File: Strategy Ch 6 Costs and the boundaries of the firm v4.doc
Date: April 2005
Words: 16,000
Chapter 6 Costs and the boundaries of the firm
Learning objectives ..................................................................................................... 2
Keywords .................................................................................................................... 2
Introduction ................................................................................................................. 2
1. The boundary of an organisation ............................................................................ 3
Figure 6.1 A set of supply chains............................................................................ 3
2. The transaction costs theory of the firm. ................................................................ 4
Table 6.1 Total costs of market exchange and of in-house supply ......................... 5
3. Costs and market transactions. ................................................................................ 7
3.1 Opportunism ..................................................................................................... 8
3.2 Limited ability to acquire and process information .......................................... 8
3.3 Asset-specificity................................................................................................ 8
Box 6.1 Asset specificity in the car industry ........................................................ 10
3.4 Asymmetric information ................................................................................. 11
3.5 Team production ............................................................................................. 12
3.6 Imperfect commitment .................................................................................... 12
4. Costs and firms. .................................................................................................... 12
Figure 6.2 Necessary relationships when engine manufacture is being done inhouse by a car assembler....................................................................................... 14
Figure 6.3 Engine manufacturing (by a car assembler) as a collection of
functional activities ............................................................................................... 14
5. Vertical integration. .............................................................................................. 15
Figure 6.4 The car production supply chain. ........................................................ 15
Figure 6.5 A vertically disintegrated car production industry .............................. 15
Figure 6.6 A partially vertically-integrated car production supply chain. ............ 16
6. Diversification....................................................................................................... 17
6.1 Costs and diversification: economies of scope. .............................................. 17
Figure 6.8 Economies of scope in the production of two goods ........................... 18
6.2 Diseconomies of scope ................................................................................... 24
6.3 Economies of scope and diversification: empirical evidence ......................... 25
Table 6.2 Potential cost benefits from scope economies in some European Union
manufacturing industries. ...................................................................................... 26
Table 6.3 Some evidence about the relationship between diversification and
performance .......................................................................................................... 27
Table 6.4 Overall diversification profile of Porter's sample of 33 companies ..... 29
Table 6.5 The acquisition record of Porter's sample of 33 companies ................. 30
7. Hybrid forms of integration. ................................................................................. 31
Figure 6.6 Governance structure, asset specificity and uncertainty...................... 33
Figure 6.7 Michael Porter’s 5 Forces Model of Competition ............................... 34
8. Conglomeration..................................................................................................... 34
Conclusions ............................................................................................................... 35
1
Learning outcomes .................................................................................................... 36
Further reading .......................................................................................................... 36
Web links .................................................................................................................. 37
Discussion questions ................................................................................................. 37
Problems ................................................................................................................... 38
Case Study: Information costs and the information economy .................................. 38
Box 6.2 Brokers and the UK insurance industry ................................................. 39
Learning objectives
Keywords
Adverse selection
Asset specificity
Asymmetric information
Costs of contracting through markets
Economies of scope
Lock-in
Moral hazard
Search processes and search costs
Transaction costs
Uncertainty
Vertical integration
Introduction
Our central theme in this book is how the strategic choices of an organisation can affect
its value-adding potential. In this part of the book, we are particularly concerned with
how strategic choices influence value through their impacts on costs. Gaining an
understanding of these impacts will provide insights into the ways in which one firm
might be able to consistently outperform its rivals through superior cost performance.
An important influence on a firm's costs is its 'architecture'. Architecture has two facets.
The first concerns the boundary choices made by the firm: the mix and scope of the
goods and services it produces, and the extent to which it deals with its suppliers and
customers externally (through markets) or internally (within a vertically integrated
organisation). The second facet of architecture concerns the firm's governance structure:
the relationship between ownership and control, the allocation of decision rights within
the firm, and the formal and informal contracts that exist among members of the firm,
including its structure of rewards. In this chapter, we investigate the first of these two
facets of architecture, its boundary choices. The second – an organisation's internal
governance structure – is discussed in Chapters 7 and 8.
2
Boundary choices do not only shape value through the dimension of costs. They may also
have important implications for revenue potential. However, we restrict attention here to
the cost impacts of boundary choices. Other chapters – particularly Chapter 3 (on
demand), Chapter 14 (corporate strategy) and Chapter 15 (on the methods of attaining
corporate objectives – will deal inter alia with how revenue potential may be influenced
by boundary choices.
The principal way in which we shall gain insight into the cost implications of alternative
boundary choices of firms is by applying the transaction costs economics (TCE) theory
of the firm. According to the TCE theory, boundary choices will depend on the relative
costs of organising production and exchange through market transactions as compared
with the costs of managing and co-ordinating production and exchange within those
institutional structures that we call firms.
The TCE framework can help us think about the best way of making boundary choices.
Why do some transactions take place between firms and others within firms? What
determines the choice of the firm's scope: the mix and range of products it produces?
1. The boundary of an organisation
The notion of the 'boundary' of an organisation is a somewhat nebulous concept,
amenable to a number of interpretations. To motivate ideas, and to explain the sense in
which that word is being used here, consider Figure 6.1. This shows the 'supply chains'
for three production processes, the production of petrol, cars and motor cycles. What is
shown here could be regarded as an extract from the entire set of supply chains of a
whole economy. A supply chain is simply a description of the set of technically separable
stages into which we can decompose any production process. Moving down from top to
bottom, the chain shows 'upstream' extraction and processing of raw materials, the
outputs of which are used as inputs into subsequent production stages. Outputs of these
stages become inputs into later 'downstream' production destined for final consumption.
Insert Figure 6.1 near here. Caption:
Figure 6.1 A set of supply chains
Figure 6.1 is just a technical description of the various activities that are involved in the
production of some final products, and the sequences in which they are usually carried
out. As things stand, it tells us nothing about how these activities are actually organised.
For example, do firms typically operate in only one of these boxes or in several? If the
latter is the case, are there observable patterns about which kinds of activities are
integrated within a single organisation? And if there are such patterns (so that
organisational forms are not simply random or case-specific) can those patterns be
explained in terms of their superior cost performance?
To establish some nomenclature, consider matters from the point of view of a business
whose core activity is car assembly. We suppose that the company is currently highly
focussed on this core activity. It makes no products other than cars, it buys or rents its
3
materials, components, and other specialised inputs through market transactions with
outside suppliers, and it sells its assembled cars on to independent distributors. Moreover,
this car assembly business is a stand-alone business: it is not a part of any broader
corporate organisation.
This highly focussed organisational structure is only one form that the car assembler
could choose. It could increase its scope by becoming involved in activities beyond its
core activity (for example, the car assembler getting involved in producing sheet steel, or
in the production of motor cycles). Extending the scope of the firm is also known as
increasing its degree of integration. Alternatively, we could describe a firm whose scope
is greater than just a single core activity as being a diversified firm.
Our car business could enlarge its boundary (or, equivalently, extend its scope, integrate
further, or become more diversified) in several ways:
1. Vertical integration, along its supply chain.
a. Upstream vertical integration
b. Downstream vertical integration
2. Horizontal integration
a. Acquire or merge with another existing car assembler
b. Geographical diversification (international growth)
c. Product or customer related diversification
i. product range extension
ii. broadening portfolio of business products offered
d. Technology related diversification
3. Conglomerate (product or customer unrelated) diversification
See pictures representing some of these possibilities (Figure 6.1b through to Figure 6.1e.)
Examples:
Vertical integration (upstream): car assembler, producing own engines and/or body
panels.
Technology related diversification: the firm using its technological expertise to develop
products which take it into new areas of activity, such as ICI in chemicals and plastics.
Customer related diversification: a firm exploiting its “brand name” to enter new
activities as when Marks and Spencer moved into financial services.
International diversification: many diversified firms are also international in scope
(Hanson, Philips).
Unrelated diversification: BAT Industries: cigarettes to financial services. But it is
sometimes not clear to an observer that there is relatedness. Perhaps a case of business
related diversification.
2. The transaction costs theory of the firm.
4
In this section, we consider a theoretical framework that can be used to help identify the
efficient boundaries of firms. Production activity could in principle be organised in either
of two extreme ways. A highly specialised world consisting of lots of simple specialised
producers involved in a single activity such as car assembly. Or a highly organised world
where all activities are integrated into a single giant enterprise. In practice we have
neither of these extremes but a world in which there is a great variety of firms. Many
specialised producers, some vertically integrated firms, some moderately diversified
firms, and some highly diversified, conglomerate firms. Is the organisation of production
the product of purely random forces or is there an economic efficiency logic involved?
The transaction costs economics (TCE) theory of the firm provides one such logic.
Consider a set of related economic activities. Undertaking and coordinating those
activities involves two types of costs:
1. The production costs associated with the set of activities. These consist of the
costs of raw materials, energy, component parts, and labour and capital inputs
employed in the activities.
2. Transactional or organisational costs. If the various activities are undertaken by
separate organisations, market exchanges will be necessary among those
organisations. Associated with market exchange are a number of costs that we
shall describe at some length below. If the activities are carried out within a single
integrated organisation, there will be a variety of costs incurred in supporting,
administering and coordinating those activities.
In essence, TCE theory suggests that the institutional mechanism chosen will be that
which minimises the total costs – production costs plus transactional or organisational
costs – of undertaking and coordinating those activities.
Consider a car assembler making a decision as to whether to continue buying engines
from an outside supplier (and so to remain an assembler only) or to make them in-house
(and so to vertically integrate upstream). We lay out one possible schedule of the
expected costs involved in each of the two routes in Table 6.1.
Insert Table 6.1 near here. Caption:
Table 6.1 Total costs of market exchange and of in-house supply
Market sourcing
Production costs
Suppliers own organisational
In-house sourcing
100
Production costs
120
50
Assemblers own
80
and transaction costs
organisational costs
(incl. capital costs)
Profit
10
5
Market price
Transaction costs
Total costs
160
50
Total costs
210
200
It might be the case that the production costs of doing things in-house would not be
greatly different from the production cost incurred by an outside supplier. However, in
our numerical example we suppose that unit productions costs are smaller when made
externally, perhaps because of larger scale or some other specialisation advantage, or
perhaps because of superior learning that would take a substantial time for the assembler
to replicate. We have ignored development costs in this example, although spreading of
these over larger volumes would favour the outside supplier.
For our illustrative numbers, external supply is favoured if production costs alone are
considered. But there are other costs to consider. If the engines are secured through
market exchange, the price to be paid by the assembler would also involve two additional
components . The first of these is a margin which reflects the supplier's own transactions
and/or organisational costs, and which it will build into its selling price. The second
component is the supplier's operating profit margin. At a minimum this must cover the
supplier's cost of capital; it may exceed that if the supplier has sufficient bargaining
power in its relationship with the assembler.
The price of purchased engines (160) still understates the total unit cost to the assembler
because the latter will have to incur a variety of transaction costs assocaited with market
exchange. Summing up productions and transaction costs, we arrive at a unitised cost of
210 per engine.
This clearly exceeds the assemblers own production costs (120), even with its relative
productive inefficiency. But producing in-house will nvolve the assember in bearing its
own organisational costs. If these were 80 per unit, then total costs per engine made inhouse (200) are lower than obtaining them through the market, and so the former
arrangement would be chosen.
We make no claims about realism of these numbers. The point is though to show in a
simple manner the nature of the cost comparison. What is particularly important is to see
which magnitudes matter. For example, the lower are trnsactional costs, the more is
market excahnge favoured. Indeed, if the processes of trade and exchange could be
arranged at zero cost to the parties involved then market-based exchange would in most
cases be superior to in-house production. However, the process of exchange generally
does involve costs, over and above the agreed price of the good being exchanged. For the
simplest kinds of exchange these transaction costs are likely to be negligible. For others,
such as hiring the services of an advertising agency, they may be large. For very
6
complex transactions involving contracting with many separate parties, such as arranging
the time-sequenced supply of complex components, these costs can be very large.
Let us now investigate the underlying factors that shape the relative magnitudes of the
costs of working through markets and through hierarchies. Section 3 below considers the
costs involved in market transactions. Section 4 considers the costs of undertaking and
coordinating activities within a firm. To provide a specific example of the general points
covered in Sections 3 and 4, we provide in Box 6.2 an analysis of the costs a car
assembler would incur in sourcing car engines externally through market transactions as
compared with the costs of doing engine manufacture in-house. It is recommended that
you now read through Sections 3 and 4 quickly, and then study Box 6.1, before returning
to Sections 3 and 4 for a second, more careful read. Try to find an example of each of the
points raised in Sections 3 and 4 in the boxed case study.
3. Costs and market transactions.
Ignoring the purchase price of the traded items, the costs involved in using markets are of
three types:
 The costs involved in searching for information about prices and suppliers. A car
assembler wishing to purchase engines, for example, needs to find out about
availability, prices, and quality through search processes. Obtaining good information
about quality, in particular, is likely to entail very substantial and costly search efforts,
particularly if the assembler requires 'bespoke' engines rather than standardised 'offthe-shelf' products.
 The costs involved in negotiation and bargaining, and in drawing up and
renegotiating contracts. When the assembler has identified several possible firms for
its engine sourcing, a process of negotiation and bargaining will follow. This might be
very resource and time intensive if its needs are specific.
 The costs involved in monitoring, policing and enforcement of contracts. Once
contracts have been signed, resources must be devoted to ensuring that the terms of the
contract are adhered to (on both sides). Quality must be monitored and controlled.
Procedures are required to deal with foreseen and unforeseen contingencies that might
arise, and perhaps for the enforcement of contractual agreements or payments of
compensation.
These kinds of costs are known as ‘transaction costs’. The Palgrave Dictionary of
Economics has defined transaction costs as:
A spectrum of institutional costs including those of information, of negotiation, of
drawing up and enforcing contracts, of delineating and policing property rights, of
monitoring performance, and of changing institutional arrangements. In short, they
comprise all those costs not directly incurred in the physical process of production.
What factors give rise to transactional costs? They arise from the interaction of two
factors:
7
(a) uncertainty in a complex and dynamic environment
(b) the self-seeking aspects of human behaviour
These two factors generate a number of consequences:
3.1 Opportunism
Opportunism refers to the propensity of individuals not only to pursue their own interests
but to do so in deceitful or guileful ways. Opportunism means that individuals are not
completely trustworthy and are likely to misrepresent situations and intentions to their
advantage. Faced with the possibility of opportunistic behaviour, contracting becomes
complex, problematic and costly, as individuals use formal written contracts, build into
them extensive contingency conditions, and establish appropriate monitoring
arrangements.
3.2 Information distortions
If people are self-serving, it also likely that individual intentions and relevant information
will not be fully and unbiasedly revealed. The costs associated with these kinds of
information-flow failures are explored in the following two chapters.
3.3 Limited ability to acquire and process information
The combination of uncertainty and self-seeking behaviour, together with the limited
cognitive capacities of human beings, means that the world in which firms operate is
necessarily complex and, in some sense at least, unknowable.
It is a reasonable working assumption that individuals are rational, at least in the broad
sense that they use the information available to them wisely. But in reality our capacity
for optimal decision making is limited or bounded by our capacity to acquire, store and
process relevant information. Moreover, the presence of guile implies that some of the
information available to us is wilfully misrepresented. These factors limit our ability to
develop fully specified, or complete, contracts. The future is too unpredictable and too
complex for us to anticipate and compute all possible events that might affect a business
relationship over its lifetime. As we cannot even imagine all the possible contingencies
that might arise, many contracts will necessarily be incomplete. This implies a possible
need to continually re-negotiate the terms of the contract as circumstances change in an
uncertain environment. Moreover, bargaining power might be very unequally distributed
in these circumstances
3.4 Asset-specificity
If productive assets were generally very good substitutes for one another they could move
from one use to another in response to price signals without losing any of their market
value. But many assets are highly specialised: they are specific to a particular use or
transaction. For example a newspaper printing press can’t easily be used to print books
and a car assembly plant can’t easily be turned to assembling computers. Specific assets
8
are those that have low value in alternative uses so that moving from one use to another
causes losses to the owner. These assets have been developed with a specific use in mind
and once developed or put in place have a value in that use much higher than in any
alternative use.
This is a problem because it creates a ‘lock-in’ effect. The parties to a transaction
involving the creation of specialised specific assets are mutually dependent on one
another. In this situation each party may behave opportunistically, trying to take
advantage of the other by attempting to obtain more favourable terms than had been
agreed before the asset was created. This is known as ‘opportunistic re-contracting’. An
example of asset specificity in the production of cars is given in Box 6.1.
The second reason concerns asset specificity. It is sometimes thought that if a company
requires highly specialised components, purpose-built for its own products, it must
manufacture them itself as no outside supplier would be willing to offer such bespoke
items. But this is patently wrong. Market contracts of this kind can be, and are, entered
into. But what becomes clear after a little inspection is that such contracting can be very
difficult. There are many reasons why it can be difficult and costly to arrange sourcing of
specialised inputs in a mutually satisfactory way through markets. These reasons include
the risks associated with one (or both parties) investing heavily in customer-specific
capital equipment, particularly where information is imperfect and one party might
behave opportunistically to exploit positions the other party’s ‘lock-in' position. We
explain these ideas more fully below. Before doing so, let us consider another alternative:
making required inputs 'in-house'.
9
Insert Box 6.1 near here
Box 6.1 Asset specificity in the car industry
Asset specificity is pronounced in the car industry. Car assemblers typically purchase
components, such as tyres, batteries and instruments from specialist suppliers. Very
often, an assembler requires components (such as car seats) to be manufactured to precise
specifications for a particular model in its range. Both sides make investments that can
only be justified if the relationship is likely to be a continuing one and involve a scale of
throughput sufficient to recover the large initial development costs. Clearly, pure market
exchange could be a very costly means of dealing with such transactional relationships,
and the risks involved from opportunistic recontracting by either side could discourage
valuable transactions.
Two of the ways in which the car industry has attempted to deal with the potential
problems of asset specificity warrant attention. One has been through greater vertical
integration (see our discussion in section x and the previous comments about the 'makeor-buy' choice). Large-scale assemblers often produce their own components rather than
purchasing them from independent specialist suppliers. By having all parties concerned
under the single authority structure of the firm, opportunistic recontracting can be
virtually eliminated. However, vertical integration can also be costly. Part of the firm’s
resource base is being directed to activities for which it may not have special capabilities,
and the advantages of specialisation are being missed.
In the last two decades car assemblers have been increasingly divesting their non-core
activities following the lead of Japanese companies in particular. To avoid the possible
adverse consequences that asset specificity brings about some major assemblers have
developed very close and long-term collaborative relationships with components
suppliers. These relationships try to get the benefits of vertical integration but at a lower
cost than would be incurred via complete merger. Patterns of close co-operation stopping
short of being internalised entirely within the firm have become critical pathways in the
process of value creation in modern business.
At the same time some component manufacturers are moving into final assembly of cars.
The Finnish component maker Valmet assembles cars for Porsche, Lada, and Saab, and
the Austrian company Steyr has been assembling for Chrysler and Mercedes. This is
partly an outcome of a policy of extensive outsourcing, extending even to final assembly
work, by the major car producers. In Japan it is said that the assemblers add only one
fifth of the added value in a car. Will we perhaps one day have a carmaker that makes no
cars?
End of Box 6.1
10
3.5 Asymmetric information
There are three kinds of reasons why the costs of sourcing car engines through markets
may be very large. One concerns imperfect information. The car assembler does not
operate in conditions of perfect information. Its business environment is uncertain, and
choices are risky. Not only is information imperfect, but the assembler and the engine
manufacturer do not have access to the same information. That is, information is
asymmetric.
Asymmetric information exists when one of the parties to a transaction is better informed
about certain aspects of the transaction than the other. The former cannot be relied upon
to disclose information honestly and the latter cannot find it except at high cost. This
asymmetry leaves one of the parties vulnerable to exploitation by the other. Think about
buying a second hand car: the seller is unlikely to tell you about its failings and you
cannot determine them precisely without paying for a full assessment. Asymmetric
information takes two forms which give rise to two separate problems for transactors.
 Hidden information. This is common in labour markets. Employees know more about
their real abilities and capacity for hard work than potential employers. It is a
particular problem in the insurance business. The sort of people who buy health
insurance, for example, will not be a random sample of the population but rather will
be those who judge themselves most likely to need the insurance. What happens when
an employer sets a wage reflecting the average abilities and capacities of individuals
or an insurance company sets terms reflecting average propensity to ill health?
Individuals who know they are above average workers or below average health risks
refuse the offer. Individuals who are below average workers or above average health
risks find the offer acceptable. The employer or the insurance company faces an
‘adverse selection’ problem. Poor workers and bad risks drive out better workers and
good risks.
 Hidden actions. It is sometimes difficult or impossible to be sure when the other party
has fulfilled her contractual obligations. For example when a worker’s productivity is
below expectations can you tell what is going wrong? Was the worker slacking, or was
his machinery playing up, or was the material being worked of poor quality, or was the
factory too warm that week? Often it will be difficult to tell, and the worker is unlikely
to admit that the problem was his effort level. Similarly, if you take out an insurance
policy on your car it is difficult for the insurance company to be sure that you will
always drive carefully and keep it locked when parked. Indeed the very fact you are
insured makes you somewhat less careful than you might otherwise be. This general
problem of verifying the reasonableness of actions taken within a contract is known as
the moral hazard problem. Moral hazard is a form of opportunism that arises when the
actions desired or required under a contract cannot be easily observed or measured.
The result is to discourage trade through market exchange.
11
3.6 Team production
When many individuals are involved in a particular joint activity but their separate
contributions to total output cannot be easily evaluated we have what is called a team
situation. This creates a measurement problem because if it is not clear who is
contributing what, then two issues arise. First, how do you reward the individuals on the
basis of their individual efforts? Second, if you cannot reward individuals on the basis of
their own efforts what is to stop individuals from free-riding on the efforts of others?
Even agreeing to split the outcome of the team’s efforts equally does not overcome the
problem. Indeed individuals who are promised the same share as everyone else have the
same incentive as everyone else which is to do as little as possible. How then do these
individuals transact with one another to get the job done?
3.7 Imperfect commitment
This refers to the fact that the parties to a contract are often unable to give totally credible
commitments about their future behaviour. Thus all drivers will promise their insurers to
drive carefully and so on. But insurers know that these promises are not credible and
cannot be relied upon. Individual drivers would gain if they could make credible binding
commitments but achieving such commitment is likely to be difficult and costly.
4. Costs and firms.
Having defined some concepts associated with complexity and self-seeking behaviour, let
us now turn our attention to the organisation of economic activity within integrated firms.
Our previous outline of what is entailed in producing car engines in-house made it clear
that this process also involves costs. We label these as organisation costs. To get some
idea about what kinds of costs might be involved, look at Figures 6.2 and 6.3. The first of
these depicts some of the relationships, both internal and external to the firm, that owners
of a car assembly firm (or managers acting as their agents) must enter into if the firm is to
build its own engines. Figure 6.3 reproduces that information in terms of a set of
functional activities that support car engine manufacturing.
It is important to be clear that in writing about 'organisation costs' we are concerned with
all those costs that are not directly associated with the production process. So, for
example, organisation costs do not include the costs of purchased materials themselves,
nor the cost of energy. But we do include the costs of building purchasing capacity, such
as information systems and purchasing staff. We do not include the cost of land and other
property purchased, but we do include the costs of property management. And we do not
include the direct cost of labour as such, but do include the costs of contracting with staff,
and monitoring and administering the way employees work.
These organisation costs could also be thought of as transaction costs, but with the
difference that they are being incurred within firms. But what is it that makes these kinds
12
of activities costly at all? The factors giving rise to organisation costs are more or less the
same as those we listed above in discussing the underling causes of costs of market
exchange. That is, uncertainty, asymmetric information and opportunistic behaviour.
Jensen and Meckling (1996)? Describe the firm as a 'nexus of contracts'. Owners of
firms, or their agents, need to search, make contracts with other members, and monitor
performance. In doing so, they also face the problems of opportunism, bounded
rationality, asset-specificity, asymmetric information, team production and imperfect
commitment that we described above. It is difficult or impossible to construct complete
contracts. It is also difficult to design incentive mechanisms that align the interests of
employees with those of the owners of the firm.
What is entailed in making car engines in-house? As before, the assembler must predict
how many engines it will need. But now it will also be necessary to install the necessary
productive capacity, and to pay the costs of planning, managing and supervising the
operation of that capacity, and of dealing with any associated contingencies that arise
through time. The firms will still need to carry out various search processes, but this time
for the components and materials of car engines rather than for fully assembled car
engines.
The costs that are of particular importance here are largely the various human resource
costs of doing things in-house. The firm will need to employ staff to undertake engine
manufacture and to carry out the associated support functions, and it will need to install
an additional managerial team to plan, supervise and coordinate the activities of its
workforce. Contracts will again be necessary, but this time with individuals within the
firm rather than with external parties. Again this may necessitate processes of negotiation
and bargaining, although much of the activities of employees might be directed
administratively rather than through explicit contracts. Many of the activities required in
the buying-in option are needed here too, but in a very different guise. For example,
quality must be monitored and controlled: but in this case, it is the quality of the firm's
own output, rather than that of supplied components, that must be monitored. On the
other hand, some costs are without any clear equivalent. For example, resources will need
to be devoted to ensuring that the efforts of managers and other employees are aligned
with the goals of the firm, and to dealing with unforeseen contingencies that might arise
within the firm, or in the state of market demand for its final product.
Once again, these 'organisation costs' are likely to be very substantial in an uncertain
world, and one in which humans are self-interested and opportunistic. Foremost among
these are the risks the firm necessarily takes in installing in-house capacity based on
forecasts of an uncertain future. This is particularly relevant where the costs of building
in-house capacity are sunk (that is, they are wholly or in large part unrecoverable).
Several of these issues are addressed in Chapters 7 and 8, and fall under the general heading of costs
associated with principal-agent problems. We shall defer further consideration of them until later.
Suffice to say that organisation costs are likely to be very substantial in practice.
Insert Figure 6.2 near here. Caption:
13
Figure 6.2 Necessary relationships when engine manufacture is being done in-house by a
car assembler
Insert Figure 6.3 near here. Caption:
Figure 6.3 Engine manufacturing (by a car assembler) as a collection of functional
activities
This suggests that we may explain the form in which a transaction takes place in terms of
the relative costs of undertaking that transaction through market exchange mechanisms as
compared within firms. Where a transaction has a cost advantage through marketexchange, one would expect that mechanism to be selected. Conversely, if the relative
advantage is in favour of a firm based transaction then that is the means we should
observe.
The advent of the information technology revolution, and of the internet in particular, may have
substantially reduced the costs of some forms of market exchange. (See the Case Study at the end of
this chapter). Indeed, in some cases it has appeared to be responsible for creating new markets
where previously none existed because search and information transmission costs were so high as
to have prevented the existence of such a market. The auction business e-Bay seems to be a case in
point. If this is indicative of a deeper trend, we might expect to see more unbundling of firms and
greater disintermediation in the future. (Again, see the case study for some examples.)
But, more generally, there are many circumstances in which the nature of the firm as an
organisational unit means that it may be able to deal with transactions problems in a relatively
effective way. For example, its ‘permanence’ and its internal authority structure means that the firm
may be less exposed to some types of opportunistic behaviour. Also, a firm should be able to
benefit from economies of scale in contracting, thereby lowering overall costs. More
generally, firms are likely to build superior abilities in organising and managing multiple
contracts than is possible via pure market exchange. The food and clothing retailer Marks
and Spencer, for example, has developed organisational skills in order to achieve tighter
quality control over the food and clothing it purchases than would be possible at the same
cost for more fragmented contractors. This creates value for both Marks and Spencer and
its customers.
Firms may also be relatively efficient vehicles for promoting effective co-operation
amongst the owners of specialised inputs, building trust through continued association,
and so facilitating learning and the development of reputations. They may also be costefficient by allowing for change and adjustment to changing circumstances without
continuous and lengthy re-negotiation of individual contracts.
The creation of a firm does not eliminate opportunism, nor can it eliminate the costs of
organising and managing transactions entirely, but it might mitigate the extent of these
problems and reduce the associated costs. In other words, the firm may be a relatively
cost efficient mechanism for organising and co-ordinating economic activity. Moreover,
the information technology revolution can also enhance these advantages of firms, so it is
unclear what the overall effect of that technology has been and will be.
14
5. Vertical integration.
We remarked earlier that any production process can be thought of as a set of technically
separable stages. A simplified schema of the stages in the production of cars is shown in
Figure 6.4, most of which consists of just one of the three supply chains portrayed earlier
in Figure 6.1. The schema shown here is an incomplete in so far as it only traces
upstream one component, car engines. We have made this simplification in order to keep
our arguments simple. But note that by doing so, some other possible upstream linkages
that might exist (or be exploited) are ignored. Note that the arrows in this diagram merely
denote technical sequencing.
Insert Figures 6.4 and 6.5 near here. Captions:
Figure 6.4 The car production supply chain.
Figure 6.5 A vertically disintegrated car production industry
A vertically disintegrated motor industry would resemble that shown in Figure 6.5. Here
each activity is carried out by one or more independent business organisations. The
double-headed arrows now denote market exchanges, rather than technical flows along
the supply chain.
A word of caution is warranted here. Vertical integration is really a matter of degree
rather than of kind. All businesses are vertically integrated to some degree. To appreciate
that this is so, consider the following points. A firm is by definition an ongoing
hierarchical organisation; any resource allocation process except individual production
with pure market exchange brings together separable stages. Activities can be described
narrowly or broadly according to the purpose at hand. For example, we have constructed
a single box in Figure 6.5 that is labelled as consisting of ‘Iron ore extraction &
processing’. This was done as we have chosen not to focus our discussions on that part of
the supply chain. But one could easily argue that this box should be decomposed further,
as it involves at least two separable, but each quite broad, activities. Indeed, any one of
the stages shown in Figure 6.5 could be seen as a collection of vertically integrated
activities. If one were to zoom-in more closely that box would resolve itself into its own
'mini supply chain' of more narrowly defined, separable activities.
Having said this, the term ‘vertical integration’ is typically used to describe a situation in
which a business with one reasonably broadly defined core activity (such as car
assembly) acquires its own in-house capacity in other stages along the supply chain. This
acquired capacity may be ‘upstream’ (closer to the source of raw materials),
‘downstream’ (closer to the final customer), or both. For example, in Figure 6.6, we
portray a case in which the car assembler has become vertically integrated upstream by
taking in-house engine manufacture and distribution of engines, but has also chosen to
integrate downstream into outboard logistics.
15
Insert Figure 6.6 near here
Figure 6.6 A partially vertically-integrated car production supply chain.
Given our earlier remark that vertical integration is really a matter of degree rather than
of kind, and so that all all businesses are vertically integrated to some degree, the
important question is what determines the extent of this integration in any particular
instance.
The TCE theory argues that the answer must be found in the balance of cost advantages
and disadvantages that vertical integration confers. It is often argued that where two
activities are technologically inter-dependent, there will be net cost reductions from
integrating those activities within one organisation. For example, steel manufacture
involves (among other things) the conversion of iron ore into steel (in blast furnaces) and
the conversion of steel into commercially-usable sheets (in rolling mills). Clearly, those
two activities are technologically related, and an efficiency gain is possible if the
activities are linked.
However, technical interdependence alone is not a convincing argument for vertical
integration. As long as the two activities are operated in a spatially-linked way
(specifically, they are located next to one another), the efficiency gains would be
exploitable. But that would be true even if the two activities were undertaken by two
separate businesses that transact with each other through market exchange.
What is needed to explain vertical integration is not just the existence of a linkage but
also the likelihood of a cost saving if the two activities are combined in a single
organisation (a single firm). One important factor here is differences in the costs of
transacting through markets as compared with the integration of activities in a single
managerial hierarchy. More specifically, there are likely to be differences in the
transaction costs of these two routes, and firms will tend to choose the method which
minimises those costs. That is, organisations will tend to be structured so that transaction
costs are low.
Go through standard arguments for vertical integration here.
See MMP notes?
This brief summary, although incomplete, gives some insight into the nature of the choice
our firm faces. Clearly we cannot predict what the firm would do without additional,
case-specific, information. What can be said, though, is that in principle at least a firm
would be expected to choose that option which minimises the total of the transactional
and organisational costs we have indicated here.
16
6. Diversification
6.1 Costs and diversification: economies of scope.
Even in a quick reading of the management literature one can find many 'motives' for
diversification. Careful study, though, reveals that the logic of many of the commonly
suggested explanations for diversification is doubtful. For example, it is sometimes
argued that firms diversify to stabilise earnings or reduce shareholder risks. But it is
difficult to see how this might add value to an organisation. Shareholders can easily build
a private portfolio of shares in specialised firms with different risk/return characteristics,
or invest through managed funds, to achieve their individual desired risk/return
objectives. Moreover, they can usually do so at costs below those that could be achieved
by complex diversified firms.
The question we should ask about diversification is how, if at all, firms can create value
through diversifying. More specifically, why might a complex firm, made up of a number
of distinctive simple business units, say one making bread and an other bricks, add value
in excess of that created by the independent simple firms? Formally, why should V(1+2)
exceed V1+V2, where V represents firm value?
Value creation of this kind would come about if diversification allowed the complex firm
to achieve either higher prices or lower costs. As far as higher prices are concerned, it
might be possible for a firm to extend the exercise of its market power in one area to
another area through diversifying. Another possibility is that a firm could use a brand
name developed in one activity to charge more than an independent firm could in another
activity. But these arguments lack substance. For example, firms such as Sony or Glaxo
undoubtedly benefit from their brand names when introducing new products as long as
they “stick to the knitting”; that is, as long as Sony sticks to consumer electronics and
Glaxo to pharmaceuticals. It seems highly unlikely, however, that Sony could charge
premium prices for new products if it moved into pharmaceuticals, or Glaxo if it moved
into electronics. So the benefits of brands and reputations are unlikely to operate very far
from a company’s core activities where its reputation was built. Thus BAT’s name in the
world of cigarettes seems unlikely to have counted for much when it moved into financial
services.
To identify what might enable firms to add value through extending scope we need to
focus on the possibility of lower costs. Economies of scope occur where a firm that is
involved in two or more separable activities, such as bread and bricks, can achieve lower
overall costs than would arise if two or more separate firms carried out the same
activities. We define an economy of scope in the following way. Let x and y denote the
output of two distinct goods. C (x, 0) is the total cost of producing good x alone [i.e. with
zero production of y] and C (0, y) is the total cost of producing good y alone. Let C (x, y)
denotes the total cost of producing x and y jointly by one firm. An economy of scope
exists if C (x, y) < {C (x, 0) + C (0, y)}.
17
Figure 6.7 uses the idea of a production possibility curve to represent the concept of
economies of scope. Consider a fixed stock of resources that can be used to produce cars,
lorries or both. The straight line connecting A and B describes the various combinations
of cars and lorries that can be produced with the available resources where there are no
economies of scope. The negative slope arises because only switching resources from car
to lorry production can produce more lorries. Along the line connecting A and B the
opportunity cost of lorries - the number of cars that have to be given up to produce an
additional lorry - remains constant.
The outward-bowed curve ending at points A and B describes the production possibilities
where there are economies of scope. Points towards the centre of the curve represent
outcomes where the resources are used in joint production of both goods. Imagine that
the resources are divided equally between two firms. One firm uses its share to produce
cars only, obtaining an output of ½A; the other uses its share to produce ½B lorries. The
sum of the two outputs is shown by point C. However, if all resources are given to one
firm which can obtain scope economies by producing cars and lorries jointly, a
combination such as D is obtainable, giving more of each good, and so a lower cost per
unit output.
Insert Figure 6.7 near here. Caption:
Figure 6.7 Economies of scope in the production of two goods
What are the sources of these economies? Previous sections of this chapter have given
some insight into possible sources of cost savings when activities are integrated vertically
along the supply chain. We shall not repeat those arguments here. Instead we look for
sources of cost saving by integrating horizontally, creating organisational structures that
span different supply chains.
Two lines of reasoning appear to be pregnant with interesting possibilities. The first of
these concerns the benefits to be gained from linking related activities. The second
concerns the sharing of resources or capabilities across different activities, or the
exploitation of under-utilised or spare resources. We shall see that the presence of one or
both of these is a necessary condition for cost reduction from integration. But neither is a
sufficient condition. Sufficiency also requires that potential benefits of linkage and/or
sharing of resources are better obtained through integrated organisations than through
market exchange.
6.2 Economies of scope from exploiting linkages
In this section, we explore the idea that there can be economic advantages from bringing
together (integrating) within one organisation the production of more than one good or
service when there is some kind of linkage or interrelationship among the different
products, either on the demand or the supply side.
18
Note that a rather broad spectrum of possibilities is encompassed within the phrases
'more than one good or service' and ‘different products'. At one end of this spectrum, we
have differentiated goods. As have seen in Chapter 3 (on demand) there are good reasons
for differentiating a product from those offered by rivals. However, we often observe that
firms not only offer a differentiated product but that they also produce a range of
differentiated products with different configurations of characteristics or attributes, but
where substitution possibilities among alterative products are quite high. Moving further
along the spectrum, we find cases where a firm offers a variety of quite distinct products,
such as cars, motorbikes, vans, and small trucks, and where substitution possibilities are
very limited.
Towards the other side of the spectrum we have sets of products that are not substitutes
for one another at all, such as the set consisting of financial services, retail clothing and
food retailing. Even where products appear to be entirely unrelated, closer study might
reveal linkages among them.
Links of this kind could arise through supply relationships, whereby it is cheaper to
produce goods jointly than singly. The most commonly discussed case is that of cost
complementarity. This describes a situation where in the production of two or more
goods the average cost of making one falls as the scale of production of another is
increased. A well-known case of cost complementarity occurs in the petro-chemicals
sector. Crude oil is a necessary source from which a number of distinct products are
derived. If the production of the set of products of crude oil is co-ordinated, this will
economise on the use of inputs compared with the case where individual products are
produced independently of one another. Another example of this scope effect is in the
manufacture of cars. Car companies typically produce a sizeable range of different
models using common components, capital equipment, assembly structures and routines.
The use of a common process or common resource in this way reduces the total cost of
the set of goods by spreading the cost of some ‘common’ activities over a variety of
products.
The common resource or process in question need not be tangible. For example, research
and development teams may generate knowledge that can be applied in innovations over
a range of different products. Some car and motorcycle manufacturers, for example,
maintain factory racing teams; materials innovations and design knowledge developed in
racing can be applied in mass-market vehicle production. The backup of a large vehiclemanufacturing base can offer consider reciprocal benefits to racing teams. More
generally, distinctive skills acquired by firms can be applied to other technically-related
areas, such as the production by single firms of various types of metering equipment or
electronic components.
Many examples of technical complementarity of this kind can be found, and the cost
reductions associated with exploiting these scope economies are often substantial. Indeed,
the organisation of some industries - for example, chemicals and oil processing - can be
largely explained by the responses of firms to this type of economy of scope.
19
Alternatively, the links may operate on the demand side, through customer preferences.
For example, individuals who value one-stop shopping might gain value from purchasing
books, DVDs, and consumer electronics jointly from a single internet site. Many web
businesses, such as amazon.com, offer sophisticated browsing and recommendation
facilities, and so can benefit by bringing together products that customers typically
purchase after browsing processes
Such linkages can sometimes make it advantageous to ‘bring together’ otherwise separate
business or product activities. The net benefits obtained in this way are a form of
synergy, known as economies of scope. The degree to which such potential relationships
are exploited is one of the major drivers of a firm’s cost performance.
But we need to be cautious at this point. There is no doubt that cost reductions can be
achieved by coordination (linking) of activities that are related through supply or demand
factors of the kinds indicated above. But coordination can be achieved in various ways.
Clearly, co-ordination can be achieved by integrating the activities inside one
organisation. But that is not the only way in which it can be achieved. A second way is
through exchange relationships (contracts) between independent business units. For
example, a crude oil refining company could locate its production facilities adjacent to
independent firms that purchase its particular outputs to produce specific petroleum or
chemical products. All necessary cost-complementarities are attainable by co-ordinated
location and market exchange processes.
So it is not the case that technical (supply-side) or demand-side relatedness necessarily
implies the desirability of diversified firms. Diversification is the best response only if
that involves lower total costs than do the other feasible responses such as market
exchange. We pursue this point further a little later. Right now, we explore the notion of
sharing resources a little further, as this is the second principle reason given for
diversification.
6.3 Economies of scope from sharing of resources or capabilities, or from exploiting
under-utilised resources or capabilities
There are two arguments here that we state and then examine.
Using under-utilised resources or capabilities
A firm may have some excess capacity, taking the form of under-utilised or spare
resources or capabilities. The use of this spare capacity in some other product or
business represents a free resource to the firm. If a way can be found of utilising
this resource or capability appropriately, this can give the firm a cost advantage in
either or both of the original activity and the additional activity to which the spare
resource is applied.
Sharing of resources or capabilities
20
A firm may have the appropriate (profit-maximising) quantity of resources or
capabilities for its existing products or businesses, and so it does not have any
spare capacity per se. Nevertheless, it may be possible to share those resources or
capabilities so they can create value in other product or business areas without
detracting from the contribution they make in their "original" uses. Once again,
such additionality of resource or capabilities usage represents a free resource to
the firm. If a way can be found of exploiting that sharing capacity appropriately,
this can give the firm a cost advantage in the activities over which the sharing is
applied.
Transferring Knowledge and Skills
It is not difficult to find examples which apparently fit these two cases. Consider first the
‘spare capacity’ argument. Proctor and Gamble, for example, share a common
distribution system for paper towels and disposable diapers. In cases such as this, it may
be that the sharing arose because the firm in question had, at some stage in its history,
some excess or unused distribution.
If so, the first question we might wish to address is how these spare or under-utilised
resources came into being. This might have been the result of poor project planning. Or it
might have happened because of chance (with outcomes deviating from what was
expected when decisions were taken). In either case, the most appropriate action may be
to capitalise those excess resources by selling them. Alternatively, it might earn income
by hiring them out? Whether these options are cost-effective will depend heavily on the
transaction costs involved.
If those costs are large (for example, where assets are highly specific), and if the spare
resources are expected to be available over lengthy periods, a more attractive option
might be to use these assets to increase the scope of the firm.
It is sometimes argued that the existence of economies of scale in distribution systems (or
the like) favours the development of large, diversified firms. But this is a weak argument.
Although such technical economies are very likely, the benefits of large-scale production
could also be obtained by hiring out those services to others, or by entirely out-sourcing
those functions to specialised firms, and obtaining low costs by virtue of their suppliers
scale economies. So while it may be correct to claim, for example, that General Electric
obtains substantial economies of scale, other reasons must be given before one should be
convinced that such economies are best realised in-house within a diversified, integrated
firm.
Let us now turn to the resource sharing argument. Of central importance in most sharing
arguments is that the resource or capability which is being shared has the non-rivalry
characteristic of what economists call a ‘public good’. That is, the consumption of it by
one user does not reduce its availability to others. As we shall see, scope economies are
common in the presence of public goods. In effect, this property creates a generalised
spare capacity for the resource. For even if the resource is being employed at exactly the
21
right level in one business area, the public goods property implies that it can be used
elsewhere too without detracting from its use in the first-mentioned business. This is a
very powerful argument for sharing, and it suggests that the strategist should look very
carefully for any resources or capabilities that possess this property.
Many human resource (or “human capital”) assets possess the public good capacity that
permits sharing. Think about the following trivial example. An investment bank employs
a forecaster to predict market trends to support its lending activities. A car manufacturer
also requires similar forecasts. A firm undertaking both activities need only employ one
forecaster, whereas if separate firms undertake the activities, each may find it necessary
to employ a forecaster. Another example of an asset that can generate scope economies is
managerial resources, as the skills embodied in those resources can sometimes be
effectively used over different products.
Leveraging managerial skills is a special case of value generation by transferring
proprietary knowledge-based skills such as production or logistical skills from one
business to another among similar ‘value chains’ in different business units. Examples of
firms which have diversified using the transfer of skills strategy are 3M and PepsiCo.
It is often observed that knowledge assets are commonly under-utilised. According to a
report in the Financial Times newspaper some observers believe that as little as 20% of
companies utilise their knowledge assets fully and effectively. A number of firms,
recognising the value of these knowledge resources, have responded with important
organisational changes: Dow Chemicals for example has appointed a director of
intellectual asset management, and several others have introduced regular knowledge
audits or valuations. Management consultancies (including Coopers and Lybrand and
Anderson Consulting) have developed software that allows many users to work together
on the same information.
Will knowledge sharing lead to integrated, and diversified, organisations? What may be
decisive here are two further matters:
 Is the knowledge embodied in individuals or is a characteristic of an organisation as
such?
 Can the knowledge be exchanged efficiently through markets?
Think about the forecaster example in the light of these two questions. If forecasting
knowledge can be readily bundled and sold through the market (to whoever might be
willing to pay for it), then the value from sharing that resource can be appropriated
without any need for integration per se. But there are many difficulties in arranging such
transactions. The first is quality assurance; unlike many products, knowledge quality
cannot be easily sampled prior to purchase, as the sampling process itself transfers the
knowledge. This problem may be averted if a knowledge seller can build a reputation for
high and consistent quality. But this does not address a second problem. Knowledge may
not be easily bundled in ways that limit what is transferred. So when knowledge relates to
routines or generalised know-how, selling that knowledge threatens to unlock or release a
cascading font of useful information. This is most likely when the knowledge in question
22
is not resident just in one or more identifiable individuals, but is an attribute of an
institution or organisation. Thus where market exchange is costly in these ways, it might
be better to exchange information internally within an integrated structure, so that losses
of valuable information to ‘outsiders’ cannot take place.
Another information resource that may be amenable to sharing is an organisation’s
reputation. This may be capable of being exploited to generate customer demand in other
lines of business, so conferring scope economies. For example, Marks and Spencer and
the Virgin group have both diversified into the provision of financial services,
presumably in the expectation that reputations developed in their core areas will benefit
the newer ventures. The cost advantage in question might then consist of reducing the
marketing expenditures that would otherwise have to be incurred to establish the
products’ availability and quality. As we noted above, though, reputational sharing does
not seem plausible when it takes place between different core activities.
Reputation, like knowledge, is often very costly to transact. As an exercise, try and
identify those costs. And then compare your thoughts with those we suggest at the end of
this chapter.
.
Sharing resources or capabilities: discussion
Michael Porter observes that “the ability to share activities is a potent basis for corporate
strategy because sharing often enhances competitive advantage by lowering costs or
raising differentiation.” But we have to be wary of jumping to unwarranted conclusions.
While sharing is often value enhancing, that does not in itself imply that sharing has to be
done within an integrated organisation. The major question that should be asked is
whether the potential added value from sharing could be capitalised in other ways and, if
they could be, whether these are superior to the integration route.
Unless distribution is a public good, the common distribution system argument seems
weak when looked at in this light. Hiring or sub-contracting distribution resources to
others is a relatively straightforward process, particularly when longer term contracts are
employed to reduce uncertainties. However, fears of being victim to the ‘hold-up’
problem we saw earlier in discussing vertical integration might lead firms to keep sharing
in-house.
Transaction costs for selling or hiring out resources are likely to be particularly high in
the case of intangible resources or capabilities that are embedded within firms rather than
merely in the individuals who work for them. Examples include brand names, reputation,
and organisational routines and capabilities. Managerial skills may also sometimes fall
into this category, particularly in the case where the managerial resources are public
goods. On the other hand, if using management time in one activity is at the expense of
its use in another activity (that is, management time is a private good) comparative
advantage arguments suggest that there may be disadvantages in departing too far from
managerial specialisation.
23
As with vertical integration, the integration of activities that are technically or
economically inter-dependent is a source of efficiency gain. That efficiency gain is
known as an economy of scope. But such an efficiency gain does not in itself necessitate
that organisational integration itself takes place. For example, it would be possible for
two legally separate entities, one producing lorries and another producing cars, to
collaborate and coordinate their activities to achieve the efficiency gain. In other words,
productive integration could take place without there being organisational integration.
So under what conditions would we expect there to be an organisational integration; i.e.
the two activities are brought together in one single firm? The answer must be when the
linkage of activities in legally separate units is economically difficult. For example, a
lorry producer and a car producer might be able to realise a (joint) cost saving by sharing
a single distribution system. One party could then contract with the other to use its
system, and negotiate a price for that service such that both parties benefit from the
sharing. Such a contract would be relatively easy to construct and cheap to administer.
But suppose that the shared resource is R&D activity. There are certainly efficiency gains
to be had from pooling R&D resources, but the contracting difficulties would be
immense, primarily because of the likelihood of opportunistic behaviour. So we would
expect to see integration within one firm when there is an efficiency gain and this could
only be obtained by non-market forms of collaboration.
The best basis for diversification is sharing activities. But the benefits of sharing
activities, or synergy as it is often described, have to be fought for because they do not
emerge spontaneously. A strategy based on sharing requires organisational mechanisms
that actively encourage sharing. Note therefore that the western preference for
divisionalisation of big firms, that is the creation of clearly identified stand-alone
divisions, is likely to be inimical to sharing, and that large Japanese firms are far less
prone to this particular organisational form for precisely this reason. Japanese firms have
identified the creation and sharing of capabilities and competencies as the basis of their
strategy. The Japanese symbol for the enterprise is the tree and the capabilities of the firm
are in the soil which feeds the whole tree and its many branches. Divisionalisation on
western lines would look to the Japanese like cutting the branches off from their source
of nutrition.
6.4 Diseconomies of scope
Can there be economic disadvantages from increasing the range of businesses or variety
of products that a single firm makes? In other words, can there be diseconomies of scope?
Three reasons suggest that this is possible in some circumstances?
 Activities are brought together for which little or no exploitable efficiency gains
are possible, and the additional organisational costs entailed in running a 'broader'
organisation outweigh potential efficiency gains.
 Substantial potential efficiency gains exist but there is failure to realise those
potentials.
24

Extending the scope of the firm (i.e. diversification) is undertaken for reasons
other than maximising the value of the firm. Principal among these reasons are
likely to be empire building tendencies of managers, and forms of the principalagent problem that we explain in the following chapter. If an economic
disadvantage is defined to be something that is firm-value destroying, then
diversification for these ‘other’ reasons is disadvantageous. However, from the
point of view of those who made the decision to diversify there may be net gains.
How likely are these situations? Some economists, taking a Darwinian view, believe that
they are unlikely. They argue that an activity can only persist in the long run if it
increases the wealth of the organisation. If this is correct, it suggests that while mistakes
might be made, the surviving examples are those who benefit. It would also rule out the
possibility of diversification being wealth destroying because of such things as
managerial empire-building tendencies.
Others are less sanguine. A substantial proportion of diversified organisations come about
through processes of merger or acquisition. It is often thought that the implementation of
mergers is beset by problems such as cultural differences and strategic behaviour in
which members of newly formed ‘coalition’ fight to preserve their original positions.
These interfere with, or slow down, the structural adjustments that are required if
efficient gains are to be realised. Secondly, many writers in the business strategy field
take the view that many, or possibly most, broad integrated corporations destroy rather
than add value, and so are more valuable broken-up into their constituent parts (or
unbundled as the business press puts it). For example a recent book by Sadtler,
Campbell, and Koch (1998) has listed a number of prominent firms in need of breaking
up including giants such as General Motors and Ford.
Each of these considerations implies that the Darwinian selection processes just described
work at best very weakly. If so, there could then be many organisations in which value
could be created by splitting-up multi-business or multi-product firms. But these cases
would have to be identified by careful case-by-case examination.
6.5 Economies of scope and diversification: empirical evidence
Many attempts have been made to study empirically the relationship between
diversification and firm performance. Most of these have used US data. Research in this
field is beset by several statistical and methodical problems that plague much of the
research in business and the social sciences, and one should bear in mind these
limitations when looking over the outputs of that research findings. Among the raft of
such difficulties, we select the following:
 Studies are based largely on samples of diversified firms at a point of time, such as the
1980 US Fortune list of the 500 largest US corporations. By definition those sampled
are survivors. Ideally, one would like to draw inferences from a random sample of all
diversified firms, both those which have survived and those which did not (because
they have gone out of business, or have been taken over themselves, or have
25
unbundled). When inference is drawn from survivors alone, there is always the
possibility of sample selection bias.
 There are serious definitional problems with diversification, and the idea is difficult to
operationalise empirically. It has proved difficult to distinguish between related and
unrelated types of diversification and researchers have found it impossible to develop
a continuous variable to measure the degree to which each company is diversified.
 There is a problem of causality. Suppose that diversification is positively associated
with good performance. There are (at least) two possible explanations. First,
successful firms are more likely to look for opportunities to diversify, hence causality
runs from performance to diversification. Secondly, diversification improves (and so
'causes' performance). Some studies have indeed strongly suggested that this is the
more likely direction of causality.
 The existence of profitable diversified firms does not in itself prove that
diversification per se is profitable. It may simply reflect the fact that the firm’s core
activities are highly profitable, as with BAT’s tobacco business. Diversification could
of itself destroy value, but that would not be apparent in the aggregate picture if the
core profitability is very high relative to diversification losses.
Given these difficulties, some are highly sceptical of any empirical evidence about
diversification. Nevertheless, we shall briefly summarise some of the research findings
available in the literature. An indication of the potential importance of scope economies
can be found in Table 6.2. This gives estimates of the potential cost savings from scope
economies in some European Union manufacturing industries. The ‘Economy of scope’
figures in the table refer to the percentage increase in average costs expected if each
producer in the industry in question halved the number of its models. Whilst many of
these estimates suggest modest gains at best, the larger numbers are far from being trivial,
and are of similar magnitudes to price-cost margins that are achieved in competitive
markets.
Table 6.2 Potential cost benefits from scope economies in some European Union
manufacturing industries.
Type of activity
Motor vehicles
Pharmaceutical products
Electrical machinery
Office machinery
Domestic-type appliances
Man-made fibres
Carpets
Machine tools
Cement, lime and plaster
Economy of scope
8
5
5
5
5
3
3
1
0
Source: Pratten (1987).
A further selection of evidence on diversification is summarised in Table 6.3
26
Insert near here, Table 6.3 Caption:
Table 6.3 Some evidence about the relationship between diversification and performance
27
Table 6.3 Some evidence about the relationship between diversification and
performance
Study
Rumelt (1974)
Sample
500 industrial companies,
1949-69
Main findings
Companies which diversified around common skills and resources
were most profitable. Vertically-integrated companies and
conglomerates were the least profitable.
Christensen &
Montgomery (1981)
128 of Fortune 500
companies, 1972-77 (subsample of Rumelt, 1974)
Companies pursuing related, narrow-spectrum diversification were
most profitable; vertically integrated firms least profitable. But
performance differences primarily a consequence of industry factors.
Porter (1987)
33 large US corporations,
1950-86.
For most of the companies studied, diversification strategies dissipated
rather than created shareholder value. A high proportion of companies
involved in his study sooner or later divested themselves of their new
activities. Some of the sample firms were very profitable but this was
often because a profitable core activity covered up a poor
diversification record.
Rumelt
273 of the Fortune 500
companies, 1955-74
Related diversification more profitable than unrelated even after
adjusting for industry effects.
Varadarajan
Ramanujam(1987)
225 companies, 1980-84
Related diversifiers earned higher return on equity and capital invested
than unrelated diversifiers
Markides (1992)
Sample of overdiversifiers
from the 1980’s
Positive returns from refocusing and reducing diversity
Luffman &
Reed(1984)
439 UK companies from
the Times 1000 1970-80
Conglomerates showed highest equity returns and growth of profits,
sales and profits
Grant et al. (1988)
305 large UK
manufacturing
Product diversity positively associated with profitability, but after a
point relationship turns negative.
.
Source: Based on an original table in R.M. Grant: Contemporary Strategy Analysis, Blackwell, Oxford,
1991.
We shall devote some attention to one of the studies listed in the table, Michael Porter's
study of 33 large and prestigious US companies over the period 1950-86. This is an
important study because it looks at diversification strategy over time and does not, like
most empirical studies of the issue, look only at the performance of currently diversified
firms. The overall diversification profile of these firms is given in Table 6.4. The extent
of diversification can be seen to be very high: on average, each firm entered 80 new
“industries” and over 27 entirely new “fields”. [It is important to note that Porter uses the
words field and industry in rather special ways. By field he means something such as
insurance which we would refer to in this book as an industry or a broadly-defined
market. Porter’s meaning of the word “industry” is much narrower than we use in this
text: he has in mind a more tightly specified grouping, roughly meaning a narrowlydefined product area.]
28
Insert Table 6.4 near here. Caption:
Table 6.4 Overall diversification profile of Porter's sample of 33 companies
29
Table 6.4 Overall diversification profile of Porter's sample of 33 companies
Average over
sample
114.8
80.1
70.3%
7.9%
21.8%
27.4
Number of total entries
All entries into new “industries”
Percent acquisitions
Percent joint ventures
Percent start-ups
Entries into new industries that represented entirely
new fields
Percent acquisitions
Percent joint ventures
Percent start-ups
67.9%
7.0%
25.9%
Table 6.5 focuses on the acquisition record specifically (as opposed to joint ventures and
start-ups), showing the percentages subsequently divested. On average, companies
divested over half of their acquisitions. In the case of ‘unrelated’ acquisitions (figures not
shown here) the percentage later divested was 74%! If the proportion of later divestments
is an indicator of diversification success, diversification in this sample seems to have an
abysmal performance. [We must be a little careful here, though. As Porter himself notes,
where the benefits from an acquisition come principally in the form of rapid one-off
gains, there is no incentive for the acquirer to hold on to the new unit once those gains
have been realised. Getting rid of the newly acquired unit is then a way of capitalising the
benefit, and not necessarily an indicator of failure.]
Insert Table 6.5 near here. Caption:
Table 6.5 The acquisition record of Porter's sample of 33 companies
Table 6.5 The acquisition record of Porter's sample of 33 companies
All acquisitions in new industries
Percent made by 1980 and then diversified
Percent made by 1975 and then diversified
All acquisitions in new industries in entirely
new fields
Percent made by 1980 and then diversified
Percent made by 1975 and then diversified
Sample average
61.2
53.4%
56.5%
20.0
60.0%
61.5%
Porter proposes three tests which any corporate strategy must pass in order to create
value:
30
 The attractiveness test: the industries chosen for diversification must be structurally
attractive or capable of being made so (this is what we have referred to above as a
necessary condition for diversification to be wealth-enhancing)
 The cost-of-entry test: any cost of entry into new activities must not capitalise all the
future profits (this is the point Peteraf makes about ex ante competition that we
explain later in Chapter x)
 The better-off test: either the new unit must gain competitive advantage from its link
with the new corporation or vice versa.
Taking the three bullet-pointed conditions together gives what we earlier referred to as a
set of sufficient conditions for wealth creation through diversification. The importance of
the latter two tests is that they correspond to the requirement that diversification is a
better way of gaining value from linkages than exploiting those gains in some other way.
If there was a better way of exploiting them – and some other firm had found that way –
then the integrated firm would be unable to outperform that rival. That is, it could not
attain a competitive advantage.
Porter finds that many firms ignored the attractiveness test, or simply got it wrong, often
mistaking short term gains for long-term profit potential. He also suggests that many
diversification ventures failed to satisfy the cost of entry test: when diversification was
achieved through acquisitions (rather than through internal growth) the acquirer often had
to pay not only what the object of its attention was worth, but a premium on top of that as
well. Where diversification took place through start-ups, high entry costs sometimes
eroded expected profits. Porter also suggests that it is common for managers to pay little
or no attention to the better-off test.
7. Hybrid forms of integration.
• Competition in the age of ecosystems
• Business ecosystems are economic communities built on a foundation of
interacting organisations and their customers and suppliers. Over time they coevolve their capabilities and roles and tend to align with the direction set by the
core company(ies). The function of the leader is to enable members to have a
shared vision, align activities, and be mutually supportive. Once established
difficult to dislodge. Wintel a prime example.
• JF Moore, ‘Death of competition’ book, 1996
Till this point we have concentrated rather heavily on binary choices: the firm or the market, make
or buy, and the like. Here choices must be made from one or another clearly specified alternative.
But in practice, organisational choices are usually matters of degree rather than of kind. They
are about whether there should be more or less integration along the supply chain;
whether more activities might better be outsourced, and if so which ones; and whether the
firm should broaden its product scope or range, or focus more narrowly on key areas of
strength. Matters of degree also enter in another way. Even where some particular
boundary extent has been selected, the degree of integration is not restricted to either no
31
integration whatsoever or complete integration in a single organisation. A host of
intermediate forms are available too.
Put another way, business architecture choices are typically about where the boundary
should be drawn and how linkages between related activities are to be structured and
exploited. Earlier parts of this chapter have been mainly concerned with the where
question: the extensiveness of the wealth maximising boundary of the firm. We now
devote a little attention to the how question.
We remarked earlier on the fact that the two cases of purchase through market exchange
and make in house could be thought of as two extreme forms at opposite ends of a
spectrum of possibilities. Lying between two polar cases of pure market exchange and
complete integration within a single organisation is a host of hybrid organisational structures.
These are hybrid in the sense that they utilise both market exchange and integrated structures.
Among the many forms that hybrids could take are long term relationships forged among otherwise
independent firms joint ventures, licensing and franchising arrangements and the adoption of
common standards.
Long term relationships: could be along the value chain (upstream or downstream) and so
are about customer-supplier relationships. Or they could be between businesses whose core
activities are in different supply chains. Transaction cost economics theory can also help to
explain when and where such intermediate forms might be optimal.
A framework developed by Douma and Schreuder (1998) is useful in helping us
understand when intermediate forms might be appropriate in vertical relationships.
Douma and Schreuder specifically use a TCE approach to analyse vertical choices among
the three possibilities of pure market exchange, long term relationships between
independent firms, and complete vertical integration. Figure 6.8, adapted from that
source, shows a matrix that spans two dimensions that we know are important in
determining the size of transaction costs. One is uncertainty, the other is asset specificity.
The permutations of the categories shown for these two dimensions yield six
configurations of asset specificity and uncertainty. For each of these configurations, the
cell contents list the structural form which Douma and Schreuder conjecture is implied by
TCE. Where asset specificity is low for both parties and there is little uncertainty, neither
party is subject to opportunistic recontracting, and the costs of market contracts are low.
Here, spot market exchange is the efficient structural form. In contrast, where one party
must invest in transaction-specific assets but the other party does not, the first is
vulnerable to hold up (opportunistic recontracting) by the second. These circumstances
favour the development of vertical integration, initiated by the party that invests in
transactions specific assets. This is likely to be true whether or not there is uncertainty
surrounding the relationship between the two parties, but is more likely where uncertainty
is high.
Now consider the cases where both parties must invest in transactions-specific assets.
The symmetry here implies that both parties are dependent on each other. If uncertainty is
32
low, this mutual dependence is likely to favour long-term contractual relationships. These
are particularly likely if transactions between the two are frequent, as repetition of the
relationship tends to generate co-operative behaviour. (See Chapter 13 for more on this
point.) However, even where asset specificity is high for both parties, vertical integration
may emerge when there is high uncertainty present. This is particularly likely if the
uncertainty relates to the quality of inputs from a supplier. Then the input supplier has
incentives to act opportunistically, and the buyer is likely to vertically integrate upstream.
Finally, consider the case where asset specificity is low for both parties but there is high
uncertainty present. If the frequency of transacting is low, the parties are likely to remain
independent, and relate through market contracts with contingent-claims clauses to deal
with the uncertainty that is present. However, if the frequency of transactions is high and
duration over time is lengthy vertical integration may be optimal for both parties,
avoiding repeated high costs of contingent contracting.
Question: Think of an example of each of these six cases. Compare your answers with
those we suggest on the web pages for this textbook.
What do these cases tell us in general? First, we note formal integration is likely when
uncertainty is high and when asset specificity is high. Second, it matters whether one or
both parties to a transaction must invest in transaction-specific assets. Third, we also note
that the frequency and duration of exchange relationships between parties can be
important conditioning factors on the architecture of organisational choices. Finally,
intermediate forms – such as the long-term contracting mentioned above- are most likely
when both parties have specific assets and complexity is low. These conditions are most
conducive to trust, with mutual high asset specificity creating conditions in which both
have much to gain by collaboration and coordination and where low complexity
generates the transparency that is needed for long term relationships.
Insert near here: Figure 6.8
Source: Figure 9.12, page 173 in Douma & Schreuder, 1988.
Figure 6.6 Governance structure, asset specificity and uncertainty
A tool much used in business strategy analysis is Michael Porter’s model of the 5 forces
of industry competition, illustrated in Figure 6.9. The value of using this tool resides in
the way it helps the strategist to study relationships between a business and the other
players it interacts with – actually or potentially – in its transactional environment. Very
often, one finds that the performance of businesses in a particular market is heavily
influenced by one of these five forces. Clearly, many of the relationships in question are
vertical in form, such as relationships with suppliers and with customers. Porter sees
these relationships in terms of relative bargaining power. Our discussions in this chapter
suggests that much of what we understand by bargaining power is really about the extent
to which one or both parties invest in transaction-specific assets, how much uncertainty
pervades the exchanges, and how frequent and durable the transactions are. Given this,
33
Figure 6.6 and its associated discussion gives valuable insights into how a business might
best structure these relationships.
Question: Look at some strategy texts, and study some of the conditions which are said
to affect relative bargaining power of suppliers (or of customers). Try and explain these
items in terms of TCE theory using the concepts of asset-specificity, uncertainty,
frequency and duration-length of transactions. Are the two forces of potential substitutes
and ease of new entry understandable in the same way?
Insert Figure 6.9 near here: Caption:
Figure 6.9 Michael Porter’s 5 Forces Model of Competition
We should also remember that boundry issues can be about disintegration as well as
integration. A topic of much current interest in this regard is outsourcing, which can be
thought of as vertical integration in reverse – vertical disintegration, if you like. This trend was
called “getting back to basics”, concentrating on the core, and more graphically, “sticking
to the knitting”. (This latter phrase comes from the best selling book by Peters and
Waterman, In Search of Excellence, which gave this as one of the rules for creating
business excellence!) We need no extra theory to explain outsourcing. Outsourcing of
some business service is likely to be desirable when the existing mode of in-house
sourcing that service is more costly than obtaining it through spot market transactions or
some hybrid form of long term contracting relationship. The widespread prevalence of
outsourcing in recent years may be explicable in terms of changes in the relative costs of
sourcing business functions internally and externally.
Question: find a recent example of an outsourcing decision. Try and explain this in terms
of TCE. In particular, try to identify what may have changed so that the costs of sourcing
the function in-house have ceased to be lower than sourcing from outside the firm.
Brief comment here on some horizontal intermediate relationships:
 Common standards
 Sharing costs and risks of technology development.
Costs and benefits of franchising:
8. Conglomeration
Conglomerate firms are those which produce a variety of goods or services, or contain a
portfolio of businesses, for which there is no 'technical' interdependence (via relatedness
of customers or relatedness of production technologies).
34
Is there any economic justification for these?
See corporate strategy notions built on ideas of parenting and/or portfolio management
later in the text.
Key point: there is relatedness here. But it is relatedness among businesses rather than
based on linkages arising on the supply-side (technical scope economies) or the demand
side (via customer preferences).
Conclusions
Why are there complex firms?
Complex firms may be able to add value over and above what would be produced by the
constituent simple firms operating independently. This chapter has given several
explanations of how that value may be added. However, it is not necessarily true that
every complex firm does add value in this way.
What is the efficient boundary of the firm?
The efficient boundary of the complex firm is where increasing its complexity no longer
promises to add value. Thus a firm involved in both advertising and public relations
might well make sense but extending its activities further, into say financial consulting,
might be harmful to value creation. Examples of firms which developed beyond their
efficient boundaries are not hard to find.
Do diversified firms make sense?
If we define making sense as adding value over and above that of the individual simple
firms then the answer is that those diversified firms that add value make sense and those
that do not add value do not make any sense. However, it is conceivable that non-valueadding diversified firms exist and persist despite the apparent contradiction of economic
efficiency logic.
Why are some firms diversified while others are not?
Diversification does make sense in economic terms but not for all firms, or for the same
firm at all times. It depends specifically on the availability of the capabilities developed
by the firm and the potential for transferring or sharing these capabilities to exploit new
opportunities for adding value. That is, it depends on the existence of economies of
scope. It depends also on the fact that because of transactional problems it is often
difficult for firms to capture the rents or profits of any under-utilised skills and other
organisational assets or capabilities by selling them to someone else. The only way to
capitalise fully on available resources and capabilities is then to diversify. However not
all firms with strong skills and other assets built up in their current activities will
necessarily be able to identify new opportunities for exploiting these in a value adding
way. Some capabilities will be easier to exploit than others. Some firms may not even be
35
aware of the potential value of their resources and capabilities and so fail to consider
utilising them more fully.
Why do firms in some industries diversify more successfully than firms in other industries?
Probably because the nature of the skills, activities, capabilities, and competencies built
up in some firms is easier to transfer and/or share than in other firms. In some firms, say a
steel processor, the skills and capabilities developed may be substantial but too specific to
steel processing to be transferred or shared. In other firms, say a chemicals processor,
there may be better opportunities for exploiting capabilities more widely.
Should firms diversify?
Yes, firms should diversify but only for the right reasons, which means adding value. The
history of the diversification phenomenon suggests that diversification was not always
pursued for the right reasons. There is a likelihood that some of the trend towards
increasingly diversified firms was motivated by reasons other than adding value. For
example, there was probably an element of fashion in the rise of diversification in the
1960s because of the attention given to certain business leaders who promoted aggressive
diversification, consultants, and business school gurus who advocated diversification for
all sorts of reasons. Some of these people, the academic Igor Ansoff for example, placed
great emphasis on synergy, and were careful to advocate diversification for the right
reasons, but others failed to make clear precisely what they thought the ultimate purpose
of productive enterprise was. If the enterprise had no clearly defined purpose then of
course diversification was as good a means of achieving this purpose as anything else.
Other people took on board Ansoff’s ideas about synergy and proceeded to oversell them.
Firms need to be careful therefore about the attractions of diversification and clear about
their ultimate purpose in pursuing it.
Point forward here to following chapters.
Learning outcomes
Further reading
Implications of the ideas of costs for accounting systems - particularly in activity-based
costing and modern managerial accounting - are examined by Shank and Govindarajan
(1989).
Douma and Schreuder (1998)
Kay, J. Foundations of Corporate Success, Oxford University Press, 1992, especially
chapters 3 and 4.
36
Miller, G. J. Managerial Dilemmas, Cambridge University Press, 1994, not an easy
read, but a brilliant analysis of the economics of organisations and organisational
objectives.
Porter, M. (1987) “From competitive advantage to corporate strategy”. Harvard Business
Review May-June 1987.
Pratten (1987).
Rappaport, A. (1992): CEOs and Strategists: Forging a Common Framework, Harvard
Business Review, May/June 1992, pages 84-91.
Ricketts, M. The Economics of Business Enterprise, (2nd Ed), Harvester-Wheatsheaf,
1994, especially chapters 2 and 3.
Good general discussions on the issues we have covered in this chapter can be found in
Kay (1992), Grant (1991), Besanko et al (1996), Sadtler, Campbell, and Koch (1998)
and Bishop and Kay (1993). Porter’s contributions are found in Porter (1985) and
(1987). Issues related to diversification are discussed in Peters and Waterman (1982).
Geroski and Vlassapoulos (1990) assess the European mergers experience. Finance texts
such as Brealey and Myers (1991) often give a good analysis of diversification,
comparing conglomerate with portfolio diversification. The papers by Love and Scouller
and Peteraf, discussed in this chapter, repay reading in full.
Web links
Discussion questions
1. What are transactional difficulties and costs and why do we need to be concerned with
them?
2. What factors give rise to transactional difficulties and costs?
3. According to the transactional view of the world what exactly are firms for?
4. According to the transactional view of the world what exactly is the function of
entrepreneurship?
5. Evaluate the view that successful high street retailers such as Marks and Spencer are
successful because they were lucky enough to have acquired the best high street locations
before their value was really appreciated.
6. Select an organisation of which you have some knowledge.
37
(a) What, if any, are the major sources in your organisation of economies of scope? Are
there other economies of scope you think that could be exploited by your organisation?
(b) Explain how the boundary of your organisation (horizontal and vertical integration,
diversification, relationships with suppliers etc.) has been affected, or might be affected
in the future, by the relative costs of doing business within the firm and through market
exchange
7. Consider the long run average cost function of a business school institution
specialising entirely in the provision of taught postgraduate degree courses in business
economics. What kinds of scope economies might be available if this institution
broadened the range of courses it provided?
8. How might a complex firm be (a) less or (b) more valuable than the sum of its
constituent parts operating independently?
9. Why do some firms diversify and then go back to basics?
10. ”The existence of profitable diversified firms proves that diversification is a valuable
business strategy.” Discuss.
11. Outline and discuss Porter’s four concepts of diversification strategy.
Problems
Case Study: Information costs and the information economy
 Economics of the internet and E-Business
38
Box 6.2 Brokers and the UK insurance industry
It has been common for individuals seeking insurance (especially car insurance) to use the services
of brokers. The specialist knowledge acquired by brokers, and economies of scale in search
activities, meant that the costs of searching for the lowest premiums available were lower than
individuals would face when searching for themselves. Overall, organised brokerage activity
offered a cheaper means of arranging insurance than direct market exchange.
But this situation has been changing in recent years. Several factors have contributed to the change.
First, increased competition in the telecommunications market has lowered the real price of
telephone services. Second, developments in electronic information storage and processing have
allowed insurance companies to offer prices closely linked to the riskiness of individual customers,
and to quote prices virtually instantaneously. Insurance companies and individual customers both
have much less to gain from the intermediation of a broker. So here is a case where one kind of firm
- the insurance broker - is finding it increasingly difficult to survive, as the relative cost of market
exchange has fallen.
End of Box 6.2
39