We consider a dynamic moral hazard economy inhabited by a planner and a population of privately
informed agents. We assume that the planner and the agents share the same discount factor, but that
the planner cannot commit. We show that optimal allocations in such settings solve the problems
of committed planners who discount the future less heavily than agents. Thus, we provide micro-
foundations for dynamic moral hazard models that assume a societal discount factor in excess of
the private one. We extend the analysis to allocations that are reconsideration-proof in the sense of
Kocherlakota (1996). We show that these allocations solve the choice problem of a committed planner
with a unit discount factor.