Integration, complexity and environmental uncertainty

After reviewing the literature on corporate strategy and structure, Caves (1980) remarks that much of it seems implicitly to be concerned with the concept of an ‘organisational production function’. The ‘inputs’ into this function, he argues, are resources devoted to collecting and analysing infor­mation and coordinating other factors of production. The ‘output’ is the ability to ‘reallocate …in response to unexpected disturbances’ (p. 89). The mode of expression betrays Cave’s neoclassical standpoint, and in Chapter 3 it was seen that many would dispute the ability of the neoclassical para­digm to cope with the truly ‘unexpected’ as distinct from the statistically ‘risky’. Nevertheless, the inputs into his ‘organisational production func­tion’ are Casson’s entrepreneurs. Casson is quite clear that the entrepreneur can be an employee (pp. 213—15) and that the institutional setting of the firm will be one of the factors underlying the ‘supply’ and ‘demand’ curves of Figure 3.1.

1. INTEGRATION, COORDINATION AND COMPLEXITY

Notwithstanding Williamson’s view reported in Chapter 2 that the distinc­tion between complexity and uncertainty is inessential since both give rise to the problem of bounded rationality, we will consider separately the influence of each on the firm. Very close and complex technical relation­ships between processes and products may imply high costs of transacting in the market because of bounded rationality, but it is not clear that coor­dinating these processes necessarily requires great entrepreneurial as dis­tinct from technical judgement. It may be perfectly obvious what actions need to be taken in any given set of circumstances, even if the exhaustive enumeration of all the possibilities in a state-contingent market contract is not feasible.

This is the situation which underlies some of the ‘conventional’ explana­tions of integration within the firm. Within the neoclassical tradition of the ‘firm as production function’ mentioned at the beginning of Chapter 6, it is natural to look for explanations of horizontal integration in the pursuit of economies of scale; or of conglomerate integration in technical com­plementarities or economies of scope; or of vertical integration in the close technical connections between one stage of production and the next. The objections of the transactions cost school to this way of thinking are most succinctly summed up in Williamson’s (1975) aphorism: ‘technology is no bar to contracting’ (p. 17). Let us interpret this point in more detail and con­sider its applications to horizontal and vertical integration.

1.1. Horizontal integration

This is the most familiar case considered in elementary textbooks. By inte­grating within a single firm the resources required to produce larger quan­tities of a single output, costs may decline if there are economies of scale. A falling average cost of production with higher output is ultimately trace­able to ‘indivisibilities’ associated with the various inputs. A particular item of capital equipment; for example, a ship for transporting cargo in bulk; will have associated with it a level of utilisation which minimises costs per tonne mile. If the ship is not fully laden, costs obviously will rise. On the other hand, tonne miles can be increased by reducing the time spent at ports or in maintenance. Such continuous operations may increase staff costs as crews, maintenance engineers, and administrative personnel are aug­mented. This example is therefore just a special case of the economist’s familiar idea of increasing and then ultimately diminishing returns to a fixed factor.

Now suppose that there are five firms each shipping a particular type of cargo (let us say grain) between two countries. They each purchase the grain from suppliers, store it in warehouses, transport it in their own ships across the ocean, and distribute it to wholesalers at the other end. It is clear, however, that the actual process of shipping is unnecessarily costly with five crews and five small ships. A single large ship would enable advantages of scale to be achieved.⁹ Does this imply that the five firms should merge their operations? ‘Common sense’ suggests that this would not be unreasonable, but, in principle, there are other options open. The most obvious alterna­tive is that a new specialist ship-owner enters the market with a giant grain transporter, and the five grain buyers and distributors contract with the ship-owner for cargo space on the larger ship. In a world of zero transac­tions costs, in other words, indivisibilities might just as reasonably lead to ‘disintegration’ and the entry of a new specialist firm as ‘integration’ and the creation of a single firm out of the five existing ones. The purely tech­nical advantages associated with particular bits of equipment do not auto­matically require reduced numbers of firms for their realisation. They may lead to integration, but the hidden and unspoken assumption in the ‘common-sense’ view is the (very often correct) one that transactions costs will inhibit the ‘market-like’ solution.

All ‘economies of scale’ involve an indivisibility of some form. Two further examples, taken from Hay and Morris (1979, pp. 44-5) may be useful. Larger size, it has been argued, will lead to economies in the use of mainte­nance staff. If a breakdown in any particular item of equipment is a random variable with a certain probability distribution, the larger the number of items of such equipment the less is the variation in the average number of breakdowns per period.¹⁰ It then turns out that a given standard of service (in terms of the probability that a breakdown will occur with no staff avail­able to cope with it) can be achieved at lower cost per machine as the number of machines increases. Although technically flawless, we can not deduce from this that larger ‘firm’ size is encouraged. Many small firms could, in principle, use an independent maintenance firm to cope with breakdowns, just as coffee machines, domestic appliances and office equipment are fre­quently serviced and mended by outside contractors. This could either be done on the basis of a payment per visit; or each small firm could pay the maintenance company a retainer, which would effectively represent an insur­ance premium, and the company would pool the risks. Either way, economies in maintenance do not logically imply the amalgamation of small firms, though the transactions costs associated with contracting in the market may, of course, produce such an effect.¹¹ Similar reasoning can be applied to the case of economies of purchasing inputs. If there is a fixed or overhead element associated with the cost of effecting each order, independent of the size of the order, it is clear that costs per unit will decline as the average size of purchase increases. This seems to favour large firms, but it is important to identify clearly the source of any advantage. The overhead element inde­pendent of the size of the order is nothing other than a simple representa­tion of a ‘cost per transaction in the market’. If there are transactions costs associated with market exchange, reducing the number of such transactions will be encouraged. The economies thus identified are transactions cost economies. If, for example, this transaction cost were peculiar to the bulk order of supplies of a particular input and did not apply to the transactions between the purchasing firms, a set of small firms could band together and order in bulk, rather as a group of students might combine to order a mag­azine or newspaper. ‘Integration’ would then occur only at the single specific level of transactions with this particular supplier and not more generally.

1.2. Vertical integration and technical links

Technological considerations have traditionally been seen as an important force leading to vertical integration in some industries. The case of the iron and steel industry is perhaps the most often cited. Thus, ‘the later stages of production are bound to the earlier by many specially close technological links. In particular a great waste of heat can be avoided’ (Robertson and

Dennison, 1960, p.27). From our discussion of horizontal integration the objection to the view that vertical integration is a purely technological phe­nomenon will be clear. The question is simply whether the close techno­logical links should be coordinated within the firm or by market transactions. Contracts might be drawn up which would permit the process of steel making to proceed with the steel maker purchasing iron in molten form. So detailed is the coordination required, however, that this is clearly not going to be the favoured procedure, and the entire process is more con­veniently brought under a single administration in the firm. In cases like the production of iron and steel, the advantages of vertical integration are so obvious that it appears somewhat pedantic to drag transactions costs into the discussion. Yet, logically, nothing can be deduced about the organisa­tion of production without reference to transactions costs because the organisation of production concerns precisely the choice of contractual arrangements best suited to a particular set of technical circumstances.

2. Integration, Coordination and Uncertainty

2.1. Generating information within the firm

In a continually changing environment, entrepreneurs gather information, interpret it to discover the opportunities latent in it, and act upon it. Acting on the basis of entrepreneurial judgement requires command over resources. It does so because the market value of an entrepreneur’s knowl­edge cannot itself be appropriated simply by trading it in the market. A buyer would require the information before an assessment of its value was possible, but once in possession of the information there would be no need to purchase it.¹² The firm responds to this problem by screening for entre­preneurial talent, placing entrepreneurs in circumstances likely to produce a flow of information suitable for the exercise of entrepreneurial judgement, providing them with resources to back their judgement, and instituting a reward system which enables those with a flair for these decisions to benefit from them. On this interpretation, the firm represents an internal capital market, a method of allocating scarce resources amongst competing uses on the basis of entrepreneurial judgement.¹³ The job is done internally because of the problems associated with transactions in markets for information.

From a property rights perspective, however, this rather rosy view of the ‘internal capital market’ requires some qualifying. Manager-entrepreneurs will be much better able to pursue their ideas and respond to unfolding events if they have the residual control rights in the assets they are using. In a multi-divisional firm, however, this will not be the case. It is the managers at headquarters who can implement a new idea without local approval rather than the other way around. As Bolton and Scharfstein (1998, pp. 107-8) put it ‘the same control that gives HQ the incentive to monitor has an adverse effect on the incentives of division managers to act in an innovative or entrepreneurial fashion’. As was seen in Chapter 4, concen­trating control rights in the hands of a single manager will reduce the incen­tive to others to make specific investments and act entrepreneurially. Independent owner-managers financed by bank lending would seem to give greater incentives to entrepreneurial alertness if the transactional problem between entrepreneur and capitalist could somehow be mitigated.¹⁴

2.2. Vertical integration and information

Integration of activities to take advantage of internally generated infor­mation would still seem to imply a fairly specialised firm. Vertical integra­tion may be seen as a way of gaining from the improved communication of information from upstream to downstream firms. Arrow (1975) considers better information about the supply (and hence the future price) of an upstream good, leading to more appropriate input decisions by a down­stream firm, as a motive for vertical integration. Integration into closely related products or processes may be favoured because information con­cerning markets or production is not totally specific, and may have wide- ranging implications.

2.3. The chain store

A useful example of integration, both as a means of discovering and analysing information, and as a means of using this, is the chain store. Many chain stores have benefited from the development of knowledge con­cerning customer wants and the identification of gaps in the market. This has not necessarily led to backward vertical integration into manufactur­ing, since the links between market information and production informa­tion may not be very important. The store concentrates on the entrepreneurial functions of identifying demands and then contracts with manufacturers directly to arrange supply. This, of course, implies that information about what is wanted can be cheaply and clearly communi­cated to potential suppliers. If the request is too innovative, it may be difficult to arrange supply from independent contractors, and backward vertical integration may occur. In section 6, we will consider the relation­ship between integration and innovation in more detail.

Information generated within the firm about markets can therefore be used to contract with suppliers, and the benefits derivable from this entre­preneurial information can then be used in every branch of the chain. A similar point could be made about information concerning the actual operation of the stores, the handling of goods, layout, warehousing, pricing, and so forth. New information can be used quickly throughout the chain of stores. As Edwards and Townsend recognised in their explanation of the growth of the chain store ‘the main advantage lies in the fact that there are large indivisibilities in knowledge’ (p.296). As will be seen in section 5, the multinational enterprise can also be seen as a means of exploiting the use of knowledge throughout a geographically dispersed firm. In this section, however, we have been emphasising the ability of a firm to adjust and to capitalise flexibly on a flow of new information. Section 5 considers more generally the ability of the firm to exploit its non­tradable resources by means of internal expansion.

2.4. The conglomerate, information and uncertainty

Conglomerate firms are much more difficult to explain on the basis of the use of internally generated information. Instead, the emphasis switches to diversification as a means of coping with unexpected events. A neoclassical approach would see the firm as holding a portfolio of risky prospects so designed to maximise the value of expected wealth (in the case of risk neu­trality) or, more generally, expected utility (in the case of risk aversion or risk loving). This ‘portfolio’ could involve the firm undertaking projects in many different product areas. A serious difficulty, however, is to understand why risks should be pooled within the firm instead of in the markets for financial claims. In a neoclassical world with low transactions costs, risks could be pooled as effectively by people holding claims to the returns from a variety of independent but specialised firms operating in different areas, as claims to the returns from a single suitably diversified firm. Thus, indi­vidual people through their portfolio of financial assets could select a desired risk-return profile, and this could be done independently of any decisions concerning the appropriate administrative framework for coordi­nating the physical assets concerned. If the firm itself takes on a risk­pooling role, it is once more to the transactional problems of the alternative institutional arrangements that we must turn for an explanation.

One obvious candidate for consideration is that people lack the infor­mation to take complex portfolio-building decisions, and that the transac­tions costs involved in holding a large number of different financial claims would prevent risk pooling. This problem, however, would appear to be met by developments within the financial markets themselves; developments such as the growth of unit and investment trusts which give people access to specialist information and, through the use of an intermediary, once more reduce transactions costs. A more persuasive explanation for diversification within the firm concerns the use of the physical assets, human capital and firm specific ‘know-how’. As was seen in some detail in Chapter 6, within the firm, information about members of the team is col­lected, inputs are monitored and complex hierarchical incentive structures are instituted. Many of the skills absorbed are firm specific rather than product specific and relate to the ability to communicate and to get on with colleagues. Further, incentives require that the firm is perceived as having effectively an indefinite ‘life’ and is unlikely to go bankrupt. At the same time, they imply that, for a portion of a person’s career, remuneration will exceed anything available on the ‘open market’. In these circumstances, the ability to relocate resources from stagnant to growing sectors is essential. Bankruptcy or stagnation and decline in a specialist firm are of no conse­quence in a neoclassical world of full information where laid-off resources are immediately reabsorbed elsewhere at the going market rate. However, in the context of the analysis of Chapters 5 and 6, it is clear that the con­sequences of decline for the operation of the internal labour market or the tournament could be very severe and imply a renunciation of the firm’s ‘implicit’ obligations.

Rapidly changing and uncertain conditions will therefore lead to con­glomerate diversification and an M-form corporate structure. The head office monitors each division’s performance and, through the internal capital and labour markets, reallocates resources towards growing and away from declining areas. This strategy is not, therefore, simply a matter of pooling risks. If it works, it will do so because the firm’s structure permits flexibility and encourages adaptation to conditions more correctly viewed in Knight’s terms as ‘uncertain’.¹⁵ Kay (1983) in his explanation for the con­glomerate emphasises this point. The firm will make use of information flows from ‘richly linked’ product markets, but this implies that ‘where links exist they may create mutual or common vulnerability’ (p. 96). In very static environments, this may not be of great concern¹⁶ but in other areas the danger of ‘technological mugging’ may lead to conglomerate diversification. For Kay, neoclassical portfolio theory is not the appropriate tool for under­standing this phenomenon: ‘The decision-maker faces a truly uncertain sit­uation in which there is an asymmetric emphasis on the possibility of life cycle decline rather than life cycle growth … All we assume the decision­maker is able to do is to order the environment in terms of surprise poten­tial’ (p.98).¹⁷ Environments in which surprise potential is great will lead firms to diversify into areas which are not strongly linked, in order to reduce the damage that might result from a sudden mugging.

Source: Ricketts Martin (2002), The Economics of Business Enterprise: An Introduction to Economic Organisation and the Theory of the Firm, Edward Elgar Pub; 3rd edition.