Conclusion

New classical models of aggregate fluctuations imply that aggregate fluctuations are mainly caused by real factors. This is why such models are often called RBC models.

New classical models are DSGE models based on optimizing behavior by both households and firms, flexible prices, and fully competitive markets.

If the competitive general equilibrium models of this kind could explain all the features of aggregate fluctuations, then there would be no need for models that stress distortions in product, financial, and labor markets, or other market inefficiencies. However, new classical models have weaknesses as models of aggregate fluctuations.

First, these models cannot fully account for the real effects of nominal and monetary shocks. For example, it is widely accepted, following the evidence presented by Friedman and Schwartz [1963], that the Great Depression of the 1930s was caused by monetary and not real shocks. Similar views are prevalent regarding the Great Recession of 2008–2009, which was one of the deepest post-World War II recessions.

The imperfect information assumption of Lucas [1972] can be used to account for the real effects of nominal shocks, but the effects of nominal shocks in such a model would be short lived and nonpersistent. However, the more recent models of sticky information, based on persistent informational frictions and rational inattention, are more successful in accounting for persistent effects of monetary shocks. So are the financial frictions models, such as the Bernanke [1983] model.

Second, even though new classical models can in principle account for employment fluctuations, they do so only on the basis of intertemporal substitution in labor supply. However, this explanation is not sufficient to explain the existence and the persistence of unemployment and the widely held view that unemployment is an involuntary condition for those who experience it, and not the result of a voluntary rational choice under imperfect information.

For these reasons—and even though new classical models are internally consistent—many macroeconomists consider them as an incomplete and not persuasive explanation of aggregate fluctuations. The alternative class of models consists of a variety of new Keynesian models. These models are based on product and labor market distortions that hinder nominal wages and prices from adjusting sufficiently in the short run to equilibrate labor and product markets. Thus, following both nominal and real shocks, output and employment deviate from their equilibrium values, resulting in aggregate fluctuations and involuntary unemployment. It is to such models that we now turn, starting with the traditional Keynesian models of aggregate fluctuations.

1. This model was implemented empirically and tested by Lucas [1973], Sargent [1973, 1976], and Barro [1977, 1978], among others.

2. The Frisch elasticity of labor supply is defined as the elasticity of labor supply when the marginal utility of wealth is held constant. See Frisch [1932, 1959].

3. Equation (14.24) could be derived using the money in the utility function approach, by allowing for real money balances to affect the utility of the representative household. See chapter 12. However, we postulate it directly for reasons of analytical simplicity.

4. We have seen how the Wicksell rule results in price level determinacy in the models of chapter 12. Here we use a version of the Wicksell rule for inflation rather than for the price level.

5. This was noted immediately after World War I by Fisher [1919], who argued for a policy of complete stabilization of the price level. The modern equivalent of the Fisher rule is the complete stabilization of inflation at π^*, the optimal inflation rate of the central bank.

6. Such a relation is often called a “Phillips curve” or a “Phillips relation” and by the 1960s, had become a major component of Keynesian models. We shall discuss this relation and its properties in chapter 15.

7. These assumptions are those of Alogoskoufis [1983] and are in the spirit of Lucas [1972].

8. We could easily assume that productivity and the money supply follow a more general stationary stochastic process, such as a linear AR(n) process. However, this would not alter our results with regard to the nonneutrality of monetary shocks, as it is only unanticipated changes in the money supply that affect output and employment in this model. Thus, the systematic part of the money supply process would be incorporated in the expectations of households and would not affect labor supply and other real variables.

9. This results from the famous confusion between absolute and relative price changes in the Lucas [1972] model.

10. See the exchange between Rees [1970] and Lucas and Rapping [1972] on the ability of the Lucas and Rapping [1969] equilibrium labor market model, based on intertemporal substitution in labor supply, to explain unemployment in the Great Depression. Lucas and Rapping [1972] concede that the model has problems explaining the persistence of the rise in unemployment. See also Bernanke [1983] for an econometric investigation of the Great Depression using a model that relies on unanticipated monetary shocks. He finds that such monetary shocks cannot explain the persistence of the Great Depression, and that additional propagation mechanisms are required to explain persistence. He thus proposes a propagation mechanism that relies on the role of financial frictions in the form of a sustained increase of the costs of financial intermediation following the extensive bank failures caused in the first years of the depression. We introduce and discuss financial frictions in chapter 19.