bullion mentions CS and that prompts me to a line of thought based upon a Quant headhunter's grasp of financial economics, a disclaimer in itself...
When I give career advice a common theme is that "easy, riskless" ways to make money are inherently unstable, enough that "riskless" should be treated in the high frequency trading sense, if you can grab the cash now then do so with my blessing, but don't treat it as a long term source of income.
I'm rather hoping that this is not in any way an controversial idea in this sort of forum.
Law was for many years on some sort of efficient frontier, indeed my wife who has formal qualifications in economics from Oxford, but who also scored a role in IBM's research labs at 18 is now a scary lawyer who has been regularly using the word "bubble" for some years in terms of fees and incomes.
The tech industry can be modelled really quite well as a series of bubbles that overlap and the sort of work that MFEs do although it is labelled "banking" not only highly technology based but shows many of the same characteristics with product types coming and going at a frequency that makes the idea of "career stability" almost naively optimistic.
But the energy industry is like that as well, and my friends in pharma look really worried and their concerns are from different causes than other lines of work.
In short, I don't think there is anything special, good or bad about law and finance both of which are showing signs of the sort of mean reversion you expect in markets where infinite demand cannot happen and increasingly supply saturates. What we're seeing in law and to an extent finance is a second order effect where the attractive risk/return attracted enough people that it helped destabilise itself. If you've done reaction kinetics then you can see in your head what I'm seeing in mine and it's worth remembering that the Le Chatelier's principle you did in market modelling actually came from physical chemistry.
So what about CS ?
My first remark is that many CS qualifications are really shit, shallow and limited in career antifrgaility and if you ain't read a preprint of NNTs book, you need to do that now, but in the mean time here is a link to my view of the mid term (3 year) market for Java skills :
http://www.theregister.co.uk/2011/10/04/programming_skills_java_bubble/
Although CS skills are increasing in supply (especially Java) the supply seems to be increasing at least as fast and now that the barriers are so low, in terms of the hardware and s/w costs (low and zero) we can expect supply to catch up and overtake demand.
So I don't see a safe harbour except medical school and in the US medical costs show every sign of a bubble of which I am aware, that will end badly, though it's beyond my wit to work out the exact mechanism.
So in the midst of dark pessimism my advice is to find out where you are most ahead of other people, where skills are valued and where you can see a way to keep those skills current with the market. Skills like stats or SQL that apply to multiple industries and technologies offer better area under the earnings curve.
