Comment

“Monetary Policy Analysis When Planning Horizons Are Finite” byMichael Woodford fits in a fast-growing literature that attempts to introduce forms of bounded rationality in macroeconomic models. Bounded rationality can be introduced in a variety ofways, depending on howwe describe the agents’ limited ability to process information, to form forecasts, and to compute optimal plans. The paper I am discussing captures bounded rationality by giving agents a finite planning horizon and exploring in depth a variety of consequences of this modeling assumption. The paper provides a nice motivation for the exercise by connecting the macro literature to existing work in artificial intelligence. In this discussion I want tomake two points, one on the role of general equilibrium effects and one on difference between finite lives and finite planning horizons. There is one dimension of bounded rationality that appears in different forms in a variety of models: the limited capacity of agents to think through general equilibrium effects in their environment. My first point is that this limited capacity for general equilibrium thinking also plays an important role in this paper. To make this point, let me use a simple example of the “forward guidance puzzle” (Del Negro, Giannoni, and Patterson 2012), inspired by Farhi and Werning (2017). Take an infinitely lived consumer, with standard time-separable preferences, who receives a deterministic stream of labor income {Yt} and has access to a single bond that pays the real interest rate rt. The optimal behavior of this consumer can be derived from the Euler equation