Purpose of the Thesis
The purpose of this thesis is to provide an econometric analysis of the staple theory of economic growth (the term used here for staple theory in its most recent form) in the contexts of the Victorian and New South Wales economies of the period 1861-1900, which is the period immediately before Federation, when the Australian colonies were effectively separate countries. The theory is intended specifically for regions such as Australia in which the factor land (in the form of natural resources) is abundant relative to labour and capital, such that there is a comparative advantage in staples, or natural-resource-intensive commodities. As has only very recently been stressed, in this theory, staple exports are essentially a leading sector in the sense defined by Professor Simon Kuznets, contributing to economic growth both directly and indirectly, per medium of spread effects. The theory is dynamic in that it is theoretically possible for the spread effects to continue long after some export stimulus has occurred and (provided that investment opportunities generated are exploited) predicts that economic growth will be a process of diversification about an export base; additional inflows of the scarce factors and the evolution of the social structure form an integral part of this process.
The analysis centres on annual dynamic linear-least squares regression equations which are directed at important economic relationships encompassed by the staple theory's framework. Considerable emphasis is given to the central relationship between export growth and Gross Domestic Product (both aggregate and per capita) using recently published G.D.P. estimates as well as various income indicators which seem statistically more reliable. Moreover, the relationships between export growth and important individual sectors of the economies, especially the manufacturing, but also the transport and financial sectors, are also analyzed. Ideally there would have been relationships between export growth and inflows of the scarce factors but, due to the nature of the available annual data (being largely the products of interpolations from quinquennial and decennial estimates), these variables are included in the theoretical framework of the analysis indirectly, which seems the most meaningful way of dealing with them in annual statistical models. During the course of the econometric analysis, attempts are made to investigate whether the aggregate impact of the export staples on the economy varied according to commodity type and, in the case of New South Wales, also regional origin.
A number of conclusions seem to follow from this study:
(1) The results, when seen against the characteristics of Victorian and New South Wales economic development, would seem strongly to support the view that staple exports contributed significantly to the economic growths of the colonies (both directly and indirectly) and that they were leading sectors in the Kuznets' sense throughout 1861-1900.
(2) Differences in the lag structures of the equations of New South Wales and Victoria would seem to highlight the theoretical points that the nature of natural resource endowments plays a vital role in shaping growth performance and also that younger regions (perhaps through less inhibiting traditions) may experience greater spread effects.
(3) The results would also seem to support the view that, theoretically, spread effects can continue long after some export stimulus has occurred and that these longer-term influences can distort and sometimes completely over-ride shorter-term relationships between export growth and economic growth.
(4) The results seem to emphasize the need for a regional approach to studies of Australian economic development of the period 1861-1900, especially in view of the finding of possible regional differences within New South Wales.
(5) The study also suggests that proxy variables can successfully be used to overcome problems created by a lack of data in econometric testing of historical problems. The various equations seem generally to have yielded consistent results.
(6) From a technical point of view, this study also shows that tests and allowances for parameter differences between subperiods should form an integral part of econometric investigations into leading sectors in the Kuznets sense.
(7) The analysis also seems to suggest that measurement error problems in historical data, caused by improvements in the efficiency of the statistical collection mechanism, can be compensated in least squares regression analysis.