The Preregistration Mistake

As the full Faculty prepares to discuss Dean of the Faculty William C. Kirby’s preregistration proposal at tomorrow’s Faculty meeting, we hope it will be an occasion for thoughtful, serious consideration of the proposal in its entirety—not just a rubber stamp. The Office of Undergraduate Education (OUE) claims preregistration could lead to far-reaching improvements in multiple aspects of student life—from shortening the coursepack line to increasing the quality of section leaders to alleviating Harvard’s advising woes. But each of these expansive claims ought to be independently scrutinized. Undergraduates have expressed overwhelming disapproval of Kirby’s proposal, not because they do not understand the potential benefits of predicting enrollment figures, but because they see the many flaws and shortcomings of such a quick-fix solution. If Harvard is serious about addressing some of the problems that shopping period presents, then it ought to consider more targeted, fundamental reforms. Preregistration is not the panacea many Faculty would like it to be.

The main concern preregistration aims to address is Harvard’s frustrating teaching fellow (TF) hiring practices. Under the current system, graduate students interested in teaching must contact professors during the preceding semester in a process that former Graduate Student Council President Lisa L. Laskin calls “catch as catch can.” But even when students receive confirmation from professors, there is no job guarantee because fluctuations during the shopping week lead to confusing estimates about eventual enrollment. This complicated problem of predicting enrollment figures, combined with an archaic, decentralized and inefficient hiring process, leaves both professors and graduate students in the dark. Consequently, as doctoral candidate in sociology Felix V. Elwert explained, shopping period becomes the “week from hell.” While undergraduates are leisurely shopping for courses, potential TFs from across the University are scrambling to keep as many job options open as possible. And as Elwert pointed out, “At Harvard, people do evaluate each other by how much they get done, and you can’t do much when keeping your foot in the door [simultaneously for multiple TF positions].” Considering that many graduate students rely on teaching fellowships to put food on the table, the current system is clearly unacceptable.

But Kirby’s preregistration proposal isn’t the answer. If students utilize the proposed “liberal add/drop period” just like they use the current shopping period, it is unlikely that preregistration estimates will be any more accurate—making preregistration powerless to ameliorate the “week from hell.” But even if estimates are improved, Harvard’s TF hiring practices will remain anachronistically disorganized.

There are several worthwhile reforms that Harvard could institute to improve the process. First, and most important, departments ought to make use of TF hiring coordinators to better estimate enrollment numbers—by surveying student interest, as some departments do for tutorials, and better employing the vast amount of enrollment data available. Currently only a handful of departments employ such coordinators—and the coordinators are often underutilized in the ones that do. As a result, only the Core Office—which employs centralized planning—guarantees TF teaching jobs far in advance. Departments ought to be able to do the same, at least with a standard pool of TFs. As Director of the Core Susan W. Lewis explained in a recent interview, predictions based on historical enrollment numbers, when available, are reliable most of the time. While new courses, both in the Core and in departments, will have less predictable numbers, the Faculty ought to remember that one experience with bad predictions does not mean the entire system is flawed. Better coordination and prediction will not fix every problem, but it will be much more useful in the long run than preregistration.

Second, the entire University ought to streamline, standardize and computerize the TF hiring process. Centralizing the job search will allow professors and TFs to find one another much more easily. Standardizing the process would better enable TFs to apply “conditionally” for multiple positions at the same time—allowing them to rank preferences and get the best teaching job they can.

These are the only reforms that can lead to tangible improvements. In a recent interview, Associate Dean of Undergraduate Education Jeffrey Wolcowitz claimed that preregistration will improve the teaching quality of section leaders because, as he explained, the last-minute dash to fill positions leads to less experienced, less interested teachers directing sections. This claim assumes that great TFs are out there—just not teaching the right courses. But adding more time—even if the preregistration enrollment numbers are accurate and hiring begins earlier—does not guarantee improvement.

The major problem with teaching quality is in departments that have too few graduate students to act as teaching fellows for its undergraduates. Another two months with the same flawed hiring system will not help those departments. They require an actual commitment to recruiting quality TFs from outside of the department and outside the University.

Beyond teaching and hiring concerns, the OUE claims preregistration can cure a number of other academic woes, but these claims are unfounded. The first is improved undergraduate advising. The rushed, obligatory conversations just prior to Study Card day are poor excuses for the advice and guidance that ought to be integral to Harvard advising. But preregistration will not alter the nature of those uninspired meetings; it will only shift them earlier—to times that are even more inconvenient for undergraduates. Under the proposal’s current time-table, the new study card signings will occur in the midst of midterms and term papers. And while Wolcowitz is right to point out that “all times of the year are busy,” the first week of classes is the least stressful and hence most productive time for serious advising meetings. (Notably, it is precisely the current shopping period that gives students even this small window of flexibility.) Additionally, these new advising meetings will naturally be less valuable—students might only have a few-line course description at their disposal: Professors are not being required to make any substantive course information—like syllabi—available before preregistration. If Harvard were serious about improving advising, the University would institute minimum advising standards across department lines, including a meeting every semester to update each student’s plan of study.

Proponents of preregistration have also claimed that better enrollment estimates will do away with the need for many courses to be lotteried, and that when lotteries are necessary, they will be conducted earlier and in a more efficient fashion. But these claims are erroneous and misguided. In fact, Lewis pointed out that “the number of courses that are actually forced to be lotteried are quite small.” And many students who are lotteried into a class then decided not to take the course. The result of asking students to opt into lotteries before they have decided to take the class is, as Lewis explained, “very long wait lists and large numbers of students who get in from them.” Preregistration will not make this process any more efficient—if anything it will make it worse. Since students will have even less information on which to make registration decisions, they will be more likely to attempt to reserve a space in a course with a lottery, thereby increasing the number of lotteries.

When courses do get lotteried, and students do get turned away, it is almost always an issue of classroom space. But the system of classroom allocation is even more antiquated then TF hiring. Currently, the registrar’s office only controls about 37 percent of the classrooms on campus. The rest are embedded in individual departments. There is no centralized—or computerized—way for course heads to reserve space, and relocation procedures require multiple forms and much patience. No course on campus should ever have to be lotteried due to space concerns, but the Office of Undergraduate Education is mistaken to believe that simply giving administrators more time through preregistration is a legitimate way to address this concern.

Others believe the proposed changes would facilitate improved distribution of our class materials and books as well as reduce the costs of coursepacks. But preregistration will not actually shift the suggested deadlines for coursepack materials or textbook lists. In fact, Harvard Printing and Publishing Services CoursePack Coordinator Geraldine Barney explained that enrollment estimates are only one of many varied reasons for long coursepack production delays—issues such as the readability of submitted materials and communication with professors are tantamount. But most significantly, preregistration will prove useless in counteracting skyrocketing coursepack prices, which are almost entirely a consequence of copyright fees. Likewise at the Coop, as Store Manager Scott Montgomery explained, preregistration will have no effect on prices. Better enrollment estimates could potentially affect whether textbooks sell out too quickly, but this concern is negligible. Montgomery estimated that with current prediction methods, textbook reorders only occur about five percent of the time, and when they do occur, it is just a matter of days before books are available—during the shopping period, they can often be replenished in a matter of hours.

The crux of the preregistration proposal, however, is whether or not it will provide accurate predictions. Wolcowitz acknowledges that during the proposed “liberal add/drop period,” student course selections will change significantly. But he said the “magic of preregistration” is such that even though students don’t take the same courses they originally select, at other schools, the course estimates stay roughly consistent. Even if one buys into such “magic,” the culture of Harvard does not seem likely to support it. Shopping period is a unique time on campus when the vast majority of students carefully plan they courses they shop, taking care to juggle their requirements, thesis plans, scheduling constraints, and other important logistic and academic concerns. It is the time when students are given the choice and freedom in course selection that they deserve and expected when they matriculated at Harvard. It is the one time at Harvard when a good lecturer is most prized, and teaching—not research—is most valued. This culture of shopping period will not disappear easily; students will pack their preregistration forms with five courses they fear will be lotteried—planning on dropping some and adding the great courses that they find later on—thus subverting the system and making it useless.

Wolcowitz said he hoped the consequences of preregistration would be a pareto improvement. If nothing else he said, “I can’t guarantee that every student will be better off, but we can make many students better off.” But as we’ve outlined above, there are many other reforms that are far more effective in actually improving life at Harvard. Preregistration stands to do the opposite.