【正文】 t_tonnement process need not exist. Although the picture is extremely simple and highly abstract, it does, I believe, reflect the essential feature of how our economy operates.V. Price and Capacity in Multiplier Process I have presented my main idea that the market exchange process can be viewed as a multiplier process as opposed to a t_tonnement process. Economists generally believe that the multiplier framework, including the IS/LM and the multiplier accelerated models, can work well only under the assumptions of fixed prices and excess supplies. For example, Barro (1994) in his recent critique on the textbook aggregatesupply and aggregatedemand model remarked that The AD curve reflects the underlying IS/LM model, and the key to this model is the presence of excess supply of goods and service. The excess supply reflects, in turn, the assumed stickiness of the price at excessive level. (pp. 3)One should first note that in the multiplier process, supply is always determined by demand (through contracts, orders and inventory adjustment) and therefore there is little sense of excess supply. However does this mean that there is some sense of excess capacity? Specifically, for the multiplier process to work well, are fixedprice and excess capacity required? To answer this question, we first need to consider the price theory associated with the multiplier process. One can find that the multiplier process that I present above is purely a process of quantity determination. It is not related to prices. Implicitly, prices are set by agents and shown publicly. Agents know their related prices when they trade. For instance, when a producer negotiates with his customers, he knows his current input prices from which he can calculate the base price of his own output. Under this existing and publicly known price structure, the multiplier process takes place as a quantity adjustment. I thus suggest that the major price theory associated with the multiplier principle is the Ricardian (or classical) costdetermined price theory, which is exactly independent of the quantity determination.. This price theory was formalized by Sraffa (1960) thirty years ago. However this need not mean that prices are fixed, or that the capacity constraint effectively exists that limits the working of the multiplier process. In this section, I will first defend the classical price theory against Hicks39。 critique. I then address some possible doubts arising from the incorporation of classical price theory into multiplier analysis. The issue related to excess capacity is also addressed at the same time. Indeed we will find the two issues are closely related. Hicks39。 Critique. Hicks (1972) listed the following problems to classical price theory. First, It is not a very stringent condition, if it is unacpanied by any rule about profits being normal。 and the normalization of profits (equalization between different sectors) is a plicated process, for which it is difficult to give sufficient time during the lapse of a single period. (pp. 78) It might be a problem if we assume producers take prices parametrically. However this leads us back to traditional equilibrium analysis. In the multiplier analysis, we automatically assume that firms are all price makers. They set the price at which they can earn a rate of profit determined by custom. If this rate of profit is known in advance, as well as the other relative input prices, the producer can immediately calculate his own price according to the Sraffa equation and shows it publicly. Having done this, he needs only to stay in his own market waiting for possible customers. Of course, it is the petitive pressure that forces the producer to set his output price at the level of rate of profit determined by custom. Otherwise no customer will e.. Of course, if the producer gets some monopoly power (including secret technology), he might set his price above the rate of profit determined by custom.Second, consider if a traded modity is nonstorable Marshall39。s fish case. When price is rigid, the advantage of beginning in this way is, however, less obvious. ... The Fixprice assumption is more awkward in the case of nonstorable than in that of storable modities. (Ibid., pp. 79) Almost all nonstorable goods are not manufactured goods, to which I believe the application of classical price theory is most appropriate. We certainly cannot imagine that the fish price can be settled at a uniform rate of profit. Yet since consumers are not assumed to e simultaneously, at least for a while (from the beginning of the market day), we find the fish price fixed. The price begins to drop off only near the closing of the market day. In the next market day, however, the actual beginning price may still be the same as yesterday39。s beginning price. Third, consider the manufactured goods, which are often storable. If, when demand exceeds output, there are stocks that can be thrown into fill the gap, it is obvious that the price does not have to rise。 ... If there are no stocks to take the strain, it is harder to stick to the assumption of rigid price. (Ibid., pp. 79) Manufactured goods, especially industry goods, are often supplied according to orders. The demanders who visit this kind of market are often there to express their desires and thus to make contracts (orders). They do not necessarily buy the output immediately. Therefore there is no sense that demand exceeds output, though in some sense, demand may exceed capacity. Other manufactured goods, especially those with many indeterminate demanders, are often supplied according to inventories. However since consumers are not supposed to e simultaneously, there is time for producers to adjust their supply according to their inventories. Even in the case of inventories having run out temporarily, what we often find is: Would you mind ing here tomorrow? Again in this case, the only sensible problem is still the capac