Background Briefing Notes 13 Autonomous Business Units, Responsibility Centres and Internal Markets
The use of “internal markets” is one of a number of new developments in the way organisations are structured and run. Other associated development include the emphasis on “lean organisations”, the “quality movement”, and “outsourcing” and the development of the “virtual corporation”. These developments, and the problems of managing the changes these involve, are reviewed later in the module.
“Internal markets” are systems used inside organisations which wish to combine the advantages of operating through an organisation with the advantages of operating through markets. They are based on the idea that central management cannot mastermind and plan (“micromanage”) everything that happens inside the organisation.
Essentially the organisation is divided into a number of “responsibility centres”, (or “autonomous business units”) covering, for example, a particular product that the organisation makes, or a particular activity which services other centres (such as transport, or computing). These centres are given a high degree of autonomy over the way that they organise their own affairs, and their performance is typically judged in terms of the profits they make, or their success in meeting a budget.
As far as possible, they are charged the true costs of their activities and (in the case of “profit centres”) credited with the value of their output. This is why an “internal market” is necessary: to put a value on goods and services that move from centre to centre. If one centre absorbs services from another, then the first needs to be charged the real costs of those services, while the latter needs to be credited with their value. If no charge is made, then it is likely that centres will over-use “free” resources from elsewhere in the organisation.
The development of responsibility centres – the multidivisional firm
The great majority of large firms are now “divisionalised” in some way. Typically, a product, or group of products, is the responsibility of a single division, with various key functions (such as production, purchasing, and so on) decentralised to the division. This can mean that a single division is responsible for all stages of production and marketing a product, rather than this responsibility being shared across a number of departments.
Sometimes there may still be functionally specialised departments which service a number of divisions, in which case the divisions will “buy in” these services from the service departments on a commercial basis. Or the divisionalisation may be by geographical area, or type of customer (industrial/retail etc). The essential point is the degree of delegation this involves, and the operating autonomy the division is granted – the divisions are sometimes termed “autonomous business units” (ABUs).
One of the earliest companies to reorganise this way was Dupont Chemicals of the USA. It produced (and produces) a wide range of chemical-based products, from dyestuffs to explosives. Chemical engineering is one in which there are very large technical economies of scale (which is why chemical firms tend to be very big) but early in this century Dupont found it was being out-performed by smaller, single product firms. Dupont concluded that this was because in the smaller firms there were individuals who were responsible for ensuring that individual products were successful, whereas in Dupont this responsibility was spread out over a number of functional departments – so ultimately no-one was responsible, below the office of CEO.
The Dupont solution was to regroup its activities into product divisions, each with its own CEO responsible for ensuring that the product(s) of the division made a profit. This new organisational structure was later adopted by General Motors (see distributed handout “Does Organisation Matter?”), and was the basis of its overtaking Ford as the leading automobile manufacturer. Ford retained a highly centralised system until after the death of Henry Ford I, after which it reorganised on GM lines. This new type of structure (variously described as “divisionalised”, “multidivisional”, “federally decentralised” and “M-Form”) was so widely imitated that by the 1970s every single firm in the US Fortune 500 claimed to be divisionalised.
One important feature of this system, as pioneered by Dupont, and developed by Alfred Sloan and William Durant at General Motors, was the separation of “operating” from “strategic” matters. The former are the responsibility of the divisions, with a minimum of interference by Head Office, which instead focuses on strategic matters, assessing the performance of the divisions, and allocating investment funds (an illustration of the advantages of delegation). This change in the way organisations are structured went hand in hand with the development of quantitative Management Science1 techniques both to aid decision making and to provide numerical data to allow the objective monitoring of performance.
Even where organisations do not formally regard themselves as divisionalised, they often make use of ABUs, responsibility centres, and of internal markets. Of particular note in recent years has been the development in the UK of the use of internal markets in organisations such as the National Health Service and the BBC. A similar development has been the reorganisation of primary and secondary school education, with individual schools given a high degree of control over their budgets, and allowed to purchase from a range of suppliers, rather than being constrained to the departments of their Local Education Authority. A number of other public service activities have also been put in the hands of semi-autonomous agencies, who are the subject of contracts with government as to the level of service they will provide, and the criteria against which their performance will be measured.
Types of Responsibility Centre
The most common types of responsibility centre are:
Cost Centres: Management is held responsible for the costs of a centre. Typically a budget is set, and the centre must deliver “on budget”. Perhaps the most common type of responsibility centre.
Revenue Centres: here less emphasis is placed on costs, perhaps because most if not all of these are outside the control of local management. More emphasis is placed on revenues, and the centre is typically judged on its sales performance. Examples include many retail stores.
Profit Centres: both costs and revenues are taken into account. This type of centre requires a high degree of autonomous control over both costs and revenue for the measure to be significant. Examples include product divisions of large manufacturing firms.
Investment Centres: similar to profit centres, but the cost of capital is taken into account in assessing costs. The centre is assessed in terms of the return that it earns on capital employed. Examples include subsidiaries of large holding companies.
Which type to use? And what sort of performance measures?
Ideally, the decision as to which type of responsibility centre to adopt should be determined by the purpose of the centre. A wrong decision can corrupt the way the centre operates. For example, many companies initially adopted call centres as a response to pressure from customers for better service, and more direct access to “real people” inside the organisation who could handle their enquiries. This implied large numbers of lines leading to staff well trained in both answering customer queries, and the products of the business.
But in many organisations, as performance was measured purely in terms of the cost of running the call centres, staff and training were cut back, customer telephone queuing time increased, and customer satisfaction declined. Redefinition of performance in terms of customer satisfaction led to the development of performance measures in terms of, for example, number of repeat calls on the same problem, and new business generated through call centre visitors – more in line with the reason for establishing the centres in the first place.
Remember the discussion of the dysfunctions of bureaucracy, and the danger that departmentation would lead to a focus on narrow “local” goals. In many ways the adoption of autonomous business units represents an acceptance of this, and an attempt to turn it positively to the advantage of the organisation, by allowing local business managers to concentrate on a limited number of local targets. But if the targets themselves are badly chosen, the results are likely to be damaging for the organisation as a whole.
What’s The Point Of Responsibility Centres?
Sometimes top management regards responsibility centres as ways of gauging the true costs of activities, or of assessing which parts of the firm are doing best. Both of these approaches miss the point, and may be actually wrong.
These systems are primarily aids to decision making and resource allocation. They enable decisions to be taken “close to the action”, which makes for speed, and also allows people with perhaps a better knowledge of local conditions to make decisions. Assuming that the goals of the centres are well aligned with those of the organisation, local decisions should also coincide with the interests of the organisations – but local decision makers will not commit resources unless this is justified.
Responsibility centres and internal markets are also ways of giving real meaning to concepts such as delegation and “organismic” structures. Managers of the various centres have real autonomy, and can make real decisions. In addition, flexibility and adaptability are encouraged – with less emphasis placed on adherence to centrally determined procedures, and more on results. These systems are based on the concept of control of outcome, rather than control of process. It should also be recalled that possibly the major advantage of delegation is the opportunity it offers for the delegator to focus on what only they can do. In this type of system the “headquarters” of the organisation is freed from concern with operating matters, and can concentrate on the strategic.
There may also be organisational benefits in terms of motivation and management education. In terms of “needs” theories, responsibility centres offer opportunities for more responsible, autonomous and challenging jobs; in terms of process theories, employees can develop an expectancy that performance is more directly related to what they bring to the job. Managers also get experience at a comparatively early stage of their career, of making “real” decisions, and integrating a variety of functional activities to achieve the goals of the organisation.
It has also been argued that these systems can help to promote the notion (popularised by Peter Senge – see article 24 in Pugh (ed)) of the “learning organisation”. New ideas can be developed and pursued in autonomous parts of the organisation, or “spun off” into new units (see the 3M’s case in Morgan, reading number 49), and innovative ideas disseminated throughout the rest of the organisation. These systems can be very flexible – new sections can be added to the organisation, or old ones closed down, or even sections “floated off” as separate entities. The organisation can switch from internal to external suppliers. All this can be done within the existing overall organisational structure, without causing widespread disruption.
Chandler, who was among the first to trace and analyse the development and spread of the multidivisional form of organisation, relates this flexibility to the strategy of the firm. Multidivisionals found it easier to produce and market a diversified range of goods and services, and to adapt this portfolio of products to changing environmental conditions.
There is also the straightforward opportunity to save on the organisational costs of co-ordinating activity. All co-ordinated human activities involve this sort of cost. In markets there are costs of seeking out buyers and sellers, setting contracts, and claiming and making payment; in organisations there are costs of planning, issuing instructions, and monitoring results. Williamson’s analysis of markets and firms (he refers to “markets and hierarchies”) revolves around these “transactions costs”: in simple terms whether an activity takes place through a market or a “hierarchy” depends on where the transactions costs are least. (Of course there is an extreme case in which the costs of co-ordination through any mechanism are so great that they outweigh the benefits of co-ordination – this is the point at which you say “The hell with it, I’ll do it myself!”)
One reason an organisation may use an internal market is because senior management has reasoned that the costs of organising things this way are cheaper than the costs of relying on traditional managerial hierarchies.
In the particular case of multinational corporations, these systems allow local operations to be adapted to local conditions – for example with regard to tax or legal requirements.
More generally, they allow organisations to structure different parts of the organisation quite differently if this seems appropriate (see Lawrence & Lorsch in Pugh (ed)).
Internal markets change the relationships between different parts of the organisation. Although they are sometimes accused of undermining co-operation, in other cases the move to a “supplier-customer” relationship can be beneficial. For example, internal service suppliers can no longer take their customers for granted, and have to be more responsive to their requirements; on the other hand the internal customers become more aware that they cannot place “impossible” demands on their internal suppliers – if they do, they will end up paying “impossible” prices. So they become more reasonable in their requests.
Although we have said that it is a mistake to see these systems only as ways of apportioning costs and appraising the performance of different parts of the organisation, a properly established system (which means realistic transfer prices: see below) should help senior management do this. However care has to be taken in interpreting the results for individual units or centres (see p6 of these notes).
For internal markets to function, a value – a price – has to be put on goods and services as they pass from one responsibility centre to another. (Whether or not “real money” has to change hands is another matter – an alternative is for notional budget balances to be adjusted by head office as transactions take place. This may be preferable to allowing budget holders to operate their own real bank balances, etc, not least because otherwise a large amount of real cash may be tied up inside the organisation simply to finance internal “trade”2.)
These transfer prices are necessary to provide a revenue for the “selling” division, and a cost to be debited against the “purchaser”.
Transfer prices have a number of functions. Much of the literature is concerned with the use (and abuse) of transfer prices in multinational corporations. Setting a high (or low) transfer price on movements of goods or services between different parts of the same firm in different countries can be a way of pushing up reported profits in one country (presumably a low tax regime), at the expense of profits reported in another (a high tax regime), or of minimising tariffs charged on the value of imports, or even circumventing exchange controls. (A recent alleged example concerns the price of “intellectual property rights” – royalties and so on – charged in the USA by non-US record companies. It was alleged – in the USA – that American subsidiaries of these companies were being charged exorbitant royalties by their parent companies as a way of syphoning money out of the USA.). Although this is a fascinating topic in its own right, it is outside our compass here. However, we should note that in reality this may be the major factor determining transfer prices, rather than the points which follow.
In some organisations, transfer prices are principally a way of keeping track of the cost of work in progress, or simply of allocating costs between responsibility centres.
The most powerful use of transfer prices is however as an aid to decision making. Given the freedom to decide, a centre will only agree to supply goods and services if the transfer price received justifies the costs incurred. Similarly, a centre will only absorb resources from elsewhere in the organisation if these can be justified in terms of value added. Important decisions about the volume and composition of production, and the production methods used, are thereby effectively decentralised to the operating units.
Use of transfer prices in this way raises a number of questions, which have to be answered (by senior management) if the responsibility centre system is to operate effectively:
how are the transfer prices to be set?
and (perhaps even more controversially)
should responsibility centres have the “freedom to trade” outside the organisation?
i) how are the transfer prices to be set?
Each unit could set its own price: but there is a danger that if they are an “internal monopolist” they can exploit their captive internal customers by, for example, inflating the price to transfer profits from their customers to themselves, or to pass on the costs of internal inefficiencies to customers. (Incidentally, it can also be demonstrated that this leads to an inefficient use of resources inside the firm, but this is rather beyond our scope here.) Similar distortions can result if a monopsonistic (sole buyer) internal customer is able to dictate the price to captive internal suppliers.
The firm could allow the two centres to haggle: but this is likely to mean that the final price – and therefore the internal use of resources, and the reported performance of the centres – is likely to reflect the haggling skills of unit management, rather than efficient decision making and operations. In addition, some organisations have abandoned transfer price systems because too much managerial time was being absorbed in haggling and bargaining over internal contracts, with little or no benefit to the organisation as a whole. This was one of the reasons the UK Government implemented changes to the internal market of the British National Health Service in 1997 – it was believed that hundreds, if not thousands, of millions of pounds was being wasted in unnecessary and ultimately unproductive transactions costs in agreeing, costing, and settling internal contracts.
We are reminded of Williamson’s basic point – that there are advantages to both “hierarchic” (organisational/ managerial) and “market” ways of co-ordinating activity.
An alternative is to use a cost-based transfer price – either marginal, average, or full cost. There are technical advantages and disadvantages to each of these, but whichever is used it should be noted that the selling centre can no longer really be regarded as a profit centre – it is better to designate it as a cost centre, and measure its performance in terms of its ability to keep costs down (produce within budget). Whatever type of cost-based transfer price is used, a decision has to be made on whether to use standard costs – what the output should have cost – or actual costs – what it really did. The advantage of the former is that it prevents inefficient centres passing on their costs to others; of the latter that buying centres know the real costs to the organisation as a whole, and can therefore take steps to economise.
Another alternative is to use the market price for the transferred goods or service. Sometimes there is no outside market for the particular goods or services being transferred – here senior management may impose a “calculated market price” – what the price would be if there was an external market. Where there is a real external market, firms can use the real external price to determine the transfer price. This raises a second fundamental question, on which senior management needs to determine policy:
ii) should responsibility centres have the “freedom to trade” outside the organisation?
In other words, does an internal supplier have the right to sell to outside customers, and an internal buyer have the right to source from outside – in particular do they have the right to do this even in preference to an internal customer or supplier? (Some firms practise restricted freedom to trade – a centre can trade externally, subject to higher level approval, or if there is no suitable internal customer or supplier. But this begs the question – what is a “suitable” internal customer or supplier? If senior management decides, then the management of the affected responsibility centre can argue that their decisions, and results, have been distorted by this decision, and that they cannot therefore be held wholly responsible for them. And we are back to a situation where senior management has decided that “it knows best”.)
As an example3, within IBM, divisional managers have been given freedom to buy their own office computers from whichever source they want. The logic here of central management is that if IBM’s own managers do not want to buy IBM machines, this is a problem for the divisions producing the machines, and should not be one for the internal customers. If the latter were forced to buy from the internal suppliers, it would “cover up” the problem. Freedom to trade exposes the problem, and also pinpoints where inside the organisation it lies.
However, it may be reasonably argued that there may be strategic reasons why senior management requires divisional management to source internally (for example, an automobile company may wish its managers to promote the company by driving its own cars) or to restrict external sales (perhaps, for example, to protect a particular competitive advantage in a process or component). Under freedom to trade, these objectives can be achieved by the headquarters paying the divisions a subsidy sufficient to induce the latter to buy or sell internally. One useful effect of this is to concentrate the minds of senior management – funds available to headquarters for such subsidies are limited, and senior management will therefore have to think carefully: are such subsidies justified in terms of the strategic benefits to the organisation?
(Incidentally, this reinforces a general point: in responsibility centre systems ideally all parts of the organisation should be responsibility centres, and held accountable for their decisions through the system. This includes “headquarters” sections. Halal, for example, has argued that the central headquarters needs to be able to demonstrate that it adds value to the rest of the organisation – otherwise why should not the various operating sections float off as independent organisations?
This would seem to be in the interests of the management and other members of those divisions, and of the shareholders, who would not have their profits diluted by unproductive central headquarters management. The only people in whose interest this would not be would be central management! (Cynics would say this is why they are sometimes less than keen to have their own sections organised this way.) Examples of organisations where the office of the chief executive is itself organised as a profit centre are given in the Halal article, distributed separately.)
When divisional results are reported, these can be compared with budgets and/or with those for other divisions. However, care has to be taken in how these are interpreted. It does not follow, for example, that a division that has apparently “underperformed” has in fact performed worse than others. It may be, for example, that a division reporting low profits has in fact done very well compared with another that has reported apparently better results. The former may be operating in very tough market conditions compared with the latter.
This reminds us of one of the advantages that organisations structured on ABU/responsibility centre lines have compared with a collection of independent firms providing the same range of activities. In the former there is a central management that has much more detailed access to operational information (even if they do not use this to “interfere”) than is available to the shareholders of the independent firms, and assessments of performance can therefore be much better informed.
This is particularly relevant for two issues: investment decisions and closure/divestment decisions
A “multidivisional” organisation can operate an “internal capital market”. It is important to remember that although an organisation may be structured as a system of profit centres, the profits earned by the divisions do not belong to them, but to the organisation as a whole. How these accumulated profits are used should therefore be a decision for the organisation as a whole. New investment programmes may be initiated by senior management: alternatively divisional management may bid for investment funds. However, these should be allocated not on the basis of current profit performance, but on the prospects of future returns. This may result, for example, in some divisions being regarded as “cash cows” to finance investment elsewhere.
It is argued (for example by Williamson) that the internal capital market of the multidivisional firm is in fact better placed to judge how investment funds should be used than are shareholders who hold shares in a number of independent firms, as they have much more direct access to relevant information. (This may be one of the major advantages that a multi-product, multidivisional firm has over a number of independent firms producing the same range of goods and services.)
Closure and Divestment
It is an advantage of divisionalised structures that activities may be closed down or otherwise disposed of without unduly disrupting the rest of the organisation. However such a decision is a strategic one and should not be based on the results of a single responsibility centre, but rather on an appraisal of what the closure or divestment will mean for the whole organisation. (Marconi has recently sold off various parts of the organisation as a cash-raising measure. Firms such as British Gas have split themselves into separate companies to facilitate more focused operations.)
Current and Future Developments
E F Schumacher, the famous author of “Small Is Beautiful”, envisaged a time when the role of the senior management of organisations could be likened to someone holding a number of lighter-than-air balloons – each of these is self-supporting of itself: only the person holding the strings keeps them together and organised as a bunch. The role of management would, in Schumacher’s view, not be to control everything, but rather to facilitate the semi-autonomous operation of the different parts of the organisation. There could even be a point where it would be more effective for senior management to “let go” completely (as noted earlier, shareholders might prefer to hold shares in a number of different companies, rather than in one conglomerate where the “headquarters” added no value to the component parts – or even detracted from overall value added).
In both the public and private sectors, an increasing number of organisations are restructuring along decentralised lines, with a high degree of local autonomy, including “freedom to trade”.
Responsibility centre systems are also useful in the context of the development of “virtual organisations”. For example, in Northern Foods, which among other things supplies ready-prepared meals to Marks and Spencer, and other major retailers, there are over 150 semi-autonomous business units, the managers of which tend to have much closer dealings with their customers than they do with the managers of other sections of Northern Foods. The adoption of these systems means that the “boundaries” between the organisation and its suppliers and customers begin to be eroded.
A particular impetus to the spread of internal markets has been the rapid development of improved information technology at the end of the 20th century. Literally at the push of a button, anybody within the organisation can obtain, for example, data about products and services available from a huge range of sources outside as well as inside the organisation. In the drive for efficiency, internal users of such goods and services are encouraged to obtain supplies from the most efficient source, whether from inside or outside the firm. (One effect of the “e-economy” has been the spread of “disintermediation”. Firms are increasingly able to “cut out the middleman” – to deal directly with firms who can directly supply their needs rather than through intermediaries, whether these be retailers, wholesalers, or other “producers” whose role is really simply to aggregate and assemble the products of other firms – if none of these add significant value. Internal markets are bringing these effects inside firms – departments are only able to survive if they add real value; otherwise internal customers are going elsewhere.)
A Reminder, and a Word of Caution
Remember that at the outset of this module we said that there are a number of ways of co-ordinating activity, of which organisations and markets were two of the most significant. Whether or not an activity would be co-ordinated through an organisation (eg through planning, specialised roles, formalised monitoring and co-ordination) or markets – with decentralised decision makers buying and selling from each other, would depend on which was most efficient.
For example markets are not without their transactions costs of setting up deals, monitoring and enforcing contracts, and so on, and an internal market is not immune from these. In addition, organisations may capture benefits that markets don’t – eg the building of a corporate culture, or the ability to get members to focus on the “common good” rather than their local interests. Too much emphasis on internal markets may increase these transactions costs, and lose these other advantages. Recall that one of the stages that Greiner identified that an organisation may go through was a “crisis of control” where it was felt that too much autonomy had been granted locally, and the organisation felt that the centre had to give much more guidance to the other parts of the firm.
After all the logical end of an unlimited drive towards internal markets, freedom to trade, and so on, would be that every individual would become an independent business, and organisations would disappear.
Read also distributed extract from P Milgrom & J Roberts: Economics, Organisation and Management Prentice Hall International 1992