EMBARQUES DE EXPORTACIÓN
CLAVES 311-312ALIMENTOS Y ALIMENTOS FORRAJEROS
It is useful to compare the monetary model and the Mundell–Fleming model directly. The differences can be considered under the following three headings: price level; income; expectations and interest rates.
6.7.1
Price level
In the monetary model, since the aggregate supply curve is vertical at all times, the price level moves with perfect flexibility to clear both money and goods markets. Then PPP ensures that the real exchange rate is preserved at its equilibrium level.
In the M–F model, the price level is simply an exogenously fixed index that can have no role to play in the domestic macroeconomy. Neither can it play any part in determining equilibrium in the open sector.
6.7.2
Income
The corollary of these contrasting assumptions about the supply side of the economy is two totally different views of the part played by income in the model.
In the monetary model, with full employment and perfectly functioning factor markets, changes in real income can only be exogenous events. Furthermore, con- sumption is implicitly taken to depend on interest rates rather than income. It follows that the only role left for income to play is in helping to determine the demand for real balances. As we saw, because exogenous increases in income swell the demand for money, they are associated with currency appreciation.
In contrast, income is one of the three endogenous variables in the M–F system, the others being the domestic interest rate and the exchange rate (if floating), other- wise the balance of payments. An (endogenous) increase in income is associated with three types of effect in this context. First, insofar as it raises the demand for money, it leads to a rise in the interest rate, other things being equal. Second, by feeding back on to consumption (the Keynesian multiplier effect) it boosts demand for goods and services. Third, and more particularly, it is associated with a worsening current account via the marginal propensity to import and, ceteris paribus, either reserve losses or currency depreciation.
6.7.3
Expectations and interest rates
Neither model provides an explicit role for expectations, although both have pro- vided the framework for more complicated models that fully rectify this omission, as we shall see later on in the book.7
Nonetheless, the flavour of the monetary model is either that expectations are exogenously given or that they depend in some more complicated fashion on the other variables (see Chapter 13). Furthermore, in the spirit of the classical tradition to which it belongs, the monetary approach assumes the real interest rate is deter- mined purely in the savings market.8When combined with interest rate parity, these
two facts imply that nominal interest rates cannot fluctuate freely to clear the money markets. That job is left to the price level. The domestic interest rate is effectively tied to (although usually not equal to) the foreign rate.
The Keynesian approach takes goods market equilibrium as dependent at least as much on income as on interest rates. The corollary is that, in a domestic context, the interest rate is free to help clear both money and goods markets. Viewed from the 184 Chapter 6 · Fixed prices: the Mundell–Fleming model
6.7 The monetary model and the Mundell–Fleming model compared 185
international perspective, the link with foreign interest rates is almost completely broken. In place of UIRP, we have something like:
r= r* + ∆se
t+ ρ (6.5)
Now if the required risk premium (the last term) rises with the quantity of domestic currency assets held, it follows that the scale of the capital inflow at any point will be an increasing function of the domestic interest rate, for a given value
of the expected rate of depreciation (the second term on the right-hand side).
This is all very well, provided, first, that the expected rate of depreciation can actually be taken as given in the face of significant macroeconomic policy shifts and, second, that the required risk premium is determined independently of the other vari- ables. On this last point, we shall have more to say in Chapter 14.
As far as expectations are concerned, it seems highly unlikely that the policy meas- ures considered in this chapter would leave expectations unchanged.9 Once it is
allowed that expectations are very likely to depend on the government’s policy stance, the logic behind the M–F conclusions begins to unravel.
Stocks and flows
The differing treatments given to interest rate determination in the two models are very much related to their contrasting approaches to defining equilibrium.
Following the classical tradition, equilibrium in the monetary model, in its float- ing rate version at least, is a steady state in the fullest sense: the stocks of both money and goods are willingly held, with no tendency for net flows in either direction. Once the economy has adjusted to a monetary expansion, for example, there is no reason why the new equilibrium should not persist forever, if left undisturbed.
Contrast this with the M–F analysis: the response to money supply expansion involves a depreciation-induced surplus on current account, which then finances the continuing net export of capital, in response to the new, lower level of domestic interest rates.
Now this cannot be an equilibrium in the full, static sense. A net capital flow into or out of a country is prima facie evidence that disequilibrium holdings are in the process of being adjusted and it is quite plausible that adjustment may be protracted. In fact, the international capital markets are never at rest.10But, however long the
process may take, at some point it will be complete. Stocks of assets will have fully adjusted to the disturbance, with no further tendency for capital to move into or out of the country – and this is the situation one would expect to find in full equilibrium.
And this is not purely a theoretical nicety. If, for example, the capital inflow following on fiscal expansion under a floating exchange rate is only temporary, the current account deficit that it finances can only be temporary. As the influx of funds dries up, the exchange rate must depreciate, other things being equal, thereby reduc- ing the current account deficit, pushing interest rates up even further and reducing income. In the limit, there would be no change in income or the interest rate. There would simply be an appreciation of sufficient scale to crowd out the additional spending.11
As has already been mentioned in the introduction, the M–F analysis was highly influential precisely because of its conclusions about the relative efficacy of monetary
and fiscal policy under fixed and floating rates. No comparable analysis is possible in the context of the monetary model, simply because it has nothing to say about the effect of changes in government spending. Why not?
One way of understanding this feature of the classical model is to go back to the goods market identity of flows of injections and leakages. Rewriting Equation 4.3, we have:
Savings≡ I + G + B (6.6)
When net government spending expands, there is a ceteris paribus rise in the volume of injections on the right-hand side of this identity. How is the equality maintained? In the Keynesian view, the equality is preserved by some combination of a rise in income, which increases savings, and an increase in interest rates, which also raises savings somewhat but which, more importantly, deters investment (‘crowding out’). The M–F analysis follows through the open economy implications of this mechanism. By contrast, the classical view12 is that an increase in the government’s budget
deficit will bring forth the requisite additional saving spontaneously. The reason is simply that economic agents realize the extra spending will have to be paid for by future taxation, even if it is financed by borrowing at the moment. They will there- fore wish to step up their current saving at any given interest rate so as to be able to pay the taxes when they fall due.
In fact, it can be shown that, provided taxation does not actually distort the sav- ings decision,13responsible, non-myopic agents will be happy to buy the additional
government debt issued to finance its extra spending at the same price as before the
change took place. They will do so because the return they earn on the debt, properly
measured,14will be just sufficient to cover the additional tax burden.
If we can rely on this mechanism of Ricardian equivalence, (as it is called, after the great classical economist, David Ricardo) we need not distinguish between spending by the private and public sectors. The government is simply an agency created by a democratic state to act on behalf of its citizens – in borrowing as in everything else. The fact that a private citizen prefers to buy now and pay later does not automat- ically raise interest rates. Why should it do so when he or she channels the whole process (spending and borrowing) through the agency of the government?15
On these and similar grounds, one would expect an increase in G on the right-hand side of Equation 6.6 to be automatically associated with the increase in savings needed to finance it. In other words, the IS curve need not shift as a result of what is simply a change in the share of spending between public and private sectors.
In practice, there are certain to be a number of breaks in the circuit leading from the nation’s budget to household budgets.
In the first place, households may simply be myopically unconcerned about the future tax burden being imposed on them. The facts are not easily accessible: government spending changes are made public only in the opaque form of official statistics, nationalized industries’ accounts, white papers and so on. In any case, it is one thing to absorb the fact that £1 of extra government spending today will have to be paid for at some point in the future by additional taxes whose present value is £1.16
It is quite another to accept that I or even my children, if I have any, will have to pay those taxes. By the time the chickens come home to roost: I may no longer be inside the tax net: I may be dead, retired, emigrated or unemployed. Moreover, unless my 186 Chapter 6 · Fixed prices: the Mundell–Fleming model
6.8 Evidence 187
coupon payments are exempt from income tax, buying government securities will yield less than I need to pay the tax bill, when it falls due.
For these reasons, the most likely outcome for an economy like that of the UK or USA would seem to be some degree of crowding out greater than zero (as predicted by Ricardians) but smaller than is suggested in Keynesian models like the one analysed in this chapter.
The question is ultimately an empirical one and one that, in spite of its import- ance, has yet to be satisfactorily answered.
6.8
Evidence
It is not of very great interest to test the M–F model in the form presented here, at least as far as the floating rate era is concerned – it seems pointless to try to explain the facts by starting from the assumption that prices are fixed. On the other hand, introducing price flexibility in one way or another changes the model quite drasti- cally, as we shall see in Chapter 7.
Nonetheless, certain episodes in the last 20 years have prompted interpretation by some economists in terms of the M–F mechanism.
In particular, the simultaneous rise of both the dollar and US real interest rates to record levels in the first half of the 1980s, led many commentators to point to the growing federal deficit as the likely culprit.
Unfortunately, whatever may be true in the criminal courts, in economics circum- stantial evidence rarely points in a single direction. In the present case, real interest rates were at or around their peak levels throughout the industrialized world, in spite of the conservative budgetary stance of some of the countries involved. Moreover, the dollar’s subsequent fall took place well in advance of any tangible sign that the US deficit could even be stabilized, let alone reduced.
Of course, the dollar’s fall, when it finally took place, was immediately rational- ized by some commentators as a market reaction to . . . the Federal deficit.
It is easy to be facetious about these ad hoc explanations. However, it should be realized they could both be correct and attempting to reconcile these two apparently contradictory positions takes us back to one of the central weaknesses of the M–F model.
The point has already been touched on in discussing the issue of capital mobility in the section on stocks and flows. Suppose that the USA was able to fund additional federal spending by borrowing from abroad (primarily Japan) in the early 1980s. In return, it offered a (very slightly) higher return. However, if the Japanese had fully adjusted their holdings of dollar securities by the mid-1980s, they would then stop exporting capital to the USA unless the risk premium rose still further, a develop- ment that failed to materialize, for some reason or other. Hence, the dollar had to fall. International money managers could hardly ignore the fact that where they had initially (in 1982) been lending to the world’s largest creditor nation, by 1987, they were being asked to lend to the world’s largest debtor – with no end in sight to the deterioration in the USA’s external finances.
This retrospective rationalization (it certainly does not merit being called an explanation) illustrates the shortcomings of the M–F concentration on flows rather