Mankiw Model of Nominal Rigidities
There are two related reasons for which firms do not frequently change prices. First, as we saw in the discussion on menu costs, the costs of price changes are not negligible and could exceed the private benefits that can be obtained by the firms in the form of increased profits. More importantly, however, the benefits to be reckoned from price changes are not so much in the private realm, but in the social realm. Price rigidity leads to unemployment, the social costs of which are much higher than the private costs reckoned by the firms in terms of menu costs. The microeconomics -based models by Mankiw, and Akerlof and Yellen clearly show that the private benefits of changing a price can be much smaller than the social benefits if there is substantial monopoly power in the economy.
We follow Dornbusch, Fischer and Startz (2004) in presenting a simplified version of the formal Mankiw model. The model relies on the fact that in monopolistic markets firms face a downward-sloping (less than infinitely elastic) demand curve and can set a price that deviates from the optimum profit- maximising price without a large swing in demand away from the firm. This is not possible in a perfectly competitive market, where every firm faces a horizontal (infinitely elastic) demand curve - a small deviation from the optimal price can in this situation lead to a large swing demand and profits away from the firm. Even if a competitive firm faces the same kind of menu costs as an imperfectly competitive firm, the loss of profits by not changing the price can be big enough to outweigh the menu costs. A competitive firm is not, of course, a price setter. Not so for the imperfectly competitive firm. Mankiw shewed that the potential profits from raising prices could be very small for such firms especially if the elasticity of demand for firm's output is low, i.e., if monopoly power of the firm is high, and if the deviation of the actual price from the optimal profit- maximising price is small. The menu cost could well be higher than the potential profits in such a case and the firm does not change its price. Other firms are likely to be similar and they too leave their prices unchanged, with the result that the nominal price level remains unchanged. The Keynesian conclusion of an increase in money supply on output rather than on prices follows from this. An increase in money supply, prices remaining unchanged, leads to an increase in real money supply. This leads to an increase in aggregate output, either through a decrease in the rate of interest (a la Keynes) or through a real-balance effect. You should note that there would have been no output effects in a classical model if prices were free to vary. An important difference between the classicals and the Keynesians is hereby established.
As Dornbusch, Fischer and Startz (2004, p. 566) put it about the papers of Mankiw, and Akerlof and Yellen:
This work provides a rigorous microeconomic justification for nominal price stickiness. Since New Classical economists attack the rigcr of the underpinnings of Keynesian models, such justification is a key piece of Keynesian response to rational expectations and real business cycle models. Not everyone agrees on the empirical significance of the formulation by Mankiw and by Akerlof and Yellen, but the work is certainly a mile stone in the New Keynesian counterrevolution.