Last Friday, we welcomed Harvard economics PhD student and JFI fellow Kathryn Holston to the office to discuss the state of macroeconomic modeling. Kathryn described major macro models and their uses. “The trade-off between realism and tractability is one of the perpetual questions in macro.” To a question on forecasting, Kathryn referenced the classic Meese and Rogoff paper on how no model outperforms a random walk, a result which “people have been trying to disprove ever since.”
Kathryn reviewed the benefits and drawbacks of DSGE models and heterogeneous agent models. Read Kaplan, Moll, and Violante on HANK (Heterogeneous Agent New Keynesian) modeling here.
What’s been happening in macro modeling recently? Since the recession, “building models that have realistic frictions has been a big focus.”
Kathryn ended the presentation with a discussion about the difficulties of interpreting and communicating macro for social policy.