This paper compares inertial behavior on schedule and hierarchically decomposed behavior, and establishes some conditions under which the latter is superior to the former. Inertial behavior relates to an agent who, facing costs of adjustments, considers at any time the option of updating the entire vector of decisions or leaving it the same. Decomposed behavior is when the agent also considers at any time the option of updating only a subset of decisions. Of the different decomposition schemes available, this paper considers the hierarchical scheme, which reflects the classic distinction between short-run and long-run behavior in the theory of the firm. The comparison between inertial behavior on schedule and hierarchically decomposed behavior is discussed for a broad class of disturbances. The problem of optimally choosing which decision variable should be updated more frequently is also studied. It is found that the more frequently adjusted variable is not necessarily the one affected by lower costs of adjustment, despite contrary claims in the literature about the selection of short-ruin and long-run variables.