There is an embarrassing trend at the US EPA for example, to get interns and volunteers to do some official work, with no real way to transition to full time with job security and health insurance. This is partly due to the unending stream of applicants, despite few positions.
So from this context, rebuilding research means .. what? Both institutional knowledge store and authoritative titled positions. After some advances in data management by Google and some other commercial players, it is not clear what role Federal science had anyway, from this desk.