It's not their fault--attending the discussion sessions is not required. It has never been since. But it was different this time because in the previous semesters, attendance was perfect. The reason for perfect attendance was the weekly quizzes. We use the weekly quizzes as what Oliver E. Williamson might consider as "hostage." Students have to attend if they want to take the quiz. The quiz, of course, matters for a student's final grade.
It is different this semester. The quizzes are administered online and during the weekend. So for students who may not really care about attending discussion sessions or who think that they can breeze through the semester attending the lectures and studying the class textbook by themselves, there is really no incentive to attend.
What is the transaction here? You may say it's the fact that the student decided to pursue their respective chosen degrees. But we can look at a more micro level. We can consider the pursuit of a degree consisting of further individual transactions, namely the enrollment in different subjects. Students enroll in a subject with the sole purpose of obtaining a benefit--passing and getting a good grade (and hopefully learning as well)
Ex ante costs aside (like monetary and non-monetary costs involved in the process of enrolling), let's concentrate on ex post transaction costs. Definitely, among such transaction costs are typical day-to-day activities students undertake going to classes and studying for tests--all with the ultimate goal of eventually passing the course and getting a grade
If attendance is not required in either lecture or dicussion sessions, and if students feel attending only lecture classes is sufficient enough to get a good grade, then for some students there is really no point in attending discussion labs. So we are really in an institutional arrangement where not attending discussion labs is a rational choice for students.
Now, if we change the arrangement, we should definitely expect a change in the behavior of the students. This turns out to be the case when we instituted a policy wherein there is an incentive for students to attend lab classes. For the remaining lab classes until the end of the semester, we will check attendance. At the end of the semester, we announced that we will randomly select two sessions where we will award 5 points of credit toward the final grade for those who are present during that session. While the effect is not a perfect attendance, there is indeed a significant change in the number of those who started to attend.
We can certainly make the case that there are three types of students in this case. The first type are those who regularly attend, and these are the students that are really interested in learning. The second type are those that started to attend after instituting the new incentive scheme. We can say that for these students, they value a higher grade and the transaction costs involved in attending the lab sessions are low compared to the benefits of getting that higher grade.
In retrospect, you wonder if the differences in the behavior of the students reflect the fact that the benefits one student gains is subjective, or could it be that the transaction costs turned out to be the one that is subjective in nature. Let's face it, some students are really constantly determining if that extra incentive is really worth all the hassles of going to lab sessions. Or maybe some students are really just plain lazy.