Right now, schools assess teachers before they teach--filtering for candidates who are certified, who have a master's degree, who have other pieces of paper that do not predict good teaching. And we pay them the same regardless of their effectiveness.
By comparison, if we wanted to have truly great teachers in our schools, we would assess them after their second year of teaching, when we could identify very strong and very weak performers, according to years of research.
The amount that's gained from teacher credential programs, measured against the prices paid for them (and there are more than just money), is scary. I can read a Time Article, 4 blogs with lengthy commentary and draft, redraft, and learn to format a blog, but I can't bring myself to work on my credential program work because I know it will be a waste of my time. Go figure.