Notes on the Lucas Critique, Time Inconsistency, and Related Issues 1 Preliminary Remarks on Government Policy Rules

ht = g(It−1) , t = 1, 2, . . . . (1) For simplicity of exposition, only deterministic policy rules are considered in this section, that is, policy rules that map each possible information state I for each period t into a single policy choice vector g(I) for period t. An alternative stochastic form would be for each possible information state I for period t to be mapped into a probability distribution PI(h) over feasible policy choice vectors h for period t. In either case, however, notice that the presumption that a government uses a single policy rule g over times t = 1, 2, . . . means this government does not engage in learning; it already knows at time 1 what action(s) it will take in each possible information state in each period t ≥ 1.