David Vines: How to make a new Keynesian DSGE model that is fit for purpose: a constructive response to the criticisms of Joe Stiglitz

Abstract:

In response to the criticisms by Joe Stiglitz of new Keynesian DSGE models I argue that a simplified version of the model that he criticizes can provide a framework for contemporary macroeconomic analysis. The simple extended new Keynesian or SENK model which I propose brings together the analysis of longer-term economic growth with the analysis of shorter-term economic fluctuations. It is a stripped-down version of a synthesis of the RBC model with capital accumulation (due to Kydland and Prescott) and the “science-of-monetary-policy” model (due to Clarida, Gali and Gertler) a model which can be found papers by Smets and Wouters and Christiano, Eichenbaum and Evans. This SENK model provides valuable intuitive insight, both in understanding discussions of contemporary economic policy, and in helping to understand the econometric DSGE models used in central banks. The new Keynesian macroeconomic research programme is itself expanding in in two complementary ways. Analytical developments include innovations which can be added as embellishments of the SENK model. These include risk premia facing firms, the existence of credit-constrained consumers, and the presence of real wage resistance. They also involve addition to the model of the kind of non-linearities needed to capture the existence of large and persistent economic downturns and booms and changes in economic structure that go beyond traditional business cycle analysis. Empirical extensions are also being made to the SENK model. The narrowly optimising microfoundations of the model are being gradually relaxed (using the insights of behavioural economics and complex systems analysis), thereby enabling a more detailed, and careful, empirical analysis of macroeconomic systems. In sum, the moves which I describe in this paper, based on the foundation provided by the SENK model, are, I believe, making new Keynesian DSGE models more nearly fit for purpose.

Keywords: DSGE, multiple equilibrium, toy models, structural economic models

JEL classification: A23, A31, B22, B41, E007