Principal-agent model

The principal-agent model appears for example when political decision-makers instruct executive bureaucrats to undertake certain actions. Think of the EU ministers (the principals) who determine a political position in trade negotiations, for the European Commission (the agent) to defend in its negotiations with global counterparts; or the IMF Board of Governors or Executive Directors instruct the powerful IMF staff to implement certain policies. There is always a certain tension in the relation between the principals and agents: to what degree maintains the agent some formal or actual autonomy, and to what extent will the instructions be faithfully executed?

In its broadest form, Waterman and Meier argue that ‘[…] agency theory assumes that social life is a series of contracts. Conventionally, one member, the 'buyer' of goods or services is designated the 'principal,' and the other, who provides the goods or service is the 'agent'—hence the term 'agency theory.' The principal-agent relationship is governed by a contract specifying what the agent should do and what the principal must do in return […]’ (Waterman and Meier 1998).

Waterman and Meier apply the model to political science, and they contend that ‘politicians (principals) and bureaucrats (agents) do not necessarily share similar goals […] If we assume that they are rational utility maximizers (politicians maximizing reelection chances and bureaucrats maximizing budgets), politicians have an interest in policies that benefit their constituents but have no interest in paying excessively for them. Because politicians and political coalitions change over time and bureaucracies develop separate interests through institutionalization and changing external relationships, a potential conflict occurs when the goals and objectives of principals and agents are at odds. Over time politicians may seek to alter established policy toward their preferred objectives, which may or may not be the same as those of the original legislation or political coalition. Bureaucratic interests also diverge from the original policy through time as politicians develop expertise about how the policy should be implemented or gain support from constituencies that favor different approaches […]’ (Waterman and Meier 1998).

Miller states that the principal-agent theory allows political scientists to explore political relationships—particularly asymmetries and incentives—between the principal (i.e. Foreign Minister) and agent (an actor in the field, i.e. Foreign Ministry staffer) (Miller 2005). Shapiro argues that this model serves ‘[…] in examinations of international organizations (e.g., central banks, international courts, the European Union) in the literature on comparative politics and international relations. […] Principals delegate to agents the authority to carry out their political preferences. However, the goals of principals and agents may conflict and, because of asymmetries of information, principals cannot be sure that agents are carrying out their will. Political principals also face problems of adverse selection, moral hazard, and agent opportunism. So principals contrive incentives to align agent interests with their own and undertake monitoring of agent behaviour’ (Shapiro 2005).

The principal–agent model starts from the observation that in public policymaking, authority is often delegated by one political player to another. “[T]he delegated authority is the authority to negotiate and to reach international agreements on trade or on the environment. Principal–agent scholars recognize that delegation comes at a cost for those actors who delegate, but that its benefits exceed this cost […] Applied to the principal–agent relation between member states and the EU negotiator with regard to international negotiations, the ex-ante control mechanisms refer to the (contestation of the) authorization of the agent on the one hand, and the instructions granted by the principals to the agent on the other hand. Also two ad locum control mechanisms are distinguished. On the one hand, the member states can often attend the international negotiations, which makes it possible to observe and to monitor the negotiation behaviour of their agent. If this happens, member state representatives sit behind the officials of the EU negotiator in the negotiating room. In this capacity, the member state representatives are often perceived as ‘mothers in law’ by the agent, which suggests that their monitoring function is taken seriously by the agent. On the other hand, the principals frequently meet with the agent during the course of the international negotiation sessions with the purpose of reinstructing (or even warning) the EU negotiator’ (Delreux and Kerremans 2010).

A critique of the PA model comes from Jankauskas et al., who argue that the linear relationship between the principal and agent is not without problems. They challenge the view that the more control a principal has over its agent, the less likely the agent is to slack (deviate from a set policy, or instructions). They state that former assumption does ‘[…] not always hold, because the intrinsic nature of the agent and how it actually behaves may vary despite formal characteristics of the delegation contract. While control may reduce the agent’s room for maneuver, it does not explain the extent to which different agents are inclined to put efforts in circumventing these obstacles. There is thus a missing element in conventional PA theorizing, and it relates to the oftentimes neglected varying nature of the agent. [They state that] the risk of agency slack is not a mere by-product of (missing) principal control but is also related to the intrinsic nature of the agent. Two implications stand out from our findings. First, not all agents demonstrate high risk of agency slack if not harshly controlled. Second, control does not always automatically constrain the agent’s behavior reducing its propensity to slack. […] The analysis shows that low control by the principal is not necessarily associated with run-away agents, whereas high control is not necessarily associated with servant-like agents. The assumed control–slack relationship can thus be distorted and determining an ideal level of control is not possible without considering the agent’s entrepreneurialism’ (Jankauskas et al. 2010).