The application of linear programming to problems of national economic policy in New Zealand

Abstract

This thesis is concerned with the development of a quantitative model which has the potential to be used by a central authority as a guide to planning economic growth. An indefatigable faith that the market forces of supply and demand best serve the objectives of society (whatever they are) does not prevent the diseconomies and waste which often occur due to imperfect knowledge of the vagaries of prices, trade cycles, market trends and so on. An overall planning agency which has both knowledge and control of the relevant variables will often be able to foresee and prevent such losses. This study is part of the Lincoln interindustry research programme. Using the same basic interindustry data the task will be to determine the maximum amount of consumption which can be achieved in the target year. Rather than choosing a vector of consumption targets and using the interindustry relationships to compute a structure which will be the basis for economic policy, the aim is to use linear programming to calculate the structure which will give the greatest level of consumption that the availability of resources will allow. Not only will this method serve as a useful comparison with earlier work, but the linear programming technique gives greater flexibility to the model builder than the traditional Leontief input-output system. In particular the Leontief assumption that each sector produces one, distinct, homogeneous commodity can be relaxed. Thus it is possible to have each of two activities producing the same commodity so that the system can choose between alternative sources of supply. It is also possible to make allowance for diminishing returns by specifying maximum levels for activities so that more of the, same commodity can be produced only by an activity which has a larger requirement of scarce inputs. The scope of this study will be restricted mainly to maximising consumption so that the linear programming solution can be compared with the earlier projections, and there will be only a limited exposition of the ways in which choice can be introduced into an interindustry model. However, it is hoped that the way will be made clear for the formulation of more adventurous programming models of the New Zealand economy.