[Market determination of the value in exchange (price) of money is considered in a general equilibrium finite horizon model. The possibility of the price of money being zero in equilibrium and the role of taxes (payable in money) in preventing a zero price are considered.]

1. INTIRODUCTION ANALYSES OF THE STABILITY of general competitive equilibrium2 have typically concerned themselves with the case of pure exchange. Sometimes that restriction has been inessential. In the tatonnement case, for example, if we are willing to make the assumption that all goods are gross substitutes, then it makes little difference where the excess demand functions come from.3 In less highly special cases, the restriction to pure exchange is a more serious matter. This is particularly so in the richer world of non-tatonnement processes. The difficulties raised by the abandonment of the pure exchange assumption in such models are three. First, such models, in assumptions or proofs, rest on considerations of utility maximization which are not immediately applicable to firms. Second, it is not fully natural to assume that stocks of goods are bounded, an important feature when non-tatonnement is being considered. Third, the creation and destruction of goods in the course of the adjustment process itself must be taken into account. The present paper shows how the first and, to some extent, the second of these problems can be handled in the context of the most appealing of the non-tatonnement models, that of the Hahn process.4 The third problem, that of the creation and disappearance of goods during the adjustment process is far more difficult. It is also not special to the introduction of firms, since the pure exchange models do not allow consumption to take place until equilibrium has been reached. Thus, this

[This paper presents a model of an economy in which the formation of equilibrium is completely explained by the independent optimizing behavior of individual agents, and thus reformulates the concept of equilibrium in a way which is essentially related to that of stability.]

[This paper contains a relatively simple proof for the stability of Edgeworthian recontracting. It also applies that proof to a random recontracting process in an economy with a finite number of utility vectors, in which the recontracting process is a Markov chain and all non-core utility vectors correspond to transient states.]

[This paper investigates the working of a multifunction tatonnement stability process in an actual economy by simulating the adjustment path of an abstract disequilibrium model in conjunction with an econometric equilibrium model, indicating that such a path is highly stable.]

[All over the world economists are busy evaluating major road proposals and other transport projects. But this work is largely confined to rural areas because the methods used are inappropriate for evaluating big transport schemes in towns, where traffic congestion is a dominating consideration. This paper discusses congestion as an economic problem of demand and supply, expressed as simple functions of the cost of travel, in time and money, to the road user. Road expansion in congested cities often seems to achieve nothing but more congestion. This paper demonstrates how this arises as a process of market equilibrium, and how one can assess the benefits of the road expansion in this situation. The paper then considers how the evaluation of road schemes would be affected if direct road pricing were introduced into cities as a means of controlling congestion.]

Simultaneity as a LimitA well-known and highly convincing position on the nature of simultaneity in econometric models is that such models are only approximations to the true state of affairs.In fact, causation takes time, and the reactions given by the equations of the model truly occur not instantaneously but with a very small time lag.Unfortunately, however, data do not come to us sufficiently finely divided in time to allow us to observe such fastmoving reactions, so we take simultaneous instantaneously-holding relations as approximations, valid between the observations which nature allows us.Time lags are thus considered negligible provided they are sufficiently small.The consequences of this position for parameter estimation when observations occur at discrete points of time separated by an interval much larger than that in which the true reactions take place have been discussed 2 in the literature.In fact, however, this particular variant of the above view does not seem a very realistic one, We very seldom have observations on the value of a particular variable at precise discrete moments in time, 3 and, if we do, we seldom use the observations in that form.Much more common is the case in which the observations either by necessity or by choice are in the form of averages or sums over a non-zero time interval.Simultaneous This view has been discussed at length by Bentzel and Hansen [2].The basic position on causation has been vigorously maintained by H. Wold in several works, e.g., Wold and Jureen [11]. 2 Strotz [10]; but see also Gorman [6].