ating under a regime of ¬‚exible exchange rates monetary expansion results in

an increase in income, with the effects of monetary expansion on aggregate

demand and income being reinforced by exchange rate depreciation. Further-

more, in the limiting case of perfect capital mobility monetary policy becomes

˜all-powerful™. In contrast, in Chapter 4, section 4.4.3 we considered how in

the monetary approach to exchange rate determination, where real income is

exogenously given at its natural level, monetary expansion leads to a depre-

ciation in the exchange rate and an increase in the domestic price level. In

what follows we outline the essence of Dornbusch™s (1976) sticky-price

rational expectations model in which monetary expansion causes the ex-

change rate to depreciate (with short-run overshooting) with no change in

real output.

In his model Dornbusch made a number of assumptions, the most impor-

tant of which are that:

1. goods markets are slow to adjust compared to asset markets and ex-

change rates; that is, goods prices are sticky;

2. movements in the exchange rate are consistent with rational expecta-

tions;

3. with perfect capital mobility, the domestic rate of interest of a small open

economy must equal the world interest rate (which is given exogenously),

plus the expected rate of depreciation of the domestic currency; that is,

expected exchange rate changes have to be compensated by the interest

rate differential between domestic and foreign assets; and

4. the demand for real money balances depends on the domestic interest

rate (determined where equilibrium occurs in the domestic money mar-

ket) and real income, which is ¬xed.

Given these assumptions, what effect will monetary expansion have on the

exchange rate? In the short run with ¬xed prices and a given level of real

income an increase in the (real) money supply results in a fall in the domestic

interest rate, thereby maintaining equilibrium in the domestic money market.

The fall in the domestic interest rate means that, with the foreign interest rate

¬xed exogenously (due to the small-country assumption), the domestic cur-

rency must be expected to appreciate. While short-run equilibrium requires

an expected appreciation of the domestic currency, long-run equilibrium

378 Modern macroeconomics

requires a depreciation of the exchange rate. In other words, since long-run

equilibrium requires a depreciation of the domestic currency (compared to its

initial level), the exchange rate depreciates too far (that is, in the short run it

overshoots), so that it can be expected to appreciate back to its long-run

equilibrium level. Such short-run exchange rate overshooting is fully consist-

ent with rational expectations because the exchange rate follows the path it is

expected to follow.

A number of points are worth noting with respect to the above analysis.

First, the source of exchange rate overshooting in the Dornbusch model lies

in goods prices being relatively sticky in the short run. In other words, the

crucial assumption made in the model is that asset markets and exchange

rates adjust more quickly than do goods markets. Second, the rate at which

the exchange rate adjusts back to its long-run equilibrium level depends on

the speed at which the price level adjusts to the increase in the money stock.

Finally, in the long run, monetary expansion results in an equi-proportionate

increase in prices and depreciation in the exchange rate.

7.7 Real Rigidities

One important criticism of the menu cost literature noted by Ball et al. (1988)

is that models with nominal frictions can in theory produce large nominal

rigidities but ˜do so for implausible parameter values™. However, Ball and

Romer (1990) demonstrated that substantial nominal rigidities can result

from a combination of real rigidities and small frictions to nominal adjust-

ment. Indeed, Mankiw and Romer (1991) identify the interaction between

nominal and real imperfections as ˜a distinguishing feature of the new

Keynesian economies™.

If all nominal prices in an economy were completely and instantaneously

¬‚exible, a purely nominal shock would leave the real equilibrium of an

economy unchanged. As Ball and Romer (1990) note, ˜Real rigidity does

not imply nominal rigidity: without an independent source of nominal

stickiness prices adjust fully to nominal shocks regardless of the extent of

real rigidities.™ However, rigidity of real prices and wages will magnify the

non-neutralities which result from small nominal frictions. The importance

of this point can be seen by considering the impact of a decline in the

money supply. Suppose initially that the presence of menu costs deters

¬rms from reducing their prices in response to this nominal disturbance.

With the price level unchanged real output will decline. Each mono-

polistically competitive ¬rm will ¬nd that its demand curve has shifted to

the left. Because each ¬rm is producing less output, the effective demand

for labour declines (see Abel and Bernanke, 2001). If labour supply is

relatively inelastic, the shift of labour demand implied by the decline in

The new Keynesian school 379

output will cause a large fall in real wages; that is, the nominal wage rate

declines to bring this about (see Ball et al., 1988; Gordon, 1990; D. Romer,

1993). This decline in the real wage rate implies a decline in marginal cost,

a decline which will be strongly reinforced if the marginal product of

labour rises sharply as the labour input decreases. As is evident from Figure

7.2, an upward-sloping marginal cost curve would greatly increase the

incentive to reduce price and would ˜swamp any plausible barriers to nomi-

nal adjustment™ unless the elasticity of demand at the existing price falls as

the ¬rm™s demand curve shifts to the left. The greater the decline in the

elasticity of demand at the existing price as output falls, the more the

marginal revenue curve facing a ¬rm shifts to the left and the less incentive

there is for a ¬rm to reduce its price.

David Romer (1993) sums up the essence of this issue as follows: ˜Thus if

the classical dichotomy is to fail, it must be that marginal cost does not fall

sharply in response to a demand-driven output contraction, or that marginal

revenue does fall sharply, or some combination of the two.™ Real price rigidity

is high the greater is the cyclical sensitivity of the elasticity of demand and

the smaller is the cyclical sensitivity of marginal cost. Hence nominal shocks

have large real consequences the greater the degree of real rigidity (see D.

Romer, 2001).

The points discussed above can be more easily understood by referring to

the familiar mark-up pricing equation facing a pro¬t-maximizing monopo-

listically competitive ¬rm (see Pindyck and Rubinfeld, 1998, p. 340). Pro¬t

maximization requires that the ¬rm produces that level of output where

marginal revenue (MR) equals cost (MC). Marginal revenue can be expressed

in the form shown by equation (7.5):

MR = P + P(1/·) (7.5)

where P is the ¬rm™s price and · is the price elasticity of demand. Pro¬t

maximization therefore requires that:

P + P(1/·) = MC (7.6)

By rearranging equation (7.6) we get equation (7.7):

P ’ MC

= ’1/· (7.7)

P

This equation can also be rearranged so as to express price as a mark-up on

marginal cost. The mark-up equation is given by (7.8):

380 Modern macroeconomics

1

P = MC (7.8)

1 + 1/·

Since marginal cost is the nominal wage (W) divided by the marginal product

of labour (MPL), we ¬nally get equation (7.9):

W« 1

P= ¬ · (7.9)

MPL 1 + 1/·

The term inside the brackets represents the mark-up, the size of which varies

inversely with the elasticity of demand (remember · is negative). Equation

(7.9) indicates that P will not fall when MC declines if the mark-up rises

suf¬ciently to offset this decline (see Stiglitz, 1984). If the elasticity of

demand does not decline, then equation (7.9) also indicates that the incentive

to change price will be small in the presence of menu costs if MPL does not

rise strongly as the labour input is reduced (see Hall, 1991). Rotemberg and

Woodford (1991) suggest that desired mark-ups over marginal cost fall dur-

ing a boom because it becomes increasingly dif¬cult to maintain oligopolistic

collusion; that is, industries become more competitive in periods of high

economic activity. During recessions implicit collusion increases, leading to

a countercyclical mark-up that acts as a real rigidity, magnifying the impact