Loading...
Please wait, while we are loading the content...
Similar Documents
The e ects of performance separability and contract type on agent e ort
| Content Provider | Semantic Scholar |
|---|---|
| Author | Rankin, Frederick W. Sayre, Todd L. |
| Copyright Year | 2000 |
| Abstract | We report the results of an experiment on the in ̄uence of performance separability and contract type on the eort levels of subjects working in an environment characterized by team eects. We demonstrate that the principal can achieve improvements in productivity through the choice of incentive scheme and/or by increasing the degree of performance separability through monitoring activities. We consider competitive, individual, and cooperative incentive schemes and two levels of performance separability. Under both the competitive and individual schemes, eort levels increase as the degree of performance separability increases. Under the cooperative scheme, eort levels are not aected by changes in the degree of performance separability. # 2000 Published by Elsevier Science Ltd. All rights reserved. This paper explores the behavior of agents in a situation characterized by a production externality where each agent chooses an eort level, unobserved by the principal, that stochastically aects his output. The presence of a production externality, which we will refer to as a team eect, creates an interdependency among the agents' performance measures, which can result in an interdependency in compensation. A team eect is de®ned as a production process that is partially separable so that each agent's observed output is a function of his and other agents' eort levels (Drago & Turnbull, 1988). In such cases, agents' performance measures are only partially separable due to a portion of one agent's eort being allocated to another agent's performance measure. Incentive problems can arise if agents are rewarded for less than the full amount of eort exerted or do not bear the full cost of shirking. For instance, in manufacturing with shift-work, if periodic maintenance of equipment and improvements to machinery or the production process increase an employee's output, it is also likely to increase the measured performance of employees in the same position on other shifts. Team eects could be present in a variety of nonproduction based situations, such as team audits, co-authored research, air trac control, and in law and medical practices. While there is some empirical evidence of such incentive problems, according to Prendergast (1996), ``A particular area of incentive provision that remains largely untested is the behavior of individuals within 0361-3682/00/$ see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PI I : S0361-3682(99 )00059-8 Accounting, Organizations and Society 25 (2000) 683±695 www.elsevier.com/locate/aos * Corresponding author. Tel.: +1-314-935-6339; fax: +1314-935-6339. E-mail address: rankin@mail.olin.wustl.edu (F.W. Rankin). 1 This diers from Alchian and Demsetz (1972) and Holmstrom's (1982) notion of team production, de®ned as a completely inseparable production process so that individual's eorts produce a single output. Such cases are rare since even minimal monitoring provides a partially separable performance measure (Drago & Turnbull 1988). teams.'' In related research, Newhouse (1973) ®nds that in medical practice groups where costs are shared, overhead costs are higher and where pro®ts are shared, doctors work less. Bailey (1970) ®nds similar results within medical practices and Leibowitz and Tollison (1980) ®nd that cost control is poorer among large legal practices than among small practices. In the presence of team eects, the principal can improve productivity through the choice of compensation scheme. Holmstrom (1982) demonstrates that with completely inseparable output and under certainty, cooperative schemes alone can remove incentive problems. His scheme requires penalties that waste output or bonuses that exceed output. The principal can also increase productivity through an increase in monitoring. In the presence of team eects, monitoring serves to uncover an individual's productive inputs and hence provides a more separable performance measure on which to base compensation. Alchian and Demsetz (1972) argue that the necessity for monitoring created by the presence of team eects is a fundamental reason for the existence of ®rm, rather than strictly market, organization of production. This research focuses on the behavioral implications, in terms of elicited eort, of simple incentive schemes and performance separability and how they interact to aect productivity in the presence of team eects. We demonstrate that the bene®ts of developing a more separable performance measure depend on how that measure is used in an employee's compensation scheme. In particular, we test behavior across an individual scheme (individual piece-rate), a competitive scheme (rank-order tournament), and a joint performance evaluation scheme (joint piece-rate) across a high and low degree of performance separability. We use a laboratory setting to test the behavioral predictions. The experimental task required subjects to make production decisions as managers of a hypothetical ®rm. Subjects were compensated under a tournament contract, an individual piecerate or a joint piece-rate. In addition, we assigned subjects to conditions of either high or low performance separability. A novel feature of our experimental design is that we include performance separability as a parameter, which allows precise predictions. We ®nd that as the degree of performance separability decreases, eort levels decrease more under the rank-order tournament, than under an individual piece-rate. The use of a joint piece-rate results in the degree of performance separability having no in ̄uence on eort. Under the individual piece-rate we ®nd evidence of social giving that is consistent with previous research on altruistic behavior. Under the tournament contract, we ®nd evidence that dierences in behavior across the two levels of separability result from strategic behavior. Our experiment is designed to test theoretical predictions derived from the economic model. We ®nd the theory to have signi®cant explanatory power, however, there are important dierences in predicted and actual behavior. Kachelmeier (1996) suggests, ``. . .that the potential to add new insights through economic experimentation can be more fully tapped if the experimentalist considers both the economic incentives of a strategic model and the behavioral forces that might challenge these incentives.'' In this spirit, we explore additional behavioral hypotheses in an attempt to explain the variance in behavior across treatments. We ®nd this additional analysis helpful in explaining deviations from predicted behavior. Hence, in addition to demonstrating that the theory does reasonably well at predicting behavior, we investigate conditions in which those predictions may not be accurate. Note, however, that this additional analysis is based on observed rather than predicted deviations from the economic theory. In this sense, these results provide a baseline for the study of behavior in environments characterized by team eects. Future research 2 In related research Choi (1993) derives optimal compensation when a production externality is present. He demonstrates that both joint performance evaluation and relative preformance evaluation can be optimal depending on the relationship between a production externality and an informational externality. Arya, Fellingham and Glover (1997) develop a twoperiod model in a setting where only aggregate performance is measured and demonstrate that it can be optimal to condition both agents' pay on a joint performance measure. Their result depends on the ability of managers to mutually monitor the ®rst period action and punish deviant behavior in the second period. 684 F.W. Rankin, T.L. Sayre /Accounting, Organizations and Society 25 (2000) 683±695 |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://directory.umm.ac.id/Data%20Elmu/jurnal/A/Accounting,%20Organizations%20and%20Society/Vol25.Issue7.Oct2000/3001.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
