I went to undergrad in the US, and there's a similar, but disguised, phenomenon. The difficult
majors (ie: faculties/degree paths) would have very high "drop-out rates", in terms of people changing tracks after the first year or two. So indeed, less than half the people who entered with a declared intention of completing a degree in Physics, Engineering, Computer Science, or Mathematics would actually do so.
Quite a few could be found in the social-sciences and humanities departments after they were "weeded out" of the hard sciences. Personally, I think this is a crappy way to run things, since it means social-science and humanities professors have to deal with lots of students who are only there for a generic degree they can manage to obtain rather than for the specific degree path.
On the other hand, I put scare-quotes on the "weeded out", because often the precise method of weeding out weaker students is to simply not even try to teach well in the first year or two of math, science, programming, and engineering courses. So for instance, my fiancee actually cleared all her differential and integral calculus requirements via Advanced Placement (exams high-schoolers can take that grant university credits), with perfect scores, but when she tried to take Linear Algebra in our undergrad math department, she found the teaching so bad that she struggled to understand anything at all. Despite this, she pulled an A- in the course and was one of the top three students in her class/section -- but insists she barely understood anything at all.
It really makes me wonder what kind of pedagogical wonders we'd see if university professors were actually trying to teach well in courses below the third year.