The firm has a name on the door, but each person crossing its threshold carries a different agreement. One is owed a wage, another repayment, another payment for a delivery, another a share of whatever remains. The company that appears from outside to act with a single will is sustained from inside by people whose claims and purposes differ. To describe it as a nexus of contracts is to make that plurality visible. The description challenges the picture of the firm as an independent creature with desires of its own. It also poses a harder question: what arrangement of these relationships produces the decisions we attribute to the creature?
The contractual account turns the firm’s unity into something to be explained. Jensen and Meckling describe an organization that serves as the common focus of agreements among individuals. The supplier’s invoice, the lender’s claim and the manager’s compensation belong in the explanation alongside the shareholder’s interest. A legal entity permits these relationships to meet in a continuing counterpart; their economic terms determine who gains, who bears a cost and who may act. Calling the firm a knot does not identify those terms. A partnership, a corporation and a cooperative can all contain many contracts while distributing their decisions and rewards differently. The useful question concerns the pattern of the knot, including the rights that hold when a particular participant departs.
Alchian and Demsetz locate one reason for a particular knot in work whose product cannot cheaply be divided among its makers. Two people loading a heavy object produce a result together. The weight delivered does not announce how much each contributed, or whether one carried more while the other eased his grip. A payment based on the team’s total may reward a member for gains another created and leave him bearing only part of the loss his own withdrawal causes. Separate contracts for separately measured products are difficult to write when the products themselves are inseparable. The problem is partly one of discovering performance, partly one of making what is discovered matter to the person whose conduct produces it.
The failure can be seen even when the production possibilities are entirely known. Consider an invented two-person team. Each person can work or rest. If both work, the joint receipt is twelve; if only one works, it is five; if both rest, it is zero. Working costs each person four, and the receipt is divided equally. When the other works, working earns six less four, or two, while resting earns two and a half. When the other rests, working earns two and a half less four, or minus one and a half, against zero from resting. Each therefore rests, although both working would leave four units of joint surplus and two for each. The arithmetic describes a payment problem. It does not describe laziness as a character trait: the stipulated reward makes rest the privately advantageous response to either action by the other person.
A monitor can improve the arrangement only by doing something that pays for its cost. Suppose, still within the illustration, that a third person can verify each worker’s effort, condition a payment of four and a half on working, and provide this service at a real cost of one. Each worker then obtains half a unit by working and zero by resting. With both working, the monitor receives twelve, pays nine and bears the monitoring cost of one, leaving two. The combined surplus is three: one less than the unattainable costless arrangement, three more than the outcome without the monitor. If monitoring instead cost five, the same production would yield a negative joint surplus. A supervisor’s presence is an expense justified by a change in conduct, and that change has to be feasible; giving someone the title does not supply the information the contract requires.
The monitor’s own incentive leads Alchian and Demsetz to a bundle of rights. The person observing others can also shirk. Giving that person the residual after the agreed payments lets him retain a gain from better monitoring, subject to the costs he incurs. Their classical firm combines this claim with a central position in the input contracts, access to information about conduct, the ability to alter team membership and the ability to transfer the contractual position. Residual income alone supplies no machinery for replacing an ineffective member. Central direction alone supplies no reason for the director to use it well. The proposed organization joins rewards to particular opportunities for action, and its advantage depends on whether that combination improves production enough to meet its costs.
The contracts explain the firm only when they explain who can change what.
A contractual origin does not make the rights inside every relationship identical. Alchian and Demsetz’s provocation was that a customer can stop dealing with a grocer, just as an employer can dismiss a worker. Yet stopping business with an independent shop leaves its shelves, premises and remaining relationships in other hands. Removing a manager while retaining the shop leaves those assets available to a replacement. Hart and Moore make this distinction central: ownership can permit a party to exclude particular people from the assets without abandoning the productive organization as a whole. A customer’s withdrawal and an owner’s replacement of a worker consequently need not have the same effects on bargaining or investment. One can recognize that both relationships involve agreement while still explaining the significance of the control assigned by it.
The authority described in the previous plate can itself be one of the things a contract assigns. A worker may agree on a range of tasks while leaving the employer to select tomorrow’s task; a supplier may promise a particular delivery while retaining discretion over the process that produces it. Their relationships differ in where choice remains after agreement. The mere presence of contracts cannot settle that difference, because contracts are capable of allocating quite different choices. Nor does a written agreement exhaust the setting in which those choices are exercised. The legal rules that make a company a continuing counterpart, the rights over its assets and the understandings sustaining repeated cooperation help determine what the contractual words can accomplish.
The nexus view is most revealing when it prevents the firm’s apparent purpose from swallowing the purposes of its participants. An instruction that improves the accounts may impose effort on workers; an investment valuable to shareholders may threaten a manager’s position; a manager’s preferred expansion may reduce what is available to both lenders and owners. Calling all these people the firm does not align them. Each proposed arrangement must explain how their different claims are reconciled, and which conflicts remain. A cooperative can place some decisions with workers; a lender can bargain for restrictions; a shareholder can delegate authority while retaining means of challenge. These possibilities are reasons to examine the constitution of each arrangement, rather than infer it from the existence of a common name.
The island survives the discovery that agreements helped build it. What changes is the account of its sovereignty. The authority within a firm has a structure: some people can direct tasks, others demand payment, others determine access to assets, and some may hold several of these positions at once. A dense tangle of promises becomes an explanation only when we can follow those lines of decision and consequence. The next plate takes up the most troublesome separation in that pattern: the people entitled to the return can entrust decisions to someone whose return depends on something else. The knot can hold the firm together without bringing the manager’s interest fully into line.