RDP 1977-07: Money and the Balance of Payments 6. Comparison of U.K. and Australian Model Results

The major differences between the Australian and U.K. model results, however, require some comment. The first is that the effects of excess money on domestic variables are stronger in the RBA76 model, and the effects on the balance of payments correspondingly weaker. Related to this is the result that the average lags with which asset markets (in particular, the markets for money and net foreign assets) adjust toward equilibrium are substantially shorter in the RBA76 model. Thus, while the greater structural detail in the RBA76 model may permit the better identification of the various channels through which monetary disequilibrium affects the behaviour of the economy, the balance-of-payments effects appear weaker (and the domestic effects stronger) than in the smaller U.K. model.

The way in which the Australian model appears to be relatively closed may reflect the emphasis on shorter term adjustments in the specification of the domestic sector of the RBA76 model, but a further point is that the estimation period for the U.K. model may be dominated by the experience of the gold standard and the post-war dollar standard,[32] while in the Australian model the most variable part of the sample period is also one of comparative exchange rate flexibility, when domestic monetary disturbances would be expected to be “bottled up” more than under a regime of less variable exchange rates.[33]

The greater structural detail in the RBA76 model is also a possible cause of the second major difference in the results: the ability to discriminate more clearly in the Australian case between the alternative monetary sector specifications. This is particularly relevant to the comparison of Models D and E; in the U.K. model case, the theoretically more satisfying Model D cannot be unambiguously distinguished on statistical grounds from the “quasi-reduced form” of Model E, suggesting that the U.K. model does not capture fully the interaction of money and capital markets, because of its smallness and simplicity. The structural detail which may be necessary for the identification of these linkages is absent from the U.K. model, while it may be adequately represented in the richer structure of the Australian model. Alternatively, the institutional and structural differences between the two economies may be such that the theoretical representation of capital flows employed in this study, although adequate for the Australian case, does not provide a good description of the working of capital markets in the U.K. over the period considered.[34]

In concluding this comparison of the U.K. and Australian model results, it should be remembered that this paper aims only to test alternative specifications of the monetary sector, within a consistent framework, for each of two sets of available data. A considerable amount of further research on the structure of the core models and the properties of the data sets would be required before any direct comparison between the two sets of results is possible.


It might be recalled here that the U.K. data period ends in 1970. [32]

It is interesting, in this context, that attempts to extend the U.K. model sample period beyond 1970 have encountered several difficulties, particularly in the influence of world prices on domestic prices and the reactions in general of the domestic economy to changes in its external position. [33]

The discussions in, for example, Sayers (1976), illustrate the volatility of the London capital market, suggesting that exogenous factors have played a major role in the determination of short-term British capital flows. [34]