RDP 1978-01: Two Essays in Monetary Economics 4. A Two-Sector Model: The Shifting Phillips Curve

Figure 3 illustrates the standard two-sector model with traded and non-traded goods in full equilibrium. Monetary disequilibrium is introduced into this model by Jonson and Kierzkowski (1975) who otherwise retain the standard assumptions about costless resource transfers and instantaneous adjustment of prices and quantities.

The standard assumptions about perfect markets can be relaxed fruitfully by recognizing that it may be rational for suppliers to delay their response to a shift in the pattern of demand, especially if the shift in demand is regarded as temporary. This can be formalized in the standard model by making the supply curves functions of past as well as current prices, an assumption that is consistent with the notions discussed above that economic agents make decisions on the basis of expected values of prices which may be sticky in the short run. It is also consistent with the usual notions that adjustment – particularly quantity adjustment – is costly. The introduction of this innovation implies that the short run transformation curve is below that which applies in the long run, and, further, is different in every period following a once for all change in the pattern of demand, although tending towards the long run transformation curve, with the exact pattern of movements in the transformation curve depending on the precise specification of the supply functions. This idea is probably quite unoriginal^[12] but its implications for the trade-off debate are worth pointing out. In the usual trade theory framework it is a minor qualification suggesting that there are transitory unemployment effects, as everyone knew anyway. It becomes more interesting when one considers such effects in a model with money, and a government sector which imposes its preferences for (say) a higher proportion of non-traded goods by printing money. Consider the initial effects of the shift in preferences illustrated in Figure 4 (which also illustrates a short run transformation curve corresponding to supply functions which depend on past as well as present prices). The rise in demand for non-traded goods raises the price of non-traded goods, and hence the general price level (since traded goods prices are assumed to be fixed in world markets). The rise in the price level induces a fall in the real value of money balances, thus reducing (“crowding out”) private demands for both traded and non-traded goods and making room for the government consumption of non-traded goods.

In the first period solution illustrated in Figure 5, production of traded goods has declined more than consumption, producing a balance of trade deficit, the monetary effects of which tend to be offset by the increase in domestic credit to finance the government consumption of non-traded goods. The assumption that the private sector uses its money balances as a buffer stock, however means that the net monetary impact will only be felt in the next period

In the next period the change in private demands will depend on the net monetary movement: if the government budget deficit exceeds the trade deficit, private demands for both traded and non-traded goods will tend to increase and, depending in part on the supply response, there will tend to be a greater trade deficit, leading to a subsequent reduction in the money stock; if the initial budget deficit was less than the trade deficit, private demand will fall still further, tending to reduce the trade deficit. If the usual stability conditions are met (and the supply response complicates this issue) the model has a solution in which the balance of payments deficit equals the budget deficit and in which output has returned to a full employment level, on the long run transformation curve (figure 6). This cannot be a steady state solution, however, because the continuing balance of payments deficit means that the country will eventually run out of international reserves. At some stage in the process outlined above the fall in reserves will mean the government will be forced to devalue the exchange rate or, more interestingly for the current analysis, reduce its spending. The latter option will set in train the opposite adjustment path, with further underemployment of resources until the economy eventually settles down to the original equilibrium at point A.^[13]

The result of this process will be higher rates of inflation and less output for the entire period that the economy is in disequilibrium than would be the case if stable government policies were followed. This conclusion – which appears likely to be highly relevant for understanding the deteriorating inflation-employment performance in many countries in recent years, when macroeconomic policy appears to have become more variable – is of course in marked contrast to that of the standard Keynesian model in which an increase in government spending automatically increases output. The contrast remains when this model includes a Phillips curve – even of the expectation augmented variety – and allows eventual crowding out of private expenditure through a tendency for tax receipts to rise more than proportionately to nominal income (although such mechanisms can be included in more complex and realistic versions of the above two-sector model). The two-sector model seems to throw further light upon the results obtained by Lucas (1973), who also argues that countries with relatively variable government policies have a less favourable trade-off than countries with more stable policies. The present model highlights the point that variable demand management policies imply resource shifts which reduce supply during the adjustment period; in contrast to Lucas, the model incorporates demand factors and provides an explicit structural view of the problem.^[14] Although similar results will obtain in larger and more complicated models, with alternative ways for governments to finance resource shifts in their favour, the simple one-asset model is likely to be highly relevant in a world in which it appears cheaper, especially in the relatively short run which dominates the horizon of most governments, to obtain resources by printing money. Yet there are at least two further reasons for focusing on “money” rather than “assets in general”.

Footnotes

I know of an unpublished paper by H.I. Kierzkowski which discusses time dependent production functions. [12]

The former option, devaluation, can lead to an alternative steady state in which government spending is financed by taxing money balances by a sustained inflation, and the exchange rate continuously depreciates so that traded goods remain competitive on world markets, to the extent that domestic demand for money is not reduced by the rising prices. Even in this case partial reversals of the tendency to draw resources to the government sector may induce temporary unemployment effects. [13]

The Lucas model could be incorporated in the present one by the specification of appropriate supply functions. [14]