Work Backward Rather Than Forward
Many people immerse themselves in library research, take copious notes, and then
attempt to organize these notes into an outline. Only at the end do they think about
the point they are trying to make. Instead, before you do any research, prepare a
“back-of-the-envelope” analysis. Rough out a tentative framing of the issue, a set of
decision criteria, and a sketch of options for addressing the problem. Rely initially on
your “ordinary knowledge” (Lindblom and Cohen, 1979), conventional wisdom,
common sense, and basic facts. Clarify what you need to find out, and discuss the
initial cut with the client or a colleague before you invest heavily in research.
Sketch a full first-draft policy memo as early as you possibly can, and review this
draft with your client if possible. As you continue the research, revise your findings
and fill in missing information.