the firm with a £390

million loan, but also revamped the senior management. In addition, as part of the

32

Page 34: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

deal finally agreed with Bombardier, £750 million of Shorts’ financial liabilities were

written off, while various aid packages were promised for the following five years.

Thus, while external resources continued to be offered, Bombardier Inc. were not

planning to rely on state subsidies; they actually instigated a radical realignment of the

cluster, moving from a centralised, hierarchical structure to a much more heterarchical

form of organisation. This reflected the decisive change in levels of competition

across the aerospace industry, with a wave of mergers occurring in response to the

radical changes in a marketplace severely disrupted by the end of the Cold War and

consequent collapse in military orders. By the 1990s, then, the cluster’s governance

system had moved away from state patronage to participating in systems that were

dominated by global capital market and corporate governance influences.

While the privatisation of Shorts Brothers would have represented a threat to many

Northern Ireland jobs, both within the firm and across a regional supply chain,

Bombardier recognised that the acquisition represented a significant asset. In the first

place, government subsidies and contracts were guaranteed, at least over the first five

years. Secondly, the workforce was highly skilled, while continued contact with the

aerospace engineering department of Queen’s University Belfast significantly boosted

the availability of expertise in what was a technology-intensive sector. Bombardier

management was also convinced that it would be possible to link the Belfast

operations into what was a global supply chain connecting activities in Canada and

the USA (Wilson, 2002: 24). By the 1990s, Bombardier had acquired Lear Jet (USA),

Canadair (Canada) and de Havilland (Canada), as well as forging links with the two

major American aerospace firms, Boeing and Lockheed, and Airbus Industrie, the

European civil aircraft maker. This has assured regular component orders, with

Bombardier group work accounting for 50 per cent of Shorts’ turnover, Boeing and

33

Page 35: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Lockheed 20 per cent and Airbus Industrie a further 15 per cent. At the same time,

although Bombardier brought capital, new technology and guaranteed customers to

Shorts Brothers, management recognised the imperative need to operate in a much

more heterarchical fashion, regionally and globally, if the group was going to compete

effectively with the newly-merged aerospace corporations that emerged after the end

of the Cold War in 1989. The organizational and spatial dispersal of resources led to a

narrowing of resource bases and deepening, and reconfigured, dependence on a wide

range of external resource providers.

A clear indication of how these external forces instigated radical change within the

Bombardier group was the termination of aircraft production at the Belfast sites. This

move had been forced on management by the inability to develop a competitive range

of civil aircraft, the SD360 failing to generate much order cover once British

government contracts tailed off in the early-1990s. On the other hand, utilizing

considerable expertise in the design and manufacture of major aerostructures and

nacelle systems, and for processes such as composites and metal bonding, Bombardier

management has played an active role in changing the organizational and production

culture at Shorts. Implementing this strategy of functional integration, the Shorts

factory no longer designs or produces complete aircraft, but retains a design team of

400 engineers focusing on engine nacelles (the pod that carries the jet engine),

fuselage construction and landing gear doors (Best, 2001; 203-4).

Apart from participating in the development of a strong cluster within the

Bombardier group and its commercial partners, another intrinsic feature of the revived

firm’s strategy was forging partnerships with local firms capable of supporting the

total quality programme. As Best reports (2001; 204), in areas like components and

tooling, Bombardier-Shorts (as the firm is now known) has created a network

34

Page 36: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

including Moyola Precision Engineering, Langford Lodge Engineering and

Huddleston Engineering. Similar arrangements have also been made with Project

Design Engineers (testing rigs and special purpose machinery), Mallaghan

Engineering (aircraft ground-handling equipment), Denroy Plastics (injection

moulded components) and RFD Dunmurry (liferafts, lifejackets and safety

equipment). Much of the firm’s work in this area is also reinforced by the creation in

1998 of the Northern Ireland Aerospace Consortium, a trade body that grew directly

out of the Northern Ireland Office’s attempts to develop effective clusters across

many of the region’s industries. This reflects the manner in which Bombardier-Shorts

were substantially aided by government in fashioning a heterarchic network within

Northern Ireland, principally as a means of strengthening its own internal reforms

initiated in the mid-1980s. This also supports the claims made by Pettigrew and

Fenton (2000) that multinationals have been pursuing both internal and external

organizational improvements, balancing heterarchical with hierarchical approaches in

enhancing competitiveness in the global economy that dominates high-technology

sectors like aerospace. The cluster had consequently made a decisive move from the

top to the bottom quadrant on the left-hand side of Figure 1, with the 1989 change in

ownership acting as the crucial catalyst, alongside decisive changes in the global

aerospace market arising from the end of the Cold War. Here, dependence on external

providers has been central to the revitalization of cluster that in 1989 was being

threatened.

We believe this framework is sufficiently robust to accommodate a wide and

disparate range of empirics; indeed, it will be strengthened by the elaboration of

further case-studies. Brief indication is given here of a few further examples. The

Birmingham jewellery district developed as a case of extreme deintegration,

35

Page 37: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

decentralization and small-scale business; it was self-sufficient and heterarchically

governed. We may situate it in the bottom right-hand quadrant of Figure 1. However,

the need to respond to external challenges (including changes in fashion and trade

policy), as well as increasing anarchic practices by some members of the district,

heightened awareness of the need for a more centralized, district-level governance

system and led, in 1887, to the formation of the Birmingham Jewellers’ and

Silversmiths’ Association (BJSA). This organization was able to discipline members

of the trade and augment local resources, whilst also reaching out beyond the cluster

to develop cultural and political resources at the national level. These developments in

no way disturbed the productive system of the district or governance of individual

firms (Carnevali, 2003a). As a result, the district moved from bottom-right to bottom-

left quadrant of Figure 1. It is argued that the BJA had played a vital role in

maintaining the vitality of the district down to the present day (Carnevali, 2003b).

In contrast, the British machine tool industry, clustered in the English midlands,

moved in the opposite direction. Several ‘hub’ firms, and Alfred Herberts in

particular, played a key role in co-ordinating the dispersed resources of this, another

classically disintegrated, industry. At a local level at least, Herberts and other firms

were also willing to provide leadership in governance issues. But on the national

stage, especially in the context of depression in the inter-war years, Herberts refused

to fulfil a similar role in providing leadership in relations with potential external

resource providers, including the government. As a result, heightened competition

began to threaten the internal cohesion of the district and industry (Lloyd-Jones and

Lewis, 2003). The transitions experienced by the industries examined here are

captured in Figure 3.

36

Page 38: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Resource dependency

High Low

Extensive Shorts Brothers

The United Alkali Company, 1890

The Potteries industrial district: static

Resource

base

Narrow Bombardier Widnes industrial district, 1848-1890 Machine tools Birmingham jewellry post 1887

Post-1920s Birmingham jewellry pre-1887

Figure 3: Transitions in chemicals, pottery, aerospace, jewellry and machine tools

5. Conclusion Our purpose here has been to propose and explore a framework within which to assess

the dynamic properties of industrial clusters and districts. In particular, we have been

interested in the forces that illuminate the process of transition in business

organization, at the level of firm, cluster and industry. Drawing on the RBV and

resources dependence literatures, the model focuses on the interdependence of

resource bases and systems of governance, defined primarily in terms of monitoring

and accountability. The principal novelty of the paper lies perhaps not only in this

37

Page 39: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

examination of the interaction of resources and accountability, but also in the

extension of both concepts to the level of the cluster. Thus, we argue that resource

bases, resource dependencies and systems of governance exist at the level of the firm

and at the level of the cluster. These forces are captured in a matrix. It is important to

stress that the model is intended to be non-deterministic. Multi-directional movement

between the quadrants of the matrix is both possible and observable.

The paper has a further purpose, however, which is to attempt to break down the

previously somewhat dichotomous positioning of two literatures within business

history; the Chandlerian narrative of the internalization of economies of scale and

scope, and the story of flexible specialization and externalization realized within

industrial clusters and districts.

Challenges remain, of course. Most obviously, further testing against the historical

record would be welcome. Important elements of the model also require further

elucidation. For example, the roles within our framework of business culture(s),

agency and choice need more careful specification. Nonetheless, we contend that, in a

world characterized by the persistence of some very old industrial clusters, the

formation of many more new ones and the splintering of the classic Chandlerian

corporation, exploration of the dynamics of clustering remains a valuable and

worthwhile research agenda.

38

Page 40: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

References Amin, A. and N. Thrift. (1994), Globalization, Institutions and Regional Development

in Europe. Oxford University Press: Oxford. Arthur, W. (1990), ‘Silicon Valley locational clusters: do increasing returns imply

monopoly?’, Mathematical Social Sciences, 19, 235–251. Barney, J. B. (1997), Gaining and Sustaining Competitive Advantage, Addison-

Wesley: New York. Bartlett, C.A. and S. Ghoshal (1989), Managing Across Borders: The Transnational

Solution, Harvard Business School Press: Cambridge, Mass. Beckman, Christine M. and Pamela R. Haunschild (2002), ‘Network learning: the

effects of partners’ heterogeneity of experience on corporate acquisitions’, Administrative Science Quarterly, 47: 92–124.

Carnevali, F. (2003a), ‘“Malefators and Honourable Men”: the making of commercial honesty in nineteenth-century industrial Birmingham’, in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp. 192–207.

Carnevali, F. (2003b), ‘Golden opportunities: jewellry making in Birmingham between mass production and specialty,’ Enterprise and Society, 4, 272–298.

Casson, M. (2003), ‘An economic approach to regional business networks,’ in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp. 19–43.

Castanias, R. and C. Helfat (2001), ‘The managerial rents model: theory and empirical analysis’, Journal of Management, 27, 661–678.

Caunce, S. (2003), ‘Banks, communities and manufacturing in West Yorkshire textiles, c. 1800–1830’, in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp.112–129.

Chandler, A. (1962), Strategy and Structure: Chapters in the History of the Industrial Enterprise. Harvard University Press, Cambridge, Mass.

Chandler, A. (1977), The Visible Hand. Harvard University Press: Cambridge, Mass. Chandler, A. (1990), Scale and Scope: The Dynamics of Industrial Capitalism.

Belknap: Cambridge, Mass. Cookson, G. (1997), ‘Family firms and business networks: textile engineering in

Yorkshire, 1780–1830’, Business History, 39, 1–20. Davies, H. and P. Ellis (2000), ‘Porter’s Cometitive Advantage of Nations: time for

the final judgement?’, Journal of Management Studies, 37, 1189–1213. Dupree, M. (1995), Family Structure in the Staffordshire Potteries, 1840–1880,

Clarendon Press: Oxford. Ewins, N. (1997), ‘“Supplying the present wants of our Yankee cousins

…”:Staffordshire ceramics and the American market, 1775–1880’, Journal of Ceramic History, 15, i–154.

Feldman, M.P. (1994), The Geography of Innovation. Kulwer: Dordrecht. Feldman. M.P. (2001), ‘The entrepreneurial event revisited: firm formation in a

regional context’, Industrial and Corporate Change, 10, 861–892. Frooman, J. (1999), ‘Stakeholder influence strategies’, Academy of Management

Review, 24, 191–205. Glaeser, E.L., H.D. Kallal, J.A. Scheinkman, and A. Shleifer (1992), ‘Growth in

cities’, Journal of Political Economy, ?.1126–1152.

39

Page 41: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

Gulati, R., N. Nohria, and Z. Akbar (2000), ‘Strategic networks’, Strategic Management Journal, 21, 203–215.

Hamilton, G. and R. Feenstra (1995), ‘Varieties of hierarchies and markets: an introduction’, Industrial and Corporate Change, 4, 51–91.

Hedlund, G. (1986), ‘The hypermodern MNC: a heterarchy’, Human Resource Management, 25, 9–35.

Hedlund, G. (1994), ‘A model of knowledge management in the N-form corporation’, Strategic Management Journal, 11, 501–512.

Jaffe, A.B. (1989), ‘The real effects of academic research’, American Economic Review, 79, 957–970.

Jones, C., W. Hesterly and S. Borgatti (1997), ‘A general theory of network governance: exchange conditions and social mechanisms’, Academy of Management Review, 22, 911-945.

Kamien, M., E. Mueller and I. Zang (1992), ‘Research joint ventures and R&D cartels’, American Economic Review, 82, 1293–1306.

Lloyd-Jones, R. and M.J. Lewis, ‘Business networks, social habits and the evolution of a regional industrial cluster: Coventry, 1880s–1930s’, in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp. 229–250.

Mathews, J.A. (2003), ‘Competitive dynamics and economic learning: an extended resource-based view’, Industrial and Corporate Change, 12, 115–145.

McKendrick, N. (1963), ‘Josiah Wedgwood and Thomas Bentley: an inventor-entrepreneur partnership in the industrial revolution’, Transactions of the Royal Historical Society.

McKendrick, N. (1960), ‘Josiah Wedgwood: an eighteenth-century entrepreneur in salesmanship and marketing techniques’, Economic History Review, 12, ???.

Nooteboom, B. (1996), ‘Trust, opportunism and governance: a process and control model’, Organization Studies, 17, 985–1010.

Northern Ireland Audit Office (1991), Sale of Short Brothers plc, HMSO. Oughton, C. and G. Whittam (1997), ‘Competition and co-operation in the small firm

sector’, Scottish Journal of Political Economy, 44. Pfeffer, J. (1992), Managing With Power: Politics and Influence in Organizations.

Harvard University Press: Cambridge, Mass. Pfeffer, J. (1997), New Directions for Organization Theory: Problems and Prospects.

Oxford University Press: New York. Pralahad, C.K. and Y. Doz (1981), ‘An approach to strategic control in MNCs’, Sloan

Management Review, Summer, 5–13. Popp. A. (2000a), ‘Trust in an industrial district: the Potteries, c. 1850–1900’, Journal

of Industrial History, 3, 29–53. Popp, A. (2000b), ‘Specialty production, personal capitalism and auditors’ reports:

Mintons Lts, c. 1870–1900’, Accounting Business and Financial History, 10, 347–370.

Popp, A. (2001), Business Structure, Business Culture and the Industrial District: The Potteries, 1850–1914, Ashgate: Aldershot.

Popp, A. (2002), ‘Barriers to innovation in marketing in the mid-nineteenth century: merchant-manufacturer relationships’, Business History, 44, 19–39.

Popp, A. (2003a), ‘Networks and industrial restructuring: the Widnes district and the formation of the United Alkali Company, 1890’, in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp.208–228.

40

Page 42: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

41

Popp, A. (2003b), ‘“The true potter”: identity and entrepreneurship in the North Staffordshire Potteries in the late nineteenth-century’, Journal of Historical Geography, 29,

Reich, R. (1990), ‘Who is us?’, Harvard Business Review, 68, 53–64. Reich, R. (1990), ‘Who is them?’, Harvard Business Review, 69, 77–88. Saxenian, A. and J-Y. Hsu (2001), ‘The Silicon Valley-Hsinchu connection: technical

communities and industrial upgrading’, Industrial and Corporate Change, 10, 893–920.

Swann, G.M. Peter, M. Prevezer and D. Stout (1998), The Dynamics of Industrial Clustering: International Comparisons in Computing and Biotechnology. Oxford University Press: Oxford.

Toms, S and I. Filatotchev (2003), ‘Networks, corporate governance and the decline of the Lancashire textile industry, 1860–1980’, in Wilson, J.F. and A. Popp (eds.), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot, pp. 68–89.

Toms S. and J.F. Wilson (2003), ‘Scale, Scope and Accountability: Towards A New Paradigm of British Business History’, Business History, 45. No.3.

Weatherill, L. (1971), The Pottery Trade and North Staffordshire, 1660–1760, Manchester University Press: Manchester.

Westphal, James D., L. Marc-David, J. Seidel, and Katherine J. Stewart (2001), ‘Second-order imitation: uncovering latent effects of board network ties’, Administrative Science Quarterly, 46, 717–747.

White, H. C. (1992), ‘Agency as control in formal networks’, in N. Nitin and R. Eccles (eds.), Networks and Organizations: Structure, Forms and Actions. Harvard Business School Press: Cambrdige, Mass., pp. 92–117.

Whittington, R., M. Mayer and F. Curto (1999), ‘Chandlerism in post-war Europe: strategic and structural change in France, Germany and the UK, 1950–1993’, Industrial and Corporate Change, 8, 519–555.

Whittington, R. and M. Mayer (2000), The European Corporation: Strategy, Structure and Social Science, Oxford University Press: Oxford.

Wilson, J.F. (2002), ‘Hierarchy versus heterarchy in British aerospace: BAE Systems and Bombardier Shorts in the 1990s’, paper given at the Academy of Management conference, Denver.

Wilson, J.F. and A, Popp (2003a), Industrial Clusters and Regional Business Networks in England, 1750–1970, Ashgate: Aldershot.

Wilson, J.F. and A. Popp (2003b), ‘Business networks in the industrial revolution: some comments’, Economic History Review, forthcoming.

Zaheer, A. and N. Venkatraman (1995), ‘Relational governance as an interorganizational strategy: an empirical test of the role of trust in economic exchange’, Strategic Management Journal, 16, 373–392.

1 Emphasis added.

Page 43: Strategy, Governance and Industrial Organization: Historical Perspectives on … · 2007-07-12 · Strategy, governance and industrial organization: Historical perspectives on the

42

2 It also follows from this brief critique of Porter that investment in external economies is the only logical source of national

competitive advantage in a world populated by MNCs. However, further consideration of this proposition is beyond the scope of

the present paper.

3 The view is an extension of Marshall for whom the market remained pre-eminent as a mechanism for allocation and co-

ordination, the ‘specialization’ found in clusters being ‘thoroughly effected without conscious effort’ (1919: 601), firms in

districts being ‘welded almost automatically into an organic whole’ (1919: 599).

4 Significant numbers of chemical firms were to be found at a number of sites close to the River Mersey and the great port city of

Liverpool, including St Helens and Runcorn, but these towns were at no great distance from one another and by the latter half of

the nineteenth century may be thought of as a single cluster centered on Widnes.

5 The UAC also contained significant numbers of firms from two other clusters, one in Northeast England and one in Scotland.

The Northwest (Widnes) cluster had however long been dominant. Indeed interactions between these three regional clusters had

long been a central element in the extra-regional linkages that formed such an important part of the governance arrangements in

Widnes.