I’ll be doing my Read column tomorrow (Saturday) and maybe a brief cooking column on Sunday, but I’m kind of behind on everything after my travel this week.
Today, a seemingly simple issue that hides a complicated mess behind it: managing the scheduling for courses for an entire faculty.
The seemingly simple job is to take every course being taught in a given semester and assign it a time and a classroom. I’m sure anybody who has studied the travelling salesman problem and its many iterations is instantly aware that this isn’t a simple task at all even if you reduce it to just matching X numbers of courses to Y available classrooms within a Monday to Friday schedule where you can use all times between 8am and 8pm.
Add variations in how long a given class needs to be, in the judgment of the faculty or the administration. (A judgment which is heavily influenced by forms of agreed-upon standardization within higher education like the Carnegie credit hour.) Add whether a class needs to have a lab session, a language drill, a performance or recital, etc. that needs a place in that schedule. Now it’s really complicated.
But not yet as complicated as it actually is, because the real challenge lurking ahead is “who wants to teach at a particular time? Why does a given faculty member believe a course should be structured in a particular way (say, in one 3-hour session; in two 90-minute sessions; in three 60-minute sessions; in five 35-minute sessions?) Why do some classes need a lab, a drill or a recital added on and not others?”
And then, “Which classes are requirements and which are not? How are classes within a particular department (and thus major or minor) to be arranged? Which classes are very similar in theme or subject despite being in different departments? In general, what’s the relationship between all the courses scheduled at a given time: if you put a required course that most students feel obliged to take (say, Introduction to Computer Science) at the same time as a required course for something like an introduction to a foreign language, are you wiping it out or accurately sorting between two populations of students who don’t overlap?”
At this point, it starts to become plain that this isn’t just a problem that can be solved algorithmically.
In this column, I often complain about administrative leaders, political leaders, trustees, etc., but the big problems with scheduling at many institutions originate in the faculty. This is the secret landscape of conflict and anguish within many institutions, not just between departments but within them. (Though it is also one of the key focal points where adjunct faculty get really screwed, e.g., being scheduled for a class and then having that class yanked away from them after they gave up other possible gigs.)
There’s two major subterranean issues roiling away underneath scheduling. The first is that a lot of faculty are trying to arrange a weekly schedule that solves work-life balance issues. Aside from classes and meetings, a lot of the rest of the work that we have to do can be time-shifted around. (Though office hours generally do need to fall within the regular teaching day and the regular working week.) Some of the work-life balance issues are inflexible, like caregiving. Others are at least slightly less so in some circumstances, like handling a long commute. (Some commutes are unavoidable, and some are pretty voluntary. It depends.) Often faculty are trying to protect at least one day to do grading, to do writing or research, to handle other scholarly obligations like peer review, during a regular working day rather than having to squeeze that into evenings or weekends.
The second underlying issue is that faculty have many opinions about the consequences of other peoples’ scheduling for their own courses. Ask someone who had a class cancelled because of low enrollments why that happened and they’ll often say “Because those assholes over in that Other Department scheduled the Huge Required Class at that time” or “Because that one jerk in the Other Department has a class very similar to this one and he scheduled at the same time and the students chose him instead.” Sometimes they’ll be absolutely right, sometimes they’ll be plausibly right and sometimes they’ll be wrong. It doesn’t matter much because beliefs in this domain are often extremely fixed and based on little specific evidence.
You might ask why anybody should care about either issue: if the sausage of a schedule gets made, who cares? But there are three kinds of consequently bad outcomes that really matter and it is a serious challenge semester-by-semester to prevent them. It is not just why universities have registrars but also to a significant extent why they have department chairs.
First is that within departments, if there isn’t a serious counterweight towards equity in scheduling, what inevitably happens is that a small number of privileged faculty will continually grab the timeslots that make their lives easier. Generally the prime times in terms of maximum work-life flexibility are Tu-Wed-Thurs between 10am-3pm. When scheduling tips out of balance, whatever the kind of inequity it is that takes hold (privileging seniority, privileging voluntary commuters, privileging whomever has the sharpest elbows or is prepared to be the biggest pain-in-the-ass) will end up pushing the losing department members into undesired slots (very early morning, very late afternoon or evening, Fridays, etc.) though sometimes there are in fact faculty who prefer those times (especially evening classes) for their own reasons. A stable inequity of this kind is a bad outcome in its own right.
You might ask, well, why not just put all the classes in the preferred time slots? Instant equity! That’s where the second bad outcome kicks in, and it’s a much more serious one, which is very seriously short-changing students, perhaps even to the point of making it impossible for them to complete a course of study in two or four years, depending on the type of institution and degree being sought. There is nothing worse from a student’s point of view than looking over the schedule for the next semester and seeing that there are plenty of classes they want or need (or both want and need) to take and realizing that they are all offered at the same time. When neither department chairs nor registrars’ offices are effective in pushing courses into all the available time slots, a university or college is at the precipice of failing its students. When everything that’s pushed out to the times deemed less desirable by most faculty are the courses of faculty who weren’t able to defend themselves against more powerful or influential faculty, that’s also a point of serious failure, simply because that often leaves many highly needed or desired classes conflicting with each other during the most congested times.
Which uncovers the third bad outcome, which is when related courses in under-enrolled disciplines are competing with each other in the same time slot. It’s bad enough when any student is looking at that kind of congestion, but when you have a major in the humanities who has a strong interest in an area of study that has an interdisciplinary reach who looks and sees that every class in that area of study is at the same time, then the institution has simultaneously underserved that student and compounded an existing issue with the distribution of faculty resources—a student who might have taken four courses on an aligned subject is now forced to just settle for one and then find three less satisfactory choices. It’s less of a problem with heavily over-enrolled courses, even though it’s never good for a student to feel like they have to take a schedule that is of little interest to them because there’s no other choice. It’s also just bad for faculty with shared methodological and topical interests to be engaged in direct competition for a zero-sum number of possible students within a given time slot—that leaves bad feelings which spill over into other domains. Faculty have folkloric ideas about whether Department A’s scheduling is causing Department B’s problems, but there’s nothing folkloric about the consequences of scheduling three courses on the politics, culture and languages of Central Asia from different departments in the exact same time slot. That’s forcing a choice that the professors teaching those courses should always try to avoid.
Equity in departmental allocation is simple to solve in the sense that it just takes a transparent rule or shared commitment to equitable load-sharing. (Which in the end is often not simple to maintain, given how much latitude individuals have to insist that their circumstances require exemptions from equitable rules.)
But excessive clumping up on popular time slots and attention to spacing out classes that are drawing on the same constituency is much harder to figure out. In a lot of institutions, the classes and schedule requests come in from departments (it is arguably the main administrative function of departments) but that means there is no obvious way to coordinate between departments when the schedules are being drawn up. Individuals who are aware of having similar interests might check in with each other every semester, but that’s a lot of work and in big faculties, you can’t be expected to know of all such possible overlaps. Departments that know they’re overlapping might have a conference between chairs but that’s also a lot of work and given that the issue at hand involved rivalrous interests, it doesn’t take much for that to spiral into dysfunction. Divisions or clusters of departments can do some coordination but at that level it starts to be hard to pay meaningful attention to the subject-level overlaps. A registrar certainly can’t be expected to identify most of the instances where two, three or more courses on similar topics are in the same slot unless there’s some kind of tagging convention in use that helps as a finding aid, and even if they do, having to renegotiate schedules with large numbers of people after those schedules have been set for the first time is a tough sell.
This might be the rare problem where a really good administrative computing system could help, e.g., if you had a ‘tentative scheduling’ tool that every faculty member and department could see, a deadline for entering tentative schedules that was well before the date that the schedule had to be finalized, and a tagging system that was sensitive enough to flag topical or thematic overlaps, you might have a way to help people clear up the logjams before they get out of hand. But even the best system (and I’m aware that saying ‘really good administrative computing system’ is already kind of a laugh within the real world range of choices) can’t guard against people who don’t use a tagging system carefully or who always choose ‘betrayal’ in prisoner’s dilemma games (e.g. withhold information with rivalrous intent).
If there’s any institution out there that has this problem completely beat, I’d love to know about it, because what I hear when I look at student conversations across a wide span of institutions is constant aggravation with schedules that massively overlap in a small number of highly contested time slots. Sometimes this might just be a nuisance, or a kind of self-inflicted wound. But when it drifts into greater dysfunction, the whole problem can outright prevent some students from achieving their goals.
Remind me to tell you about the year that I was chair of the Educational Policy Committee, and tasked with the problem of why 75% of our courses were scheduled on T, Th from 10-2.
My finest moment was sitting down with the English Department, which refused to make any changes at all, as they explained to me that pedagogically, it was *impossible* to teach any of their lower-level courses on M, W, F in the morning.
My response? "How interesting! Because Swarthnmore teaches Shakespeare in that slot. And Oberlin teaches the 20th century American novel in that slot. And Williams...."
And then I said: "What are they doing *wrong*?"
Crickets.