There's also a big problem in building up the mathematics inventory. At my undergrad institution we have two main streams of economics studies, the standard one, and the mathematical one (which is explicitly said to be gearing toward to graduate studies). However, even the mathematical one does not require the hardest maths (and does not even require third year Analysis, but only recommended, and did not even recommended the harder, math major version).
Therefore, if a standard undergraduate comes in and (unknowingly) take cheap calculus for the first two years, there is no way she will be ready to take Real Analysis by third or fourth year.
So our profession is counting on the natural math wiz kids (who somehow don't go into math/compsci/physics) or only the kids surfing the web for "econ phd" even before starting college (me). Is that really what we want it to be?