Stopping Inflations, Big and Small
用一个包含价格调整固定成本的模型，解释为何温和通胀的结束会伴随产出损失而恶性通胀可快速消除，对研究反通胀政策的经济学者有参考价值。
This paper attempts to reconcile Gordon and Sargent's disparate observations using a single theoretical model of inflation and disinflation. The model's key feature, a fixed cost of price adjustment, makes firms unwilling to immediately reset their nominal prices following a small change in inflation. Most of the disinflationary episodes studied by Gordon began from moderate rates of inflation, and indeed, the end of a small inflation in the model presented here is accompanied by losses in aggregate output, although these losses are minimized when the disinflation is gradual. At the same time, however, firms incur the fixed cost to adjust their prices following a large change in inflation; big inflations, like those studied by Sargent, can be eliminated quickly with no loss in aggregate output. Parts of this story appear elsewhere. Ball, Mankiw, and Romer (1988), for instance, develop a model of costly price adjustment in which monetary shocks have larger effects on output at lower rates of inflation; the real effects of money tend to vanish as inflation rises. Because of the technical difficulties associated with solving models featuring fixed costs of price adjustment, however, previous studies have been unable to consider the full effects of large changes in policy, such as those required to implement a disinflation. Danziger (1988), for example, addresses the problem of disinflation in a model of costly price adjustment, but confines his analysis to cases in which a small inflation is brought immediately to an end. He therefore stops short of answering the questions considered here: how does the cost of disinflation depend on the initial inflation rate, and how does it depend on the speed of disinflation? Other studies of disinflation, including Phelps (1979), Tay...