There is definitely a problem with an oversupply of PhDs relative to the job market for physics (and likely biology). For a position at say Berkeley for a biology faculty position there used to be approx. 600 applicants per position. For physics at first tier or second tier institutions the number may drop to 200. Even if we are cruel and suggest that half of those are unqualified, that still leaves a large pool of extraordinarily qualified people competing for a rather small pool of jobs. I see this regularly when there are young postdocs with good publication records (Nature, PRL, etc.) who are having trouble finding permanent positions after their postdocs. Part of this may be related to decreased state funding and hiring freezes (in several states, there have been furloughs). Even for postdocs who have decided that they would prefer to work at an undergraduate institution and teach, the competition is fierce. Oddly, even for those that want to teach at a public high school, it's hard because of the education requirements (you can run a facility, teach freshmen at an elite college--but teaching high school seniors....). Things are so fierce that it's rather hard to have much selectivity about geography. This can wreak havoc with relationships and in physics is known as the two body problem--where a couple in science has difficulty finding positions in the same zipcode. As one colleague told me, she'd be happy to just have the same timezone....
For my subfield, industrial research positions have been gradually drying up (at least for doing physics rather than engineering). A number of companies in the past were able to use monopoly profits to drive research (think of AT&T Bell Labs which is now but a shadow of it's former self--when I was there as an intern, it was amazing....). However, many have scaled back. Thus, I have seen a number of people pursuing various exit strategies.
During the internet boom (where I had decided to drop computer engineering as a major because physics was more fun), a number of people who could code dropped out an joined startups. Later, people from Ivy institutions joined consulting firms such as McKinsey (with a "mini-MBA"). Later, a number joined in the gold rush of financial engineering. While that continues, many go through a brief masters first to get their foot in the door. A few turn to more engineering related work. So, while the unemployment rate for physics PhDs is low--not so many are actually still doing physics research.
For myself, I'll take on undergraduate and high school interns. No graduate students. I really respect String Theorists who for years intentionally limited the number of students they would accept due to the paucity of permanent positions. For years, I'd been reluctant to take on a postdoc due to the current situation. Now, I've taken on my first postdoc and will do my best by him--but I have to be honest about the job market and I'm having him learn some programming as a plan B. Plan C is that I'm very confident that he'll be able to get a position in his home country afterwards.
I've seen some people who are bitter (think of the opportunity costs!) when they leave. But, I've seen some who are mellow--"At least I got to work with something beautiful for awhile....".Part of the difficulty is that for scientists, you don't go into it for the money (at least I hope you don't!), you go into it for love. So, doing science becomes not just a job, but rather a calling and a way of life. So, someone's sense of self may often become tied to being a scientist--and that's hard to leave behind...
So to summarize, while all fields of science are not cutthroat, given the level of competition, it is very hard to find a job. Also, given the level, then people have to work extremely hard and it takes a toll on people's personal lives (it's hard to have one when average work weeks extend to 60-80 hrs for a number of experimentalists--my solution has been to sleep less, but I'm told that's unhealthy...).