I've attended two schools with honor codes. At Rice, where I was an undergrad, we used to pledge in writing not to cheat on exams, all the while taking an in-retrospect-bizarre pride that the "honor code" (and by extension "honor") applied only to schoolwork and not to (for example) pub nights or cast parties.
Serious accusations of cheating resulted in trials whose abstracts were publicly posted, with enough identifying details redacted so that you could never be sure which ~~student-athlete~~ student or course was being written about. Usually you could guess, though, and if someone unexpectedly disappeared for a semester or two, that was a good signal you were right. To the extent I thought about it, I guess I assumed the honor system worked pretty well.
The Caltech honor code is a bit more reasonable (to my mind), in that it specifies that "no member of the Caltech community shall take unfair advantage of any other member of the community." Alas, I attended Caltech as a grad student in economics, perhaps the only academic field in which cynicism is integral to the methodology. Unlike their counterparts in other departments, economics professors behaved (and tested, as much as they could) as if their students had no honor code at all.
Anyway, I never bothered reconciling the two positions. My guess is that some students are going to cheat either way, some aren't going to cheat either way, and the honor code makes some difference at the margin. I also suspect that an honor code raises the cost of collaborative cheating ("three may keep a secret...") relative to individual cheating (e.g. looking in the book when you're not supposed to, surfing the web for answers, and so on), shifting behavior accordingly.
One solution to "the cheating problem" is some version of Bentham's panopticon:
No gum is allowed during an exam: chewing could disguise a student’s speaking into a hands-free cellphone to an accomplice outside.
When a proctor sees something suspicious, he records the student’s real-time work at the computer and directs an overhead camera to zoom in, and both sets of images are burned onto a CD for evidence.
But if you are a good-hearted person, whenever someone insists that a police state is the only way to solve a certain problem (*cough* SSSCA *cough*) you ought to spend some time thinking about whether the problem is itself so bad. Is it so terrible when people get outside help solving their problems?
After all, when I run into problems getting Python to talk to Oracle, or getting LaTeX to typeset my manuscript, no one insists that I figure out the problem on my own. I'm allowed to talk to people and ask questions and see if someone has already solved the problem. In an academic setting these might be "cheating" but in a real-world setting they're how people get things done.
That doesn't mean that the ability to solve problems without outside help isn't valuable. For some (but not all) problems it is, and it's not that unreasonable that you'd want to test for it. (I am skeptical that proctored exams in the surveillance prison are doing this, but the article doesn't give enough detail to say.)
Anyway, rather than forcing people to take examinations naked and gumless so that we can place them into carefully bell-curved buckets, we might consider less-amenable-to-cheating ways to simply give them a thumbs-up or thumbs-down. Oral examinations, for instance, ought to be tougher to cheat on and are probably a much better gauge of what students have learned and understood. (On the other hand, they're much more work for teachers, and they don't facilitate meaningless, fine-grained distinctions between students, so they'll probably never happen.)
Of course, if you ask questions like "why have grades at all?" you tend to get dismissed as a heretic and burned at the stake. (Also, the schools that opt for this approach are mostly dirty hippie colleges that you'd never want to be associated with, which probably scares people away from this point of view.)
Nonetheless, one of the primal fitness blogs I read wrote along these lines recently:
While all are working within the system -- in many cases very effectively -- to change things, there's something that can be done instantaneously.
1. Tell your kids you don't care about them doing their homework assignments.\ 2. Don't ask to see their report cards or inquire about their grades. You shouldn't care.\ 3. Let them know it's fine to pursue a different passion per week until they find their true one, and if it doesn't involve going to a top university, or any university at all, that's just fine.
You can't imagine doing this, can you?
And perhaps the most heretical approach of all is to ask, "Why school?" But I've said too much.