This Appendix advances a model that captures the key features of Giblin’s analysis in Australia, 1930. In particular it captures the hybrid character of the analysis: containing both a ‘Keynesian’ emphasis on the benefits of demand, and a ‘Classical’ emphasis on the benefits of wage cuts.

The modelling assumes the existence two factors of production- labour and land – and three goods: food, manufactures and services. Food uses only land in its production and services uses only labour. All manufactures are imported, all food is exported, and services are neither imported nor exported. The money price of manufactures is normalised at 1. The money price of food in Australian currency is P, and determined by the exogenous exchange rate and world price of food in sterling. The money wage is a given, V. The production of food, F, is given by the fixed supply of land.

National income is the sum of the value of services and food.

Y = Z + FP

As there is zero saving, national income in Australian pounds, Y, is entirely spent. As all spending is on services, Z, or manufactures M, we may write,

Y = Z + M

The two equalities imply

M = FP

But outlays on manufactures (equal imports) are proportional to income,

M = m Y

Thus,

This last equality captures the leading ‘Keynesian’ thesis of Australia 1930: national income is a multiple of the value of production of the exportable, FP.^[36] A decline in FP will produce a decline in national income several times as large. A decline could arise from (i) a fall in the volume of food production, F, (ii) a fall in P due to fall in the pound sterling price of food, and (iii) a fall in P due to the appreciation of the Australian pound. Thus this equality captures the vulnerability of Australian income to world terms of trade, as well as to the nominal exchange rate.

But certain ‘Classical’ features of the analysis are also apparent. As all employment is located in the services sector, which uses no other factor than labour we may write:

L V = Z

And since Z = (1-m) Y we may infer,

This amounts to a negative relation between employment, L, and the real wage in terms of food, V/P, as hypothesised by ‘Classical’ economics. Thus the impact of a fall in P may be read in terms of it increasing the real food wage, V/P and thereby reducing employment. The restoration of employment will require a reduction in real wages, that will be secured by a reduction in the nominal wage, V

It may be objected that V/P is not a measure of ‘the real wage’ when, by our present assumptions, no food is consumed. But the consumption of food can be allowed for without attenuating the negative relation between employment and the real wage. Suppose that c of national income is spent on food. Then

Y = Z + M + cY

Given Y = Z + FP we may infer,

and so,

The negative relation between employment and the real wage remains.

Two further points about this last expression:

(i) The size of the multiplier contracts as c rises.^[37] Thus c operates like the propensity to save in the ordinary Keynesian model, even if the mechanism is different. c does not represent an increase in ‘leakages’ as income rises; it represents a decrease in ‘injections’ as income rises. For c is the amount by which exports are reduced by an extra unit of income. This implies that if c + m = 1 then all of an increase in income goes to either reduce exports or increase imports, and there is no multiplier at all, as Giblin appears to have completely appreciated.^[38] The upshot is that the existence of a multiplier requires that at least part of an extra pound of income be spent on the one thing that is neither importable nor exportable: services.

(ii) If m = 0 – the closed economy - then the model reduces to the de Lissa model: national income equals ‘production’ divided by the propensity to spend on production.