preferring exchange rate changes to general deflation in deficit countries
and/or inflation in surplus countries. The classic case in defence of using
exchange rate adjustment was that of Milton Friedman (1953), who made
the analogy with changing to daylight saving time; it is easier to change
one price, the exchange rate, than to change' the prices of everything an
economy produces, just as it is easier for everyone's clocks to be set back
an hour than for everyone to change his or her schedule.
However, it is hard to credit the case that prices of goods, and especially
wage rates, are so flexible that the huge real exchange rate changes needed
to eliminate current external imbalances could have been accomplished
quickly through inflation and deflation. The problem is essentially one of
coordination within an economy. Although the discussion of this problem
is familiar in macroeconomics, it is perhaps less familiar in the international
context, and so will bear one more discussion.
Suppose that, as typical estimates suggest, to balance the U.S. current
account it was necessary that US wages fall 30 percent relative to foreign
wages from their 1985 peak. For an individual worker a 30 percent wage cut
is very drastic; one would imagine that bringing wages down by that
much would require a protracted and bitter struggle between employers
and employees. However, if all U.s. wages fall by 30 percent, the real wage
rate will fallmuch less, say, only 3 percent, since the bulk of U.S. consumption
is domestically produced. This means that if all wages could change
simultaneously, and each worker could know that other workers would
take the same wage cut, it might be possible to get such an adjustment
fairly quickly and painlessly. However, this would happen only in a world
of hyperrational agents, with no long-term contracts. In the real world
nominal wages never fall that much except in the face of a collapsing economy.
What a currency depreciation does is to solve the coordination problem
by lowering all domestic wages relative to foreign wages at the same time.
Figure 1.1 shows the behavior of U.S. unit labor cost relative to its competitors
and of the nominal dollar effective exchange rate, both as calculated by
the IMF. The figure surely shows that there is a prima facie case that
exchange rate changes do produce short-run changes in relative labor cost,
and thus can facilitate such a change when one is necessary. The figure also
shows the huge magnitude of the fall in U.S. relative wages that has already
occurred since the dollar's peak. If one believes that a relative wage change
of this magnitude was necessary, it is worth imagining what it would have
required to achieve this with a fixed exchange rate.
As years of debate in closed-economy macroeconomics have shown,
someone committed to the belief that prices are perfectly flexible cannot be
convinced of the existence of some inertia on the basis of evidence, since
evidence can always be rationalized away. However, for those less committed
the prospect of attempting to achieve large real exchange rate
movements without changing nominal exchange rates must surely look
unappealing.
Read More: How Useful Is Exchange Rate Adjustment?
