When performance outcomes are impacted by the actions of a team of people working together and when individual contributions are difficult to measure, it may be appropriate to design team based incentive schemes. For example, many business projects (e.g. marketing plans, launching new products) involve teamwork. However, pay-for-performance plans are much more complex for teams and are much more sensitive to strategic manipulation and “gaming” by team members. Thus, managers must be extremely careful in designing team pay-for-performance schemes to avoid negative unintended consequences.
When several members contribute to a project but there is no good measure of any one member’s “output,” there is potential for free riding because it is hard to determine who has contributed the most. Most people can probably remember instances in which team members shirked on group projects in high school or college. When pay is contingent on team output, incentives for each team is muted because (a) low effort by an individual can be easily disguised and (2) the reward for one’s contribution is spread over the entire team so that it has reduced marginal impact on the individual’s pay. For example, for a given output, Q, the team is awarded some dollar amount P . This P would be disbursed among all team members so that each gets P/n for a team of size n . That is, everyone
gets paid the same even if not everyone has worked hard. The bigger the team, the bigger are these problems. Only in the case when team members are fully awarded for their contribution will optimal incentives arise, but this is difficult to accomplish when individual contributions are hard to measure. It is also impossible to pay each team member P because it would simply be too expensive and not self financing (payments will exceed gains from the incentive plan).
One solution is to hire a third-party to manage the team. This third party would promise to pay each member the full amount P , but in exchange, each member has to make an up-front payment to the third party (e.g. entrance fee), but this is rarely feasible in practice.
Also, there is evidence that group incentive plans may lead to high turnover of better performers (Malkovich and Newman). Top performers may feel that it is unfair that incentives for high individual performance should be diluted and spread across all employees in the group, regardless of their contribution. Top performing employees may leave if they feel that there is too much free riding and their “just share” of compensation is distributed to others who free ride.
An upside of team incentives is that it can also signal to workers that cooperation is a desirable norm and are often enthusiastically accepted by workers. Therefore, it is often not difficult to get many employees to buy into a team incentive plan.
Possible solutions to the team incentive problem:
- Have team members monitor each other. But this can be difficult if teammates don’t want to “snitch” on each other. In some cases, it works. For instance, many professional sports teams have designated captains whose jobs are to provide leadership and crack a whip on those who are not trying hard.
- Create awards in which members can nominate teammates. Provide a cash bonus for being nominated. This puts a positive spin on the idea of team members monitoring each other.
- Improve performance measures of individual activities if possible. For example, rather than measuring customer satisfaction of an entire department, see how customers of specific individuals rate their experience.
- Might have both team output goals and individual goals. That is, individual incentives kick in only after team goals are met. This signals that cooperation is desirable and at the same time addresses the free rider problem. However, this requires that individual contributions be measurable.
A more detailed discussion of these issues can be found in the book by Brickley, Smith and Zimmerman, and the book by Milkovich and Newman.
A Note on Relative Performance Plans
Relative performance plans are pay-for-performance plans that are based on how an employee performs relative to other employees. For example, one can design a tournament scheme where those that rank in the 80 th percentile or higher receive a bonus and those that rank below the 20 th percentile are fired. In the broiler industry, tournament style compensation plans are built into the contracts written by integrators for growers. Growers are typically compensated based on how they perform relative to other growers under contract with the integrator.
Are relative performance plans desirable? The answer is it depends on the type of risk facing the employees. If there is a common random factor that affects all employees, then relative performance compensation is desirable. For example, suppose the entire tournament cohort of growers is affected by a common weather shock which simultaneously reduces the productivity of all growers in the group. Because grower compensation under tournaments is determined by relative performance and not absolute performance, growers would be insulated from a common shock. On the other hand, had the growers operated under a fixed performance standard contract, each grower would feel the full brunt of the weather shock because it becomes more difficult for each grower to meet the fixed performance standard. Thus, tournaments are an important instrument for mitigating certain types of risks.
As another illustration, students typically prefer grading curves (relative performance scheme) if there is great uncertainty about the difficulty of a professor’s exams. If the professor writes an exceptionally difficult exam, an absolute grading scale would harm the students. For instance, if nobody scores above 90%, no “A’s” would be awarded and overall class great point average would be low. On the other hand, a curve would protect students from unreasonable exams.
Possible downsides of using relative performance schemes are that they may promote excess competition and employees may try to sabotage each other. Moreover, some researchers have found that people may perceive them to be less fair than absolute performance payment systems so that employers may have to pay more to induce employees to participate in a relative performance plan (Wu and Roe). Thus, relative performance plans should only be used in the following circumstances:
1. When there is a large common random factor.
2. Teamwork is not important.
3. There are clear performance measures for each individual.
4. There is little opportunity for employees to sabotage each other’s performance; i.e. they work relatively independently of each other.
Brickley, J.A., C.W. Smith, and J.L. Zimmerman. Managerial Economics and Organizational Architecture. 2 nd Edition. Boston , MA : McGraw-Hill, 2001.
Milkovich, George T. and Jerry M. Newman Compensation , 6 th Edition. Boston , MA : McGraw-Hill, 1999
Wu, Steven and Brian Roe. “Tournaments, Fairness, and Risk.” American Journal of Agricultural Economics 88(2006):561-573.