Too much to manage

Any organization wages a never-ending war against chaos. It commands an army of rules, forms, processes, and procedures to force the naturally occurring complexity in a set of manageable, similar things that can be dealt with in a uniform manner. Otherwise, chaos will take over and make any mission impossible to implement. Chaos is simply complexity that got out of control. Universities, for example have courses, programs and other requirements fixed in catalogues, and schedules that map these out in repetitive time blocks called semesters. Students aided by an army of advisers must turn the complex catalogue information and apply it to their schedules, while taking into consideration their jobs, families, and other obligations as well as availability or scarcity of class offerings. Like in any war, it is important to not underestimate the enemy. Sometimes complexity is too great to manage, chaos becomes inevitable, and it is time to retreat.

Here is one example. Many of our freshmen cannot translate the catalog requirements into a sensible schedule in their first year on campus. Advising notwithstanding, they make so many errors, that the chaos creates real damage to their academic careers, sometimes delaying graduation by years, and sometime derailing their college plans altogether. To reduce the errors, we implemented a program of block scheduling, where every freshman receives a pre-created schedule. This was designed to reduce the errors and relieve freshmen orientation anxiety. That is a wonderful, sensible idea, not at all unique to our institution. However, we discovered that creating over 3000 individual schedules for 64 bachelor's degree programs with 70 concentrations, and registering them all was… hard. OK, it put our resources on the brink of total collapse. Why? – Because the entire university registration system was designed around the minuscule act of a student registering for a course. The system was error-rich, but required no direct management. Once we centralized it, we became like the Soviet Gosplan, a body that was so spectacularly unable to cope with running the huge planned Soviet economy. While market economies are prone to terrible errors, and unintended consequences, they are not trying to manage the unmanageable. In the long run, the distributed self-regulation works. I remember when someone in Gosplan forgot to plan for toothbrushes, and the entire country went into a panic-fueled buying spree that Soviet stores of any toothbrushes for years. Then it was sugar, pantyhose, toilet paper, jar lids, and almost everything else. A contemporary economy with tenths of thousands of consumer products is too complex to manage. Student schedules are very close to that, although the university can figure it out, albeit with much effort. Like any borderline situation, its value is in demonstrating the limits.

Now we are thinking about the Fall re-opening. We do not know what the health regulations will be in effect: do we still need to maintain the 6-feet social distance and an hour break between classes for cleaning? Or will it be just a request to reduce the campus population to a certain level? Will we be asked to impose an absolute cap on class size? We do not know how many students, faculty, and staff will not be able to return for health reasons, and how many will not want to return, and what kind of policies we may be able to have to compel people to return. Most importantly, we do not know when we will know what we do not know now. With seven thousand sections, any major revamp of the master schedule will take many weeks. No one knows what the solution should be, and I am no exception/ However, I am pretty sure it should not involve a team of tired chairs and associate deans manually assigning students to classes and classes to rooms. The quest is for a solution that would allow for thousands of informed micro-decisions, and somehow make the whole puzzle work relatively quickly.

This is just in case you are wondering what they mean by “Planning for the Fall re-opening.”