The myriad miseries of graduate school are reserved to no one discipline, but there may be something to the contention that biology graduate programs are particularly bad. Here's what Mike the Mad Biologist says, in response to Science Professor, and I think he's quite right:
The basic problem stems (so to speak) from too many biology Ph.D.s and not enough funding, leading to an immensely cutthroat environment--and one that is psychologically damaging to boot. . . .So why does this dysfunctional cultural paradigm exist? I think it has to do with two things: specialization and Ph.D. training. When you go to Ph.D. school in biology, especially biomedical sciences, you learn a great many difficult techniques requiring lots of skill--it's not for dummies at all. The problem is that most of the skills you learn are only useful in . . . the biomedical sciences. Most don't learn enough 'generalist' skills, such as high level math or serious programming skills, to have other career alternatives if academia doesn't work out. Worse, many of the skills they learn become obsolete. A decade ago, sequencing was a Ph.D. activity, or at least, an activity supervised very closely by a Ph.D. Now, it's largely automated, and the machines are mostly run by technicians with bachelors degrees.
This is something I've tried to explain many times to nonscientists: most of the esoteric techniques I mastered during my thesis aren't useful outside a Drosophila lab. They're not transferable to any other field of biology, let alone any other scientific or nonscientific profession. Those skills I picked up on my own - speaking, writing, teaching, how to think about problems and dig into the literature unaided, how to handle severe setbacks, find ways to motivate myself - those are all transferable to virtually any career. But you have to teach those to yourself.
A friend of mine recently complained to me about a postdoc who wasn't living up to her promise. The postdoc would spend far too much time on simple activities, doing them to death, while making no progress at all on her new research project; how, my friend wondered, did the postdoc ever get a PhD in the first place?
I think that postdoc is probably terrified. Once you learn a lot of nontransferable skills, you don't know how to go outside the box. You want to work hard and prove yourself, but you can only perform the few procedures in your comfort zone; going outside that zone just reminds you how little you know, and how no one has taught you to learn new skills. The biology PhD is not readily self-renewing - it's more like a craft apprenticeship where you learn a specific, often homegrown and idiosyncratic set of skills from your mentors. That those skills are highly idiosyncratic is demonstrated when a lab relocates: it can take six months to get experiments working, cells growing, or model organisms breeding again - even in the same people's hands with the same equipment and same techniques. That's just scary, and we take it for granted in biology.
I've seen graduate students focus on perfecting their nontransferable skills to the exclusion of reading the literature or thinking about their thesis, frantically cranking out data without having a full grasp of the context of their work. In a big lab, that's okay; you're collaborating with senior postdocs, cranking out data makes you feel useful, and your productivity ensures the goodwill of others. But when you're the postdoc, you can't do that. You have to know how to leave the box, and if you never practiced that, you don't know how.
When I talk to biology graduate students about getting into science policy, the most important message convey is to acquire transferable skills. You will have to create opportunities for yourself to do so - volunteering, blogging, teaching, etc. - but unless you seek out those opportunities, you won't have either the skills or the confidence to make a meaningful contribution away from your model organism and apparatus of choice. Within your research, also, you should focus as much as you can on learning broadly applicable skills - programming, statistics, how to handle large datasets. Later on, your obscure skills will impress nonscientists, but it's your transferable skills that will make them want to hire you, and help them envision what you could do for their organization.
The problem with this advice, of course, is that it's often in tension with what your PI or committee want. They (often) don't want you teaching on the side when you could be in the lab. They don't want you learning programming, if you have a postdoc who does all of the lab's programming already. They don't want you reinventing the wheel or duplicating effort; they want you to publish within your allotted niche. That's why Mike's post rings so true: acquiring transferable skills is only a good investment if everyone accepts that graduate students are not going to follow their PIs into academia in exactly the same field using the same techniques. Until the "traditional model" of academia dies, graduate schools will be turning out students prepared to compete for a handful of academic jobs, and unprepared to do anything else. That's just not right.