Periodic Reporting for period 1 - FeedbackTeamContest (FEEDBACK IN TEAM COMPETITIONS)
Reporting period: 2015-04-01 to 2017-03-31
Many firms employ team-based incentives in production and the use of such incentives can affect worker productivity in different ways. First, team incentives can affect how much effort individuals put into their work, as team compensation introduces incentive to free-ride off of co-workers’ efforts or, alternatively, may encourage individuals to work harder so as not to let down their teammates. A large body of theoretical and experimental work has studied the motivations of workers in team production environments, with experimental evidence regularly finding that free-riding is lower than predicted, especially when teams are engaged in a competition with another team. Second, the use of teams can affect how workers direct their efforts. If workers have complementary skills, then organizing them into teams may enable workers to allocate more time to the tasks at which they personally excel. While productivity gains from worker complementarities is considered a primary advantage of organizing workers into teams, this effect is largely unstudied in the experimental literature. Further these complementarities can affect how teams are formed ""endogenously"" which also has not been studied in the literature. Most studies in the literature assume either that all team members are engaged in the same, substitutable task or that their efforts are perfect complements, our environment allows us to study whether teams allocate work in order to exploit gains from complementary skills as in the real world.
If firms implement team contests with the goal of maximizing workers' output, then the question of whether teammates can self-organize and divide work efficiently is fundamental to understanding the productivity of teams. Overall objective of the project is in addition to the significant contribution to the literature, to contribute to workplace innovation policy of Europe, which aims at “… improving staff motivation, thereby enhancing labor productivity, organizational performance, innovation capability, and consequently business competitiveness”."
Several main insights emerge from the experiment. First, individuals and output-substitutable teams are similarly productive. Second, as predicted, individuals and teammates who cannot collaborate allocate most of their time to the task at which they are weakest; however, they still allocate too little time to their weaker task relative to the output-maximizing prediction. Third, despite the potential for coordination failure, teams in which individuals can collaborate and potentially specialize produce significantly more. Finally, while communication has little effect on team production when teammates cannot collaborate, it has a strong positive effect in collaborative teams.
In the second part of the project, we aimed to test following hypotheses through an experiment: 1) There is assortative matching when endogenous matching is allowed in input- and output-substitutable treatments. 2) In the preference lists, subjects rank more balanced types higher in the output substitutable treatment, and they rank more extreme types higher in the input substitutable case. 3) Average output is higher in the input substitutable treatment where subjects can specialize on the tasks. 5) Average output is higher in the endogenous treatments (under both team production functions) where subjects can choose whom to be matched with.
Several main insights emerge from the experiment. First, although endogenous team formation affects team performance positively in input-substitutable team contests, it does not affect the performance at all in output-substitutable team contest. Team performance in input-substitutable treatments outperforms team performance in output-substitutable treatments due to specialization. Second, as predicted, while subjects rank extreme types higher in input-substitutable treatment, they rank balanced types higher in output-substitutable treatment. Third, when we look at the realized matches both in input- and output-substitutable treatments, we observe most matchings occurred assortatively.
Contribution to the literature:
• This is the first work using real-effort task with induced heterogeneous skills.
• This is the first work showing the effect of specialization in team contests.
• This is the first work showing that subjects do not free-ride in team contests when they should by using a real effort task.
• This is the first work showing the effect of endogenous team formation on team output when team members' specialisation is possible or not.
• This is the first work showing how team members' preferences change when specialisation is possible or not.
Contribution to the policy advises:
• Agents should be grouped into teams while having competitions and should be allowed to divide tasks within a team by themselves.
• Agents should be allowed to form their own teams when they can specialise on different tasks.