Contract Enforcement and Investment: Evidence from a Field Experiment

The intervention consists in sending a contract to teams engaged in teamwork. Team output is graded and counts for 30 percent of the final grade in a real university course. The grade is the same for all students in the same team, which creates a free-riding issue: every student has the incentives to let others do the work and benefit from the team grade. In one third of the teams, students will write a team contract to set firm expectations from the beginning and will send the contract to the professor. In another third of the teams, students will write a contract and send it to a teaching assistant who has no authority to change the grades (to test whether the strength of enforcement matters). A final third of the teams will serve as a control group.

2020-09-10

2021-12-20

The primary outcome is free-riding, measured by effort on the teamwork, measured by:

• questions 10 (Indicate who contributed a fair share to the teamwork), 11 (Indicate who was on time to the team meetings), 12 (Indicate who respected the deadlines set by the group), 13 (Indicate who showed enthusiasm and a positive attitude in the team)

• questions 18,20,22,24 (What was the contribution of student X in the teamwork?) in the questionnaire

• questions 3 (How many times did your team meet to work ?), 4 (How many hours did you work on the teamwork this past week ?)

These outcomes will be looked at only if there is sufficient variation in the data.

I will use a difference-in-differences analysis. Multiple hypothesis testing will be with the Sharpened False Discovery Rate (FDR) adjusted q-values (Anderson, 2008).

The research plan includes subgroup analyses by:

• number of friends before the teamwork (question 16: Who were you friends with before the teamwork?) since a contract might be less needed in teams with a greater proportion of friends.

• first or second teamwork in the course (contracts may not be necessary with repeated interactions)

• high- or low-ability teams (contracts may be less needed when the rest of the team is of high-ability, peer effects may substitute to contracts).

• Team performance: grade on the teamwork

• Academic performance: grades on quizzes and exam, as well as question 5 (How many hours did you work on the rest of the class (lectures, quiz…) this past week ?)

• Opinion about the teamwork (positive experience, stress, friendships after teamwork): questions 1 (Did you appreciate the teamwork assignment ?), 2 (Do you think the work done helped you develop your teamwork skills ?), 6 (Did your group appropriately split the tasks among team members ?), 7 (What kind of problems, if any, did you encounter in your team ?), 9 (Do you think that this work a. Allowed you to acquire new knowledge in development economics b. Allowed you to get familiarized with new methodologies c. Allowed you to increase your knowledge about the topic d. Was globally beneficial for you in your studies), 14 (Was there a leader in your team), 15 (Indicate the members who taught you something during teamwork), 17 (Who are you friends with now), 27 (How stressed were you by the teamwork?)

• Number of complaints to the professor (and to the TAs)

• In one third of the teams selected randomly, students write and sign a team contract, which is sent to the professor for enforcement.

• In another third of the teams selected randomly, students write and sign a team contract, but only send it to a teaching assistant who has no authority to change the grades.

• Finally, in the last third of the teams, there will be no team contract.

The randomization will be done in office by a computer.

Teams of 4 students will be randomly formed. The randomization will be at the team level.

Yes

The planned number of clusters is a minimum of 120 teams.

The planned number of observations is a minimum of 480 students.

40 teams with team contract to the professor, 40 teams with team contract to the TA, 40 teams control.