Who is the principal? Who is the agent?

The right which is vested in the agent is that of judging the state. Who is to delegate this right? In dualist systems, because the control of the administra­tion is at stake, the state itself is the principal.

In contrast, a monist system insists, in the first place, on the necessity to protect citizens who indeed are the source of the rules of law - the latter is the unintended result of the interactions between individuals. In this case, the rationale for the existence of the state is not constitutional but, rather, conven­tional or customary. This means that not only does the state have no capacity to impose itself as the principal but also that the principal appears to be the impartial spectator of the common law, in the sense of Hume (1992) or Smith (1976). This conception adopts different forms. Countries like England use procedural constraints which are imposed on the state, under the rule of law.

Different principals may not always have different agents. However, in the case which is considered here, these agents differ substantially. The ‘dualist’ principal creates an agent, the administrative judge, who is a judge in the administration. He/she addresses the objective legality of administrative decis­ions and tries to balance public interest and the protection of citizens. Thus, the discretionary power of the state is controlled and limited by the state itself. This may seem a peculiar way to solve the paradox of government. However, substantive norms, which are a priori defined, delineate the set of possible actions of the agent. These interpretative schemes provide the Kelsenian framework (Kelsen, 1962) within which administrative courts judge and thereby administer. As far as the ‘monist’ agent is concerned, ordinary jurisdictions provide ordinary judges, whose set of possible actions is delim­ited by procedural fairness and stare decisis. Although in practice they are becoming increasingly specialized, these judges none the less belong to the common law judiciary. Some countries may of course have more intricate institutional arrangements.

Other countries have a quite different view of what is covered by adminis­trative law. Such is the case in the United States, where regulation is at the core (Rose-Ackerman, 1992). The aim is an efficient control of economic activities, particularly but not only with regard to public utilities. In principle, separation of power implies that the judiciary should control the legislative and executive powers. In practice, numerous independent and specialized agencies give rise to many agents. Congress grants them a quasi-jurisdic- tional status, but at the same time they may endorse decision-making systems that are similar to common law procedures. Thus the legislator and the impartial spectator may have to share the principalship.

Having defined the prominent characters of this agency problem, we shall now briefly outline the incentive mechanisms and the competition for control of the agency, with reference to the founding cases of France and England.