Harvard is known for many things, its rigorous academics, its crisp New England campus, its secret societies, and now, what may be the most extensive cheating scandal in Ivy League history. A total of 279 students are now under investigation for collaborating on a take-home exam, with the threat of a year's expulsion hanging over their heads if found guilty.
Matthew Platt, professor of the course in question (Introduction to Congress), brought the tests before the school's administration after noticing similarities on a few of the exams, and the investigation mushroomed from there. Students were not permitted to work together on the exam (officially), but now there's a lot of talk about the instructions, the expectations, and the questions themselves being unclear. I would bet that there are a number of aspects to this situation that led to such a widespread web of cheating.
In general, lack of clarity in expectations is a great instigator of dishonesty, after all, when no one tells you what you can and can't do, it becomes much easier to decide for yourself what probably is and isn't okay. For instance, it might seem that asking a peer what he or she thinks a question means if the wording is unclear is pretty reasonable. Then, naturally, that discussion of intent might lead to what the answer could be. In this case, the instructions seem fairly clear, stating that "students may not discuss the exam with others." However, it appears that the professor cancelled his office hours before the tests were due, which would make it a lot more difficult to clarify any questions. This makes for easy justification.
Also, the subject of the class was Congress, which is itself an institution shot through with ambiguity and famous for its lies and liars. Extensive discussion of corruption could easily engender more dishonest behavior in those taking part (in psychology we call it priming, where we expose participants to a stimulus that alters their behavior as a result, for instance, asking people to do math problems when we want to induce logical thinking). It's hard to imagine a better primer for dishonesty than a class on Congress. Maybe one on modern financial institutions.
Moreover, people generally agree that cheating in the social domain is often acceptable--we call them little white lies. Like when a friend asks how she looks in something and you say "great!" when you really should say "passable"; that's often excused from the realm of dishonesty. Or another friend asks what you think of his new girlfriend, and you say "she seems nice!" instead of "she seems boring and self-centered!" We tell these little lies to keep the peace. Yet we generally deny that this is acceptable in the business domain. If you ask your accountant how much money is in such and such an account, giving a number twice as high to make you feel better would be inexcusable. We need to consider that for students, the social and professional circles vastly overlap, which makes it more difficult to separate what's permissible and what isn't. This is not to absolve students who cheat, but it's something to consider. Students often live in the same place they go to class, which is essentially their workplace. Their friends are also their colleagues, and their "bosses" (professors and TAs) are often their friends. All this blending makes can make lines of conduct a bit more indistinct.
None of this is meant to make light of the problem of cheating, or to imply that it's excusable. But if we want to prevent such things from happening again, we need to think about not just the students, but also the system in which they live and operate. Thus, professors need to work on being crystal clear in instructions. Telling students, for instance, "speak to no one other than the professor or your TA about any aspect of the exam" leaves no gray areas. All that said, it will be interesting to see how things at Harvard shake out ...