As I've been out of school for a few years now, I've realized that the memorized material from school fades unless you use it regularly. It boggles my mind that we still effectively test students on memorization ability rather than research and synthesis abilities, even in 'hard' sciences. Sure, there are some processes, procedures and standards which we expect professionals to just 'know', but that should not be the vast majority of knowledge tested.
Plus, in the 'real world', I'll happily hire an engineer who doesn't know how to solve an interview problem, but can adequately describe how they'd go about learning how to solve it. I'd trust that person not to just go with their gut or what they think they know, but to rather do some research.
Granted, mass cheating, copying answers, etc is still wrong. People should still have to 'do the work' so to speak.
According to some faculty, the U.S. military academies (or some of them) have rampant cheating (and other corruption) and little penalty. Look up The Cost of Loyalty by Tim Bakken, for example.
The best way to corrupt an institution is to idealize it.
Cheating all on your own is difficult if you also have to hide it from your peers.
I really think it was a great system. But, it's not perfect. I believe the military academies are all on similar honor systems, and I remember a few times of mass-cheating scandals, which have pretty heavy penalties, including expulsion in those environments.