When I was young I studied physics. The initial idea of string theory is kind of cute. It kind of answers the question what the quantum version is of a rubber band. But that does not really work out as a fundamental theory because it has to live in 24 dimensions, contains tachyons, and there are no fermions. To fix this people invented the fermionic string. Already at that point this seems a somewhat artificial thing to do. The thing is that ever since the question was posed it is still quite unclear how to bring general relativity and quantum field theory together. String theory was interesting for so long, I think, because it complicated enough to contains lots of interesting mathematics but not complicated enough to be completely intractable.
String theory had great marketing and so was able to convince politicians they would get answers in ten years if they just got money, which in turn meant those marketers had jobs. I call it marketing because regardless of the truth, a lot of money went into it for years with no useful results. Maybe a few more years of string theory research will discover it is right and makes some useful predictions - but so far it hasn't and most people outside have given up on it. (and it appears most in physics today expect it will never amount to anything - but since I'm not deep into physics that might be wrong)
None of the people involved in scientific funding decisions are politicians in the sense of being elected or caring primarily about public opinion. The pejorative you're looking for is "bureaucrat." :-) Although in most cases the ground-level decisions are made by peers (which means "professors.")
Meh, I am not buying this. String theory is not very expensive. Picture a guy writing down formulas on a large piece of paper. How much does that cost? Not very much, really. And, at the same time, you also get teaching out of the same person. Experimental high energy physics is orders of magnitude more expensive.
