RDP 9812: An Optimising Model for Monetary Policy Analysis: Can Habit Formation Help? 8. Conclusions

A model to be used for monetary policy analysis should be closely related to the underlying objectives of consumers and firms, should explicitly model expectations, and should capture the dynamic interactions among variables that are exhibited in the data. While many recently developed models explicitly model expectations, and purport to build close ties to underlying agents' objectives, most simple optimisation-based macroeconomic models fail to replicate important dynamic correlations in the data. A direct implication of these models' failure to replicate key dynamic correlations is that the models are unlikely to represent agents' dynamic behavioural decisions. As a result, such models are not suitable for monetary policy analysis. In many cases, the model's empirical failings are not widely understood, because the authors have not attempted rigorous empirical testing of the model.

This paper suggests a reasonably rigorous empirical standard for dynamic econometric models, and makes some progress towards a model that meets the standards itemised above. It does so by including a particular form of non-time-separability in the utility function, namely ‘habit formation’, or the assessment by consumers of utility relative to a habit level of consumption. The paper develops evidence that shows that augmenting the model in this way allows the model to replicate key dynamic correlations among consumption, output, interest rates, and inflation to a degree that standard models cannot. In particular, the model can match the hump-shaped response of consumption to income, interest rate, and inflation shocks. The habit formation specification improves upon the standard specification because it imparts a motive for consumers to smooth the change, as well as the level of consumption.

Other specifications may also afford improvements in the empirical performance of the standard model. This paper suggests, however, that only specifications that impose some smoothness on the change in consumption will be successful empirically. The gradual or hump-shaped response of consumption to shocks that is found in reduced-form and other empirical studies is a statistically significant feature of the data.

The specification set forth in this paper might not be robust across shifts in monetary or other policy regimes. But only through rigorous econometric testing of this and alternative specifications across regime shifts can observational equivalence (or empirical dominance) of alternative specifications and stability of any one specification across policy shifts be determined. I believe that this paper takes a small step in the direction of developing a rigorous standard of empirical validation for macroeconometric models, and in the implementation of that standard to provide a modest improvement in optimising models for monetary policy.