An appraisal of input-output models

Abstract

The aims of this thesis have been (a) to examine the theoretical framework of the input-output technique of economic analysis, and (b) to set out the results of a project which was concerned with the treatment of imports in input-output models. Accordingly, Part I contains a discussion of the basic Leontief model - the input-output model in its most simple form. The assumptions made in this method of analysis, and the essential properties and most important uses of the model, are examined. The various procedures involved in the operational use of the basic input-output model, and the methods of solution available, are dealt with in some detail. The simplifying assumptions made in the basic input-output model may constitute a serious departure from reality in certain circumstances, and the possibility of relaxing these simplifying assumptions is considered in Part II. It would of course be quite feasible to draw up a single theoretical model, which incorporated all the modifications and/or extensions of the simple model and, provided the necessary statistical data was available, to apply such a model operationally. However, such a complex model would be unsuitable for examining the different modifications independently. The procedure adopted has therefore been to introduce only one or two possible modifications at a time. The nature and effect of each of the various modifications and/or extensions of the basic input-output model are thus more clearly exposed to examination. Furthermore, it is most likely that only limited modification of the simple model would be necessary, in order to deal realistically with a particular problem, and to obtain results which had a reasonably high degree of accuracy. Parts I and II have drawn on the work of many economists . An effort has been made throughout to simplify the exposition of the various models as far as possible. In a number of cases, a clear examination required a more detailed treatment than had been undertaken in the available literature, as for example, the discussion of the treatment of imports. Much attention has been directed to the methods of finding solution values for each model, and, where it was felt to be appropriate, e.g. the solution of the linear programming models, rather different methods of solution have been adopted than those which had been proposed by the original authors. Some models have been modified, to a certain extent, to make them more realistic, and to overcome criticisms which could have been directed at them in their original form. Part III is devoted to the application of both a conventional method, and a refined method, of dealing with imports, to Australian data derived for 1946/47. The results obtained suggest that input-output studies, for economies such as Australia, should incorporate a more refined method of dealing with imports than the methods normally used. There is some doubt as to the reliability of results obtained in a conventional analysis. The application of each method of dealing with imports has been discussed in considerable detail, so as to provide a clear illustration of the computations necessary in the operational use of the input-output model.