This semester I am a Teaching Assistant for Stat 2120: Introduction to Statistical Analysis. Coincidentally as we talked about the principal agent problem in class yesterday, I experienced this in my lab section on Monday. In a classroom with 90 students, there are four undergraduate TA's, and a PhD TA. We the undergraduate TA's are paid for showing up, regardless of how many questions we answer. This creates an incentive to shirk , by not answering students questions. When one of the four of us shirk, there is a marginal effect, but the slack is easily manageable to be carried by the three other TA's. However, when three of the four TA's shirk, it is evident that there is too much slack for a single TA to carry.
Unlike politicians, who can fail to create policy and face little to no repercussion for an absence of legislation beyond losing reelection, the TA's have a quota of questions we must fill; the students will ask questions they have, and still want their questions answered regardless of the number of TA's doing their job. The overwhelming amount of slack for a single TA to carry combined with the students' collective opportunity cost of their time spent waiting for help harms the students. Perhaps they choose to forego their question, because waiting 10% of the duration of the lab to ask a simple question is a waste of their time.
In this case, the Professor (who does not attend lab sections) is the principal, and the TA's are the agents. The TA's getting paid to sit in the corner, face the wall, and do homework for other classes are shirking, and both the students and the TA's choosing to help the students are harmed. You'd think that the PhD student would serve as a monitor of information about the TAs' performance in lab, but he fails at this task because he is too nice. While implementing some sort of method to record how many students a TA aids in lab may be tedious, it could cut down on this problem. The statistics department could make TAs reapply each semester to continue teaching, as a form of shortening their contracts of guaranteed employment, or define more clearly that it is expected TA's won't ignore their students.
1 comment:
Hi, Brandon! I've noticed certain classes suffer from less-than-perfect TA engagement, but I never considered it as a principal-agent problem until now. That being said, it makes great sense considering that there rarely seems to be an opportunity for recourse. The most common tool students have for evaluating TAs comes at the end of the semester in Collab Course Evaluations. Therefore, professors lack monitoring tools besides spending their valuable time observing the TAs. In our discussion of the principal-agent problem, we considered Kalt and Zupan's claim that shirking among politicians will likely decline as a reelection nears. Perhaps this notion could be tested with TAs.
It would be interesting to see if the TAs had a job review based on student feedback towards the middle of the semester and at the end of the semester. Perhaps a survey evaluation by students could be used to determine bonuses based on performance, or a ranking system that could be factored into their applications for future TA positions. The two impending dates in which a survey would impact the TAs' personal goals (bonuses, future positions) might incentivize them to match their output with the expectations of the principal. In this manner, maybe the professors can overcome their inability to monitor their agents, reducing the cases of shirking.
Post a Comment