Introduction

Organizations are formed by agents working together for a common purpose. Two sets of obstacles may prevent organizations from reaching their goals. First, incentive problems: Achieving a common goal may be difficult because individuals pursue different objectives. Second, even when all agents share a common purpose, they face coordination problems due to cognitive costs. They must exchange information, coordinate their actions, and make joint decisions. Such activities require talking, writing, reading, and thinking – all tasks that require time and energy.

On the first set of problems, starting with the work of Hurwicz (1973) and Groves (1973), mechanism design and contract theory have made huge progress. Contracts are designed under asymmetric information so that principals can motivate their agents to align their individual goals with those of the organization. The typical contract-theoretic contribution strives to find the optimal contract, holding constant the role of the agent in the organization and the informational environment. For instance, in the classical moral-hazard problem, the set of possible actions that the agent can take is given as is the information that the agent and the principal observe; moreover, the relation between agents also is given exogenously. This corresponds to assuming an exogenous solution to the second problem discussed previously.

However, in practice, organizations can choose – at least partly – how to allocate tasks to agents, which monitoring systems to put in place, and which channels of communication to establish. In a world with cognitive limits, these organizational choices come at a cost. Designing the organization requires choosing between alternative structures given the cognitive limits of its agents (or the agents that could be hired).