4.2.1.1. Agency relations and agency theory

It was the contribution of [SPE 71] and [ROS 73] that led to the development of agency theory. In agency theory there are generally two tendencies¹. The “positive” approach is based on the works of Jensen and Meckling [JEN 76]. The second, approach “normative” [BRO 93] is commonly known as the “principal-agent theory”. It originated in the theoretical developments of [SPE 71] and [ROS 73]. The agency theorists [ROS 73, JEN 76, FAM 80, PRA 85] analyzed the issue of risk-sharing between individuals or groups [ARR 71] through integration of the agency relationship. This situation occurs when the contracting parties have different (or even antagonistic) aims, which leads the appearance of collective forms of labor division and re-configuration. The agency theory aims to determine the various organizational forms that will minimize contract costs within an organization. First and foremost, it is interested in the firm and shows how contracts can create cooperation between asymmetric partners, emphasizing the importance of incentive mechanisms.

[JEN 76] understood the firm as a “nexus of contracts”, and examined the ideas of property rights theorists Alchian and Demsetz [ALC 72], who stated that contractual clauses define the internal structure of property rights [LEP 85]. Alchian and Demsetz provided a contractual view of the firm (as well as of any organization), asserting that all the relationships that structure a firm can be considered as agency relationships. They even regard the firm as a particular type of “contract”, somehow denying the notion of organization itself.

Agency theory broadly refers to property rights because their structure plays a key role in both individual decisions and contracts. Jensen and Meckling pointed out that they expect to analyze “the behavioral consequences of property rights as specified in contracts” [JEN 76, p. 31]. According to [ALC 72], individuals and firms are determined by property rights. The decision to acknowledge the complex nature of the firm enables them to reject, on the one hand, the limited scope of microeconomic theory, which assimilates the firm to a “black box” and, on the other hand, the “holistic” paradigm of the organization. Rights share two features: exclusivity and transferability. The analysis of the firm carried out by [ALC 72] fell within a framework based on very specific behavioral assumptions about individuals, particularly on the concept of homo œconomicus, which states that economic agents maximize their utility function while they pursue their individual interest. It is individual behavior that makes it possible to explain the actions of the firm. While transactions correspond to freely negotiated contracts (a priori), the firm has been traditionally defined by the hierarchical nature of its relationships (based on authority and discipline).

Agency theorists have broadened the basis of property rights theory (which had so far only retained employer–employee contracts) by considering all the bilateral contracts established between the firm and its environment:

“The relationship of agency is one of the oldest and most common codified modes of social interaction. We say that an agency relationship has arisen between two (or more) parties when one, designated as the agent, acts for, on behalf of, or as representative of the other, designated the principal, in a particular domain of decisions. Examples of agency are universal.” [ROS 73].

There are numerous examples of this: the case of the manager and the owner; the employee and the employer; the lender and his debtor and the service provider and the client. This relationship can be explained because the agent possesses particular expertise or information. This definition may be enriched by that of [JEN 76], who considers the agency relationship as “a contract in which one or more persons use the services of another person to perform any task on their behalf, which naturally implies the delegation of decision-making power to the agent”.

4.2.1.2. Moral hazard and adverse selection

An agency relationship involves an agency cost which originates in the non-fulfillment of the contracts [KLE 83] established between the actors of the organization. It is impossible to fully predict all the possible contingencies that could take place, as well as to write an optimal contract, unless an extremely costly procedure is used [GRO 86].

This uncertainty, closely linked to information asymmetry, manifests in two ways [LEV 88]. On the one hand, uncertainty leads to the phenomenon of “adverse selection”, described by Akerlof [AKE 70]. Certain characteristics of the transaction are known to one party and cannot be discovered by the other partner without a cost. Below is the famous example of used cars, or “lemons”, introduced by Akerlof [AKE 70].

“The example of used cars captures the essence of the problem […]. Suppose (for the sake of clarity rather than reality) that there are just four kinds of cars. There are new cars and used cars. There are good cars and bad cars (which in America are known as “lemons”). A new car may be a good car or a lemon, and of course the same is true of used cars.

The individuals in this market buy a new automobile without knowing whether the car they buy will be good or a lemon. But they do know that with probability q it is a good car, and with probability (1 – q) it is a lemon; by assumption, q is the proportion of good cars produced and (1 – q) is the proportion of lemons.

After owning a specific car for a length of time, the car owner can form a good idea of the quality of this machine; for example, the owner assigns a new probability to the event that his car is a lemon. This estimate is more accurate than the original estimate. An asymmetry in available information has developed: now the sellers have more knowledge about the quality of a car than the buyers. But good and bad cars must still be sold at the same price- since it is impossible for a buyer to tell the difference between a good car and a bad car. It is apparent that a used car cannot have the same valuation as a new car - if it did have the same valuation, it would clearly be advantageous to trade a lemon at the price of new car, and buy another new car, at a higher probability q of being good and a lower probability of being bad. Thus the owner of a good machine must be locked in. Not only is it true that he cannot receive the true value of his car, but he cannot even obtain the expected value of a new car” [AKE 70].

This situation of information asymmetry about the quality of the product creates generalized distrust, insofar as the characteristics of all the goods exchanged (or likely to be exchanged) are not known by all the actors in the same way. Not even the price can properly play its signal role. The way to remedy adverse selection is to appeal to a procedure that makes it possible to obtain information about the intrinsic quality of a product or service [CAH 93].

The problems of adverse selection result from “the pre-existing distribution of information” [BRO 93]. However, the contract itself may be the source of informational asymmetries. Uncertainty refers to a problem of moral hazard, the origin of which may be twofold. First, the moral hazard reflects the principal’s inability to observe accurately and without cost all the efforts undertaken by the agent [HOL 79]. In a second case:

“The uninformed agent can observe the action but cannot verify whether it is appropriate, because he cannot observe the circumstances in which the action takes place … When there is moral hazard, the problem is to succeed in encouraging the agent who has private information to make an optimal decision for the uninformed individual” [CAH 93].

Some of these actions, which are not observable, can hamper the smooth running of the agreement and thus affect the final result. If these actions were intentional, we would be confronting opportunistic behavior.

Moral hazard is closely linked to the notion of agent rationality [BAR 86]. The “principal–agent” models do not incorporate the limited rationality aspect of transaction cost theory. The theory of the agency postulates a broader rationality than the one developed in Williamson’s analysis. This is what leads Levinthal to describe the principal–agent approach as:

“The neoclassical response to questions raised several years ago by March and Simon concerning the behavior of an organization of agents pursuing their personal interest with contradictory goals in a world of incomplete information” [LEV 88].

Simon developed the concept of limited rationality and established a difference between substantive (or substantial) rationality and procedural rationality [SIM 78]. Substantive rationality corresponds to “rationality” as it is conceived in the neo-classical framework: economic agents make a choice that maximizes their utility from a set of possible alternatives. On the contrary, procedural rationality pushes agents to seek situations that will allow them to attain a certain level of satisfaction in their aspirations.

“In the past, economics largely ignored the process that a rational man uses in making resource allocation decisions. This was possibly an acceptable strategy for explaining rational decision in static, relatively simple problem situations where it might be assumed that additional computational time or power could not change the outcome. The strategy does not work, however, when we are seeking to explain the decision maker’s behavior in complex, dynamic circumstances that involve a great deal of uncertainty, and that make severe demands upon his attention. As economics acquires aspirations to explain behavior under these typical conditions of modern organizational and public life, it will have to devote major energy to building a theory of procedural rationality to complement existing theories of substantive rationality. Some elements of such a theory can be borrowed from the neighboring disciplines of operations research, artificial intelligence, and cognitive psychology; but an enormous job remains to be done to extend this work and to apply it to specifically economic problems” [SIM 78].

The agents are therefore sufficiently rational as to take advantage of any “empty spaces” [MAM 92] left through incomplete contracts. As each agent is able to rationally anticipate the same behavior in the other, he/she seeks to protect himself/herself against these opportunistic attitudes, incurring into a cost known as an agency cost. Jensen and Meckling distinguished between three sorts of agency costs in the principal–agent relationship:

• – the monitoring costs borne by the principal in an attempt to limit the opportunistic behavior of the agent and to ensure that the decisions made by the agent comply with his own objectives;
• – bonding costs (non-pecuniary as well as pecuniary), which are the expenses that the agent is prepared to incur so as to build trust and to convince the principal that he/she is working in an optimal way;
• – residual (or opportunity) costs related to the “loss of utility” suffered by the principal in the event of divergence of interest between the principal and the agent. The existence of opportunity costs stems from the following logic: it is in the interest of the protagonists to reduce as much as possible the loss of value that would result from the opportunistic behavior of an agent, and it would also be productive for them to bring their respective utility functions closer together. If the principal adopted a permanent and total control strategy, the marginal cost of his/her action would quickly exceed the marginal income that the supervision would provide him/her. In that case, abandoning control defines the very nature of the residual costs.

Agency costs can be reduced through the establishment of either monitoring/control procedures or incentive systems. These two measures are considered as alternatives. Incentives may totally or partially substitute surveillance and repression systems:

“Property rights economy and agency economy ignore the permanence of the confrontational character of cooperation, and the subsequent need for creating control and direction systems” [BRO 89].

In agency theory, the principal must find an incentive mechanism [EIS 89a]. An incentive mechanism is formulated as a “set of procedures intended to induce agents to disclose their preferences or information and to accept the consequences of their own activities” [MEN 90].

According to Holmström [HOL 79], this mechanism may correspond to a monetary compensation that will encourage the agent to adopt the best possible behavior. He considered the fact that in the case of team production, the role of the principal is not limited to monitoring the efforts made by the agents. The implementation of an incentive mechanism may solve the free-rider problem.

4.2.2.1. Risk-sharing and arbitration between reassurance and incentives: the outsourcing contract

A subcontracting relationship can be interpreted in light of the results of agency theory. It can be defined as a contract by which a firm (an agent) makes a commitment to supply a principal with productions that are exclusively reserved to him/her and that meet precise specifications:

“In the subcontract, the object is specified ex ante […], but nothing guarantees a perfect execution of the transaction ex post. At the time of the initial pact, the principal only purchases “potentiality”, and not a product. No explicit mechanism can guarantee ex ante that this potentiality will meet the expectation of the sponsor. Only time will validate, or invalidate, the initial contract” [BAU 92].

One of the outsourcing contracts is the “framework contract”, concluded for a period equal to or longer than 1 year. In some cases, this type of contract may be similar to cooperation. In fact, it differs from the punctual contract, representative of “classical” subcontracting, which is periodically renewed in the case of recurring relations (repetitive control). The formula for the framework contract avoids having to regularly renegotiate the initial order and thus incur additional costs. It also saves transaction costs, especially costs associated with contract drafting [WIL 85].

Outsourcing is a principal–agent relationship whose moral hazard is linked to hidden action and information. From the moment that the prime contractor does not produce (or no longer produces) the good in question, he/she has difficulties assessing (or at least, without error) as well as controling the efforts provided by the subcontractor. [CAH 93] pointed out that models with hidden action reveal that the final result depends on two important elements: the effort provided by the agent and an “unforeseen risk”, which can be determined by “nature” (e.g. weather conditions). We then join Laffont’s definition of moral hazard as the “combination of non-observability of action and unforeseen risk over the product that makes it difficult to interpret it in terms of action” [LAF 87].

In addition, the subcontractor possesses specific knowledge which the principal no longer has in relation to the costs of production. Thus, the subcontractor can eventually cheat on production costs. A solution for this would be for the contractor to multiply tenders in order to reduce this information asymmetry.

In general, the subcontracting relationship is accompanied by a higher risk aversion for the subcontractor than for the main firm. It is normally considered that the principal does not have risk aversion in the measure that he/she has diversified assets, which enables him/her to hedge against risks. Moreover, in the case where the principal suffers from risk aversion, the main results will not change radically [SHA 79]. Several cases may arise.

In a symmetrical information situation, the principal may suggest a first-class contract that totally insures the agent. This situation is relatively infrequent.

In an asymmetrical information situation, in general the principal can no longer totally insure the agent. Risk-sharing between the lead firm and the suppliers should then be considered. This possibility was studied by Aoki in 1988 in a model borrowed from [KAW 87]. The subcontract model developed by [AOK 88] is written as follows:

The important variable is that of risk-sharing: α. Extreme cases are those where α = 0 (the risk is totally borne by the supplier) and α = 1 (the principal bears the full risk). If 0 < α < 1, then the risks are shared by the two protagonists. The unit price of the supplied good is represented by p, whereas b is the price initially set by the contractors, or target price. The initial fixing of this price raises the question of the disclosure of costs of the subcontractor, which reveals a situation of information asymmetry. With regard to the re-negotiation, this should be based on the increase of a subcontractor’s production costs and, on the other hand, on the reduction of costs provoked by an increase in the duration of the collaboration [BAU 91]. In this model, the principal offers insurance to the subcontractor on the price of the supply. In return, the supplier must ensure the regularity and quality of the products supplied.

“The keystone of this system lies in the possibility of sanctioning the failing supplier and rewarding the deserving supplier, in function of their respective reputation: there is a struggle to classify suppliers […] The impact of the phenomenon of reputation is considerable” [BOU 94].

In the event that the subcontractor makes innovations, he will have to transfer the benefits of these efforts to the main firm. Under these conditions, his motivation to innovate could be reduced. There is a compromise between risk insurance and incentives [GAF 90].

Agency theory also makes it possible to justify the efficiency of selective or franchise distribution contracts. These are hybrid forms of organization [BRI 97, NOR 88, CAR 91, REY 91] for which the questions of moral hazard, free-riding and the appropriation of quasi-rents arise naturally. Competition in these practices is based on not only price but also intangible assets, such as branding. In the field of distribution:

“Loss leader pricing or free-riding practices are typical examples of the implementation of opportunistic behavior. In the first case, the very image of the manufacturer’s brand may be affected. In the second, the commercial strategy is disorganized. Free-riders represent a category of distributors who live at the expense of their colleagues who locally agree to comply with the manufacturer’s instructions and organize, for example, advertising campaigns to provide information, free demos or maintenance. Benefiting from the favorable external effects thus created, free-riders (who refuse to do the same) can then sell the products at lower prices: consumers will gather information from the former and buy from the latter. Distributors who are loyal to the policy desired by the manufacturer will tend to suspend their offers of additional services due to customer loss” [GLA 92].

Under these conditions, regular checks to discourage free-riding practices are essential. However, the establishment of monitoring mechanisms proves to be very expensive when the number of agents is high. The solution to this problem usually involves limiting the number of authorized distributors. Ultimately, as has been explained by [LEP 89], with selective distribution by an authorized dealer system the temptation of the free-rider is practically reduced to nothing.

4.2.2.2. Arbitration between different organizational forms

The association of agency theory with the theory of transaction costs (see Chapter 2) allows us to provide explanations concerning the compromise that a firm must reach when confronted against certain organizational choices, such as choosing between a merger and an alliance. Mergers and acquisitions (M&A) and alliances are different means of development available to businesses. It is interesting to show how the arbitration between these two types of transactions is done through the combination of agency theory and transaction costs, plus the theory of transition costs introduced by Ciborra. High transaction costs lead a firm to either favor full integration through a merger or establish an agreement with other companies.

Arbitration between integration and alliance is a function of the degree of transaction specificity (idiosyncrasy) and of the frequency of relations (Williamson). In some cases, a merger may be more profitable than cooperation and is justified according to this framework whenever the asset is very specific and the frequency of trade is high:

“In situations characterized by information asymmetries, the existence of specific assets and human resources, for which markets are very imperfect or non-existent, merging may be aimed at absorbing a competitor who holds information, reputation capital, a brand image, a distribution network or management that could improve the dynamics of the acquiring firm. In contrast, the use of market relations to obtain these benefits may take too long, become too costly or impossible” [JAC 89a].

In these situations, the company has an interest in merging. This is the case for integrations made through skill acquisition of suppliers or subcontractors, or thanks to the benefits that customers can offer. According to [HEN 90b], an alliance in the form of a joint venture will be preferred to an acquisition as a means of obtaining assets only when a double condition is met: when the assets in question are “public goods” and when the targeted assets cannot be separated from the undesired ones.

Transition costs introduced by Ciborra [CIB 91] refer to organizational learning. They are defined as the costs that companies need to bear when they have to re-structure so as to develop new strategies. In this case, according to the Ciborra, the choice of an alliance to the detriment of a merger takes place in a situation where transaction costs and transition costs are high (Table 4.1). However, it may be considered that a merger may be preferred to an alliance even in a situation where transition costs are high. Indeed, a merger is often accompanied by a re-organization of the activities of the firms concerned. However, this important re-structuring, which ranges from the composition of the board of directors to the assortment of goods offered, requires significant financial resources and can be assimilated to transition costs.

Finally, agency theory also allows us to shed some light on the choice between alliance and merger. Initially, agency theory focuses on analyzing agency costs within an organization. As Muldur cleverly suggested, the enthusiasm of managers for external growth (takeovers through M&A) is explained by an increase in agency costs, linked to conflicts of interest between the company’s managers and shareholders, which are particularly intense at the stage of strategic redeployment:

“Instead of distributing the available cash flow to shareholders, managers prefer to use it to acquire new businesses. They thus increase their power and the chances of promoting their own employees. But if they allocate these cash surpluses to unprofitable projects, this can provoke an increase in agency costs and, consequently, a reduction in business efficiency. This theory (the agency theory) tends to show that while it is true that M&As may bring about conflicts of interest between shareholders and managers, at the same time, they may be the response to the problem” [MUL 88].

The relationship between two companies involves specific costs, that is, agency costs. In the event of high agency costs, it seems that the merger will be preferred. However, it is possible to imagine that in such a situation the signature of an agreement would invite a stable and lasting collaboration, ensuring the partnership’s continuity. This will require a retribution/punishment mechanism to engage the different protagonists and potentially solve arising conflicts.

The existence of agency problems due to the development of a form of property that separates ownership from management functions is reflected in the theses of Galbraith on the emergence of technostructure in the 1960s. In large enterprises, power is legally in the hands of the owners of the public limited company, but the reality of power belongs to those who make decisions behind the scenes. This is the reason why the notion of management replaced that of entrepreneur. Those who constitute the brain of the company (or “technostructure”) bring specialized knowledge, specific skills and experience to decision-making groups. In this analysis, the object of scarcity is no longer capital, but sophisticated knowledge. As organized intelligence is the new object of scarcity, this would legitimize the exercise of power through knowledge and thus explain the origins of technostructure.

4.2.2.3. Toward a synthesis of the theories of the firm

The Mahoney [MAH 92] approach is interesting in that it studies arbitration between different organizational forms from three theoretical axes: property rights theory, transaction cost theory and agency theory, which helps us to make a synthesis of Chapters 2, 3 and 4. Mahoney began by recalling important theoretical concepts, specifically, transaction costs and agency costs, from which he derived a practical analytical grid.

As discussed in Chapter 2, the theory of transaction costs emphasizes the notion of asset specificity. Agency theory focuses on the asymmetry of information (between the principal and the agent) due to team production, which leads to a problem of product inseparability [ALC 72]. Alchian and Demsetz explain the nature of the firm highlighting the impossibility of ensuring a remuneration that depends on the marginal productivity of each individual factor. A second important variable in agency theory concerns the knowledge of the transformation process or task programmability [EIS 85, OUC 79]. Low-task programmability reduces the effectiveness of monitoring efforts.

According to Mahoney, if we meet three criteria (asset specificity, task separability and task programmability), a firm could be confronted with eight situations: