Where is the comparison to the employment ratio of the entire workforce? Without this it is difficult to look at subsets (like 16-19 yrs old, 55+) and draw much of a conclusion. I would guess that population/macro effects probably tell a lot of the story.