Content area
Full text
This paper explains why high-powered incentives are more common than low-powered incentives in market arrangements, while low-powered incentives are more common than highpowered incentives within firms. In a firmlike principal-agent framework in which a common principal participates in the multiple agents' production processes with his own productive efforts, social efficiency can be obtained by relative performance schemes when the agents are risk-neutral. We derive a group of relative performance schemes which achieve a socially efficient outcome. They are different in their pay-for-performance sensitivity, ranging from a high-powered pricelike relative contract to a seemingly low-powered promotion-based contract. We show that the high-powered relative contract is the most efficient among the first-best relative contracts when the agents have private information, and the promotion-based contract is the most efficient when the agents' limited liabilities are of serious concern. (JEL Classifications: D80, J00)
I. Introduction It is observed that 'high-powered' incentives are more common than `low-powered' incentives in market arrangements, and `lowpowered' incentives are more common than `high-powered' incentives within firms.l For example, franchisees pay to franchisers fixed upfront fees and royalties which are proportional to their sales, whereas managers of company-owned outlets generally receive fixed salaries which are independent of their performances.2 Lafontaine (1992) finds that in her sample 7% of franchisers charge a zero royalty rate and the remaining 93% charge only around a 6% royalty rate, implying that the franchisees face highly sensitive pay-for-performance schemes.
High-powered incentives found in market transactions are consistent with what standard agency theory suggests. According to the standard agency theory, highly sensitive compensation schemes must be designed in places where incentive problems are prominent. However, low-powered incentives within firms provide a puzzle for the theory since workers in firms also face various incentive problems.3
Based on the standard principal-agent model, several theories have recently been offered to explain why low-powered incentives are employed in firms. They argue that low-powered incentives within firms arise from the incompleteness of contracts (Williamson 1985), multiple tasks of workers (Holmstrom and Milgrom 1991), and obscure performance measures for workers (Baker 1992). These theories basically provide stories of why firms should attempt to reduce their workers' incentives.
The main purpose of this paper is to explain the above puzzling phenomenon from a different perspective. As observed...