The traditional view of the firm is of a unitary entity with a single set of objectives. Neoclassical analysis assumes that the maximisation of profit is the sole objective of the firm, and does not typically look behind the veil of the firm to consider how the interests of individual human agents generate this objective. More recent analysis of firm behaviour has taken account of the separation of ownership and control, by applying principal-agent analysis to the delegation of corporate control by shareholders to managers, and in turn from managers to other employees lower in the hierarchy.
The view of the firm as representing the internalisation of contractual transactions goes back to Coase (1937) and has been developed by a number of other writers, most notably Oliver Williamson. In a sense, any firm other than the simplest will typically represent a vertical structure in that workers of different kinds provide a variety of inputs to production which itself may take place in several sequential stages. Williamson has called this view of the firm, in which organisational structure is seen as providing an alternative to vertical contracting, the ‘markets and hierarchies approach’.
The markets and hierarchies approach attempts to identify a set of environmental factors which together with a set of related human factors explain the circumstances under which complex contingent contracts will be costly to write, execute and enforce. Faced with such difficulties, and considering the risks that simple (or incomplete) contingent claims contracts pose, the firm may decide to bypass the market and resort to hierarchical modes of organisation. Transactions that might otherwise be handled in the market are thus performed internally, governed instead by administrative processes.
The presence of hierarchical or divisional complexity clearly raises the question of whether the pursuits of the firm such as profit-maximisation can be considered on a unitary basis or whether each component of the firm has to be considered separately. It is tempting to suppose that, in spite of different divisions having different objectives, some sort of consensus objective emerges when these are combined. However, empirical studies suggest that this rarely occurs in practice.
The internal dynamics of the firm
One response to the problem which these organisational ramifications raise for the ‘black box’ view of the firm is to regard the concept of ‘firm’ as merely a helpful theoretical construct. In that case, we could argue that while the presence of a vertically integrated structure creates problems for the assumption that the firm is a single profit-maximising entity, these problems are not different in nature from those which have been identified as applying to any firm. So long as profit-maximisation continues to be used as the basis for modelling competitive behaviour, therefore, it is appropriate to use it to analyse any firm, to some extent regardless of its precise organisational structure. However, conclusions about vertical merger derived from models predicated on this assumption may be misleading, given how crucial the question of intra-firm relations is to an understanding of the effects of vertical integration. The issue therefore deserves a certain amount of attention.
Legally, a firm which takes over another firm gains the right to control that firm. Control here means the power to make decisions about all the activities of a firm, including production, pricing, investment and so on. Some of these powers may of course be delegated, so that in practice decisions about the activities of a firm which is taken over are often left with the management of that firm. In spite of agency problems arising from delegation, it might be thought that an integrated firm would behave much as a single firm would.
Contrary to this, some models of vertical integration have tried to incorporate some of the insights of agency models of the firm by suggesting that the allocation of control rights is different in a firm which is the product of a vertical merger than in one which is not. In the model of Grossman and Hart (1986), it is only the right to control the firm’s assets which is transferred by acquisition, interpreted as control over production decisions. The target continues to make autonomous decisions about investment because of the hierarchical structure of the integrated business and because investment choices are non-contractible.
It should be stressed that we assume that separate managers are needed to choose [investments for the two firms or divisions] under any ownership structure (p.706)
ex ante investments ... cannot be specified in the contract either because they are too complex to be described or because they stand for nonverifiable managerial effort decisions. (p.697)
The benefits from each firm’s investment and output choices are also assumed to continue passing to each independent manager without aggregation; again, because of non-contractibility, in this case of profit streams.
After investment decisions are made ex ante and [the two output levels have been chosen] manager i receives benefit Bi. This benefit is again supposed to be nonverifiable and hence noncontractible. That is, Bi is a private benefit, accruing directly to firm i’s manager, that does not show up in firm i’s accounts. For example, Bi might stand for managerial perquisites or effort. A consequence of B1 and B2’s not being verifiable is that it is impossible to write in the [initial] contract that firm 1, say, should transfer its benefit B1 to firm 2. (ibid., p.698)
Thus, while all three elements of a firm’s objective function – investment levels, output levels and profit streams – are non-contractible in Grossman and Hart’s (1986) model, control over output is the only one of these three which is posited to pass from target to acquirer on integration. The intuition for this is unclear, but may be based on the relationship between ownership and control over assets. However, the equation of ‘control over assets’ with ‘control over output decisions’, to the exclusion of control over investment decisions or control over profit streams, is questionable. Grossman and Hart explain the identification of control over output with residual control rights as follows.
The [output decisions] represent rights of control over firm i’s assets, which are assumed to be ex ante noncontractible ... we have in mind a situation in which it is extremely difficult to think about and describe in advance how the production allocation should depend on the ‘state of the world’ ... Since [a firm’s output decision] is ex ante noncontractible, it qualifies as a residual right of control, and our assumption is that the owner of firm i has the right to choose it (ibid.)
Grossman and Hart do not explain why control over investment, being non-contractible ex ante, is not also a residual control right and therefore transferred with ownership of the firm. There is little argument to support the claim that the acquiring firm does not obtain full benefit of the acquired firm’s profits, an assumption which seems in conflict with commercial reality.
Without adequate justification, it is hard to avoid the impression that the hypothesis of selective transfer of control is an ad hoc solution to the problem of why there are limits to vertical integration, designed to generate predicted ineffiencies within a vertically integrated firm. The following quote could be interpreted as confirming that suspicion.
It may be useful to comment briefly on the motivation for our assumption that ai, qi, and Bi are all ex ante noncontractible. We shall see in the next section that, if either the ai’s or the qi’s are ex ante contractible, the first-best can be achieved under any ownership structure, and so the degree of integration of the firms is irrelevant. [...] Hence, in order to develop an interesting theory of ownership, it is necessary to assume that the ai’s, qi’s and Bi’s are all at least partly noncontractible. (ibid, p.700)
There are several reasons why Grossman and Hart’s definition of vertical integration is unusual. First, the production choices of the integrated firm merely serve as the threat point for the Nash bargaining outcome: hence vertical integration does not allow the integrating firm to internalise the full benefits of its choice of production variables. This resembles renegotiation of an initial contract concerning vertical control, rather than integration. Second, it assumes that investment decisions cannot be directly affected by vertical integration, a departure from the transaction cost literature which views vertical integration as a means to alleviate opportunism and the resulting underinvestment.
Pace Grossman and Hart, I believe that the correct way to view a vertical merger is that full control over the target’s assets is transferred to the acquirer. This means that all production decisions and investments decisions should be regarded as under the control of a single entity, and that all cost and revenue streams should be seen as accruing to the same single entity. This perspective seems, indeed, to be shared by Hart in a later paper on vertical integration (Hart and Tirole 1990) in which he acknowledges that the residual control rights perspective is in fact consistent with the traditional view of merger as representing full integration of control rights and benefits.
While there are agency problems within a vertically integrated firm, these can, as Hart and Tirole themselves note, be resolved by appropriate incentive schemes.
... diversion problems are not completely eliminated by integration. In particular, if [firm B] owns [firm A], B can use its residual control rights to divert money from A. However, as long as B diverts on a proportionate basis from both units A and B – and as long as this diversion is less than 100 percent – A’s subordinate manager can be given a compensation package that is some fraction of A’s and B’s joint profit. Given this, A’s subordinate manager will have an incentive to choose pricing and trading policies that are in the interest of the company as a whole.
Another argument can be given as to why a merger reduces conflicts of interest over prices and trading policies. Under integration, a subordinate manager will act in the interest of the parent company, since otherwise he or she will be dismissed. But the pressure on the manager of an independent unit to act in the interest of another independent contractor is to sever the whole relationship with the unit (the contractor cannot fire the unit’s manager alone). (ibid., p.207)
The above discussion supports the idea that a vertically integrated firm should be regarded as a single entity with full control over investment and output decisions, and that this is what gives rise to the benefits that can flow from vertical merger.
REFERENCESCoase, R. (1937), ‘The nature of the firm’, Economica, 4, 386-405.
Grossman, S. and Hart, O. (1986), ‘The costs and benefits of ownership: A theory of vertical and lateral integration’, Journal of Political Economy, 94, 691-719.
Hart, O. and Tirole, J. (1990), ‘Vertical integration and market foreclosure’, Brookings Papers on Economic Activity (Special Issue), 205-276.
