Effects Of Task Dependence And Compensation Mechanism On Sabotage Behavior In Team Tournaments: Experimental Evidence

As competitions become more and more fierce, Tournament mechanism as anincentive strategy, which is based on the comparison of individual’s relativeperformance rather than absolute performance evaluation, is widely used in many fields,no matter in Sports events such as National Basketball Association、FIFA World Cup, orin promotion system of enterprises and public government departments. With thedeepening of specialization，the organization form of tournament begins to transformfrom individual to team. In a team, the effect of cooperation among members dependslargely on the distribution mechanism of team compensation and team task dependence.Especially in those team tournaments which based on relative performance comparison,team compensation distribution mechanism and team task dependence will make teamcooperation and competition become more complicated. For example, two teamscompete for the same goal (whether obtaining a patent or a contract), the way ofdistribution of the bonus, whether based on the relative contribution size or shareequally, certainly will affect internal team cooperation. At the same time, in the teamtournament, there are two options for a team to win the game: one is to improve theoutput by exerting more real efforts and the other is to reduce output of the competingteam by sabotage. However, it costs to sabotage, so for pursuing personal benefitmaximization, who prefers to sabotage more to compete with the other team, it is thetop member or the weak ones? In addition, the marginal contribution rate of a team willdepend on team task dependence. So who will be sabotaged the most? The strongestin the team (capture the ringleader first in order to capture all his followers) or the onewho is the weakest (Bucket effect)?In order to solve the above questions, the paper intends to conduct an real taskexperiment in the laboratory to discusses the effects of the task dependence andcompensation distribution mechanism on the agents’ efforts level and sabotage behavior.Our research takes the real task experiment into the controlled laboratory, which isdifferent from previous researches. Through controlling the different task dependenceand the conditions of compensation distribution, we design the following fourexperiments to discuss and study the above two variables on the effects of the agent’sbehaviors. They are including: task dependence, team’s compensation distribution onrelative contribution size; task dependence, team’s compensation distribution on average; task independent, team’s compensation distribution on relative contributionsize; task independent, team’s compensation distribution on average.By employing96undergraduate and postgraduate students and carrying out fourreal task experiments, We come to the following conclusions:①Through thecomparison of experiment one, two and experiment three, four, we found that: when thetask is dependent, the agent’s effort level is higher in the case of the average allocationof team compensation, and group incentives should be highly emphasized. However,when the task is independent, the agent’s effort level is higher when the team’scompensation is distributed on relative contribution size, and individual motivationshould be more emphasized.②By analyzing the subjects’ sabotage behavior,We getthat in the dependent task context, when team’s compensation is distributed on relativecontribution size, high-ability individuals are more likely to sabotage others; whenteam’s compensation is distributed on average, low-ability individuals sabotage morefrequently. However, in independent task context, the relationship between team’scompensation distribution and sabotages is not significant.③By analyzing the data ofsabotage, the first and second experiments show that: when the task is dependent, theleast able member of the team, who has suffered the most serious sabotage. And thiswill not be influenced by compensation distribution system. This conclusion isconsistent with the Buckets effect, which considers that the real height or strength of agroup is determined by the lowest or weakest part of the group. On the contrary, thethird and fourth experiments get the conclusion that when the task is independent, thecompetitive team almost puts all sabotages together to the most capable members of theteam, and team compensation allocation mechanism has no influence on this result. Thisis also confirmed the hypothesis that “to catch bandits, first catch the ringleader” in thispaper. Finally, considering this two points of view, we conclude that the degree of taskdependence plays a certain role in regulating the sabotage object.In conclusion, for enterprises, in the designing of the compensation of a team, it isnecessary to build up a dynamic salary system which is based on the characteristics ofthe work, it can make the compensation system of the organization match the taskcharacteristics and fully motivate the employees. In addition, in the context of teamtournament mechanism, the task dependence will affect the choice of the sabotageobject of the competitive team. And the allocation mechanism of team compensationwill lead to the existence of different sabotage subjects. Creating a fair competitionenvironment for enterprises is particularly important. Thus a performance-based team compensation plan designed for enterprise which links income and personal reputationis recommendable.