If the principle of private autonomy were uncompromisingly applied to the law of agency, only an actually authorized agent could create legal rights and obligations for his principal. The will of the principal would then precisely define the boundaries of the agent’s competence; however, a third party cannot always make a reliable determination of whether the agent has acted within the scope of this authority. Rather, he must often rely upon the principal’s manifestation of the agent’s authority, which may go beyond the more restricted authority actually communicated by the principal to the agent. Often he must rely upon the fact that the agent holds a certain position, such as wife, partner, or employee, in the belief that the normal incidents of authority implied by such a position are present, even though special restrictions on the agent’s authority may have been made. And sometimes a third party must rely upon an earlier declared authorization that has since been revoked internally by the principal. Obviously, a third party cannot be expected to check all the details of the agent’s internal authorization, especially since representation makes sense only when it functions efficiently and since use of an agent is also for the benefit of the principal. Therefore, the risk that the agent has acted without power of representation must be apportioned between the principal and the third party, with individual factors determining who must bear it in a particular case. This compromise is the subject of the doctrine of apparent authority.

It is obvious that the expression apparent authority is inappropriate since in none of the above cases is the competence of the agent illusory. With respect to the third party, at least, the authority of the agent is no less real than it would be if it were in conformity with the will of the principal. The distinction between apparent and real lies rather in the justification of the agent’s acts in relation to the principal. In contrast to apparent authority, real authority is more than mere power of an agent, for along with the agent’s legal power “looking out” (posse) stands his privilege “looking in” to the lawfulness of his conduct (licere). In the case of a real authority, power and privilege to exert the power are coextensive as against the principal, while, in the case of apparent authority, the agent has only the external legal power to perform without the corresponding internal justification vis-à-vis the principal.

Although the expression apparent authority spread from continental Europe, where it was used by, among others, the influential French classicist Robert Pothier (1699–1772), into English law, where Lord Ellenborough applied it in Pickering v. Busk (1812), its development in the two legal systems was independent and went systematically different ways. Nevertheless, the results obtained in concrete situations by the courts of both systems often seem rather similar. A similar disparity marked the development of the doctrine within Europe itself, with the formulation of different rules in the French system, which largely identifies authority and mandate, and in the German system, which distinguishes the two ideas. Another general tendency that has arisen in both legal systems has been to objectify the more subjective aspects of the doctrine, thus making the legal consequences more certain and predictable. Today both Anglo-American and continental courts permit the legal consequences of an agency relationship to result even when the principal is not willfully responsible for the appearance of authority.

Similar situations can also lead to opposite approaches in some matters. Here also it depends in the first instance upon whether the principal or the third party must bear the risk, since in general an agent acting without authority is not personally responsible. The question arises as to whether and to what degree the authority of the agent to bind his principal is affected by the death of the principal. According to the traditional English view (Campanari v. Woodburn; 1854), this event automatically terminates the agent’s powers, irrespective of whether the agent or the third party knew or should have known of the death. While the rule originally might have rested upon the “fiction of identity” between principal and agent, today it is supported more by the technical argument that what a dead man cannot himself do, he cannot do through another, with further support garnered from the so-called doctrine of the meeting of the minds that is necessary for concluding a contract.

In contrast, the continental legal systems have evolved less doctrinaire solutions to this question based more on considerations of protection of the concerned parties. Article 2008 of the French Civil Code even goes so far as to treat all transactions of an agent who acts in ignorance of the death of his principal as valid. The more balanced solution offered by the courts on the Continent, however, is to make the good faith of the third party the determinative factor, since in the usual case the real interest of the agent is simply to avoid any personal obligation.

Disclosed and undisclosed agency

Continental European laws restrict the application of agency rules to cases where the agent acts openly in another’s name. Thus, French jurists infer from article 1984 of their Civil Code, according to which agency is the act of the agent pour le mandant et en son nom (“for and on behalf of the principal”), the negative conclusion that in case an agent does not disclose that he is acting as an agent for a principal, the consequences touch only the “agent” himself. The hidden principal is not concerned by the effects of the transaction at all. Section 164 of the West German Civil Code expressly provides that “an agent, who acts without disclosing the fact that he is acting as agent, is the only one to acquire any rights and is exclusively personally liable.”

In contrast to the continental view, when an agent contracts in his own name without disclosing his principal, the common law allows the undisclosed principal under certain conditions to sue or be sued by the third party. Such conditions include that the agent had power to make the contract and that the parties eventually learn their respective identities. This wider concept of agency has no counterpart in continental legal tradition.

The use of this basic doctrine in the common-law countries gives rise to questions regarding the identity of the undisclosed principal, the election of remedies that must be made by the third party, the extent of the respective liabilities, the right of the third party to setoff (the amount of its own damages from any sum that might be awarded it), etc. A solution to these conflicts of interests must in final analysis rest upon an evaluation of the extent to which the relationship between the undisclosed principal and the agent should influence the contract made by the agent with a third party.

The external, unilateral act of authorization

It is still a characteristic feature of the French Civil Code and of other codifications following its model (for instance, those of Spain, Portugal, Romania, and Brazil and other Latin-American countries) that agency is not recognized as an isolated institution. These legal systems conceive of agency only as a subordinate instance or external effect of mandate. The result is that they consider the power to act as an agent as a mere part of mandate and do not have a general concept of authorization as a distinct legal institution. In contrast to this approach, the more modern codifications of Scandinavia and of such countries as Germany, Switzerland, Japan, Poland, Italy, and Greece draw a sharp distinction between the unilateral organizational act on the part of the principal authorizing the agent to act and the internal contractual relations between the principal and the agent. This distinction, one of the major achievements of 19th-century European legal scholars, is also followed by modern English and American legal writers, even though the classical concept of mandate is unknown in the common law.

The insight that authority can exist independent of the underlying contract, and even without it, opens several new practical possibilities. For example, it explains the rule that the authorization of a minor can be valid if he is in fact mentally and physically capable of transacting business, even though the minor is not competent himself to conclude the (internal) contract of employment for himself that establishes fiduciary duties. Thus, the minor is not subject to the liabilities ex contractu (from or out of a contract) of an adult agent although he is authorized.

A few countries still prescribe a special form for every authorization. According to the Russian and Brazilian codes, for example, an authorization must be given in written form. This general rule, which has been mitigated to some extent by the courts, is influenced by a distrust of informality and by a concern “to protect the individual and to be better able to examine the lawfulness of a legal transaction.” The opposite tendency is followed to the extreme by the German Civil Code. It not only adopts the general principle of “freedom of form”—with the rare exception that a special form is required in those cases where a statutory article expressly prescribes it—but even states that the granting of authority need not be in the form prescribed for the judicial act to which the agency power relates. In spite of this express statutory language, German courts do require compliance with formalities in certain extreme situations under pressure of practical necessity, in order to avoid frustration of the aim of the formal requirement in the principal contract. This attitude thereby approaches the English rule according to which, in most cases, no particular formalities are required, even if the agent is to make an agreement for the sale or lease of land, an agreement that must be in writing. More attention to the connection between the authorization and the act of the agent appears in provisions like that of the Greek Code of 1940, which says that the authorization must take the form required for the legal transaction for which the authority is issued unless the particular circumstances lead to a different solution. Other modern codes like those of Poland and Italy do not even contain the “unless” clause and prescribe only that an agent’s power of attorney must be granted with the formalities prescribed for the contract that is to be entered into by the agent.

The internal agency contract

The internal bilateral relationship between principal and agent—apart from agency of necessity—rests upon what is commonly termed a “contract of agency.” This term encompasses all sorts of contracts of employment including gratuitous services (in those cases in which the agent is willing to render a friendly favour).

This internal relationship between principal and agent must begin with an agreement, since otherwise the principal would have no rights against the agent—even the right to beneficial service presupposes the agent’s consent to serve. The consent of the principal may be given expressly by a written power of attorney or implied by his conduct, such as an established course of dealing. Generally, the formation of the contract of agency requires no formal ritual. Moreover, in many cases the obligation to act as agent is only a part of the greater complex of duties that forms a person’s job, profession, or office (for example, those of factors, bailiffs, attorneys, or brokers).

The basic principle of agency is that the agent, in fulfilling his obligation, concludes legal transactions on the part of his principal. When these transactions result from the authorized legal acts of the agent, the result is that only the principal is bound by them. The competent agent himself remains outside the contractual relations with the third party (but compare “undisclosed agency” above) provided he has not also contracted for himself at the same time—for example, when the third party is not content with the principal’s credit and insists upon having the agent’s liability in addition to the principal’s; in such a case the agent binds both his principal and himself. This general rule that the agent does not become involved may change in the exceptional case of an authorization conferred for the benefit of the agent. Thus, such an agent has been held personally liable by German courts for mistakes made in the course of contract negotiations (culpa in contrahendo).

Rights and duties between principal and agent

Continental European codifications generally do not treat the contract of agency as a separate type of contract. The standards governing the principal–agent relationship must therefore be derived from the general legal rules governing the mandate, the contracts for performing work, employment contracts, and partnership contracts, together with the more specific rules (e.g., of the commercial codes) regulating particular agency relations and the individual transaction at hand.

Differing from this general approach, the common law has recognized a number of more specific rights and duties between principal and agent. The principal must provide the agent with a regular opportunity for service under the contract of employment and has a positive duty to aid, and not inhibit, the agent’s performance of such service. He must compensate the agent for his services, provided they are not gratuitously rendered (the agent’s right of remuneration), and must account for those amounts due to the agent, including indemnification for economic loss suffered by the agent on account of the agency relationship (right of indemnity and right of lien by way of charge on the principal’s goods in his possession). The principal also has the more abstract duty to conduct himself so as not to harm the agent’s reputation.

The primary duties of the agent to the principal are those of care, obedience, and loyalty—similar to those of a trustee. More specifically, the agent must act solely for the interests of his principal and therefore must account for any financial benefit (“secret profit”) he might derive from a transaction. It follows that it is a violation of an agent’s fiduciary duty to his principal to have unrevealed interests adverse to those of his principal. He must not compete with his principal and may not use or disclose confidential information except for his principal’s benefit. If an agent has received money or other property from or for his principal, he must account for it. An agent also may not normally delegate his task to a subagent, since the principal is assumed to have placed his confidence in the person of the agent and not in a subagent. Finally, the agent also has the abstract duty of conducting himself so as not to bring disrepute upon the principal.