We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Delegated Expertise.
- Authors
Demski, Joel S.; Sappington, David E. M.
- Abstract
We have presented a principal-agent model with structural features designed to capture multiple facets of the managerial task and costly communication. To make modest headway in exploring the setting, we have confined ourselves to two activities (planning and implementation) and prohibitively costly communication. To focus on the idea that different activities carry different degrees of "consumption" for the manager, we have assumed planning is personally costly while implementation entails no disutility for the manager. Casting the model in policy space, we have a standard principal-agent model, to which the stock of available results and insights may be applied. Focusing on the additional structure of the, model, however, allows us to exhibit additional nuances. For example, the most efficient way to control the moral hazard surrounding the expert's planning activities may entail a spillover of motivation concerns into the implementation activities. Casually put, we may wind up with a delicate balancing of risk-return considerations in the implementation phase simply as a result of motivating the desired planning activities. In a larger sense, we should expect a web of responses to the overall control problem. Our focus on induced moral . hazard was designed to illustrate this point. Similarly, costly communication from the informed expert is a friction in our model. And one of the responses to blocked communication may be to alter the implementation activity. Again, then, we see a mixing of efficient responses to a contracting friction. Numerous extensions could be made to the general ideas of our structural approach. For example, placing private information in the hands of the principal would allow for a study of communication timing questions, such as whether it is most efficient for the principal to release information before the manager becomes informed, after the manager becomes informed, or perhaps to suppress the information until it is time for payments to be made in this way, substantive questions regarding the most efficient time to revise a budget arise. As a second example, a principal in a hierarchical setting could be viewed as a nonexpert who receives no private information. The direct subordinate of the principal could be thought of as another nonexpert, but one with access to some private information. The agent of this subordinate might best be regarded as the expert of our model. It turns out that truthful communication in this hierarchy is not necessarily optimal. And the important communication timing issues noted above would persist in the hierarchical setting. A third extension of our analysis would allow limited or partial communication between the expert and nonexperts. Though an expert is often not able to detail costlessly all. of his knowledge to other organization members, some coarsification of this information may be communicable. It would be interesting to explore in detail the optimal trade-off between the costs and benefits of coarsified information transmission. A fourth extension, permitting the principals to make productive decisions following the expert's choice behavior, would allow us to analyze' a setting in which firms explicitly compete in the product market. Finally, movement to a repeated setting would allow the introduction of dynamic aspects of the expertise setting-in particular, competition among experts as they develop reputations over time.
- Subjects
EXPERTISE; COMMUNICATION specialists; CONTRACTING out; PLANNING; STOCKS (Finance); PAYMENT
- Publication
Journal of Accounting Research (Wiley-Blackwell), 1987, Vol 25, Issue 1, p68
- ISSN
0021-8456
- Publication type
Article
- DOI
10.2307/2491259