Elements of International Economics-Springer-Verlag Berlin Heidelberg (2004)
Elements of International Economics-Springer-Verlag Berlin Heidelberg (2004)
Elements of International Economics-Springer-Verlag Berlin Heidelberg (2004)
Elements
of International
Economics
With 41 Figures
and 9 Tables
i Springer
Professor Dr. Giancarlo Gandolfo
University of Rome La Sapienza
Faculty of Economics
Via del Castro Laurenziano 9
00161 Roma, Italy
[email protected]
Modern economies become more and more open and the external sector of an
economy becomes more and more important. This textbook aims at clarify-
ing how an open economy functions, in particular at explaining the determi-
nants of international fiows of commodities and financial assets. It also aims
at examining the effects of these fiows on the domestic and international econ-
omy and the possible policy acti.ons at the national and international level.
Particular attention will be paid to the problems of international economic
integration at both the commercial and monetary level.
Students will be able to read and interpret the balance of payments of
a country, evaluating the various types of balance, to explain the behaviour
of commercial fiows in the light of the theories studied, to analyze fiows of
financial assets according to interest-rate differentials and other elements, to
study the forces that determine exchange rates and cause currency crises,
to understand the reasons behind international economic integration such
as the European Union, to evaluate the effects of national and international
policies.
A peculiarity of this textbook is that it tries to bridge the gap between un-
dergraduate and graduate texts in international economics without being too
bulky. Drawing on my two graduate texts, International Trade Theory and
Policy (Springer Verlag, 1998) and International Finance and Open-Economy
Macroeconomics (Springer Verlag, 2002), I have written a concise textbook,
where the treatment is at a level suitable for undergraduate courses without
sacrificing the topics treated in graduate textbooks (for example, an elemen-
tary introduction to the modern approach to international macroeconomics,
namely the intertemporal approach, is given).
I am grateful to Marianna Belloc, Andrea Bubula, Giuseppe De Arcange-
lis, Daniela Federici, Alberto Felettigh, Michael D. Goldberg, Lelio Iapadre,
Manuela Nenna, Francesca Sanna Randaccio for useful comments on earlier
drafts. The boxed inserts that now and then appear in the text have been
mainly prepared by Daniela Federici and partly by Manuela Nenna.
The University of Rome "La Sapienza" provided the ideal environment for
the development and testing of the material contained in this book: I have,
in fact, used it both for courses given in English to undergraduate students
coming from all parts of Europe under the Socrates/Erasmus programme,
VII
VIII
1 Introduction 1
1.1 International Economics as a DistinctSubject 1
1.2 Structure of the Book . . . . . . . . . . . . . . 3
1.2.1 International Finance . . . . . . . . . . 3
1.2.2 The Theory and Policy of International Trade 4
1.2.3 Small and Large Open Economies . . . . . . . 5
I The Basics 7
2 The Foreign Exchange Market 9
2.1 Introduction: The Exchange Rate 9
2.1.1 The Real Exchange Rate. 11
2.1.2 The Effective Exchange Rate 13
2.2 The Spot Exchange Market .. 14
2.3 The Forward Exchange Market .. . 17
2.3.1 Introduction . . . . . . . . . . 17
2.3.2 Various Covering Alternatives; Forward Premium and
Discount . . . . . . . . . . . . . . . . . . . 19
2.4 The Transactors in the Foreign Exchange Market 22
2.4.1 Speculators . . . . . . 23
2.4.2 Non-Speculators.... 24
2.4.3 Monetary Authorities . 24
2.5 Currency Derivatives 24
2.5.1 Futures . . . . . . 25
2.5.2 Options . . . . . . 26
2.5.3 Swap Transactions 27
2.6 Eurodollars and Xeno-Currencies 29
2.7 Selected Further Reading . . . . . 30
IX
X Contents
Index 323
Introduction
1
2 Chapter 1. Introduction
The descriptive part, as the name clearly shows, is concerned with the
description of international economic transactions just as they happen and
of the institutional context in which they take place: flows of goods and
financial assets, international agreements, international organizations like the
International Monetary FUnd, the World 'Irade Organization, the European
Union, and so forth.
The theoretical part tries to go beyond the phenomena to seek general
principles and logical frameworks which can serve as a guide to the under-
standing of actual events (so as, possibly, to influence them through policy
interventions). Like any economic theory, it uses for this purpose abstractions
and models, often expressed in mathematical form. The theoretical part can
be further divided, as we said above, into trade and monetary theory each
containing aspects of both positive and normative economicsj although these
aspects are strictly intertwined in our discipline, they are usually presented
separately for didactic convenience.
A few words are now in order on the distinction between international
trade theory and international finance.
International finance (also called international monetary economics) is
often identified with open-economy macroeconomics or international macroe-
conomics because it deals with the monetary and macroeconomic relations
between countries. Although there are nuances in the meaning of the various
labels, we shall ignore them and take it that our field deals with the problems
deriving from balance-of-payments disequilibria in a monetary economy, and
in particular with the automatie adjustment mechanisms and the adjust-
ment policies concerning the balance of paymentsj with the relationships
between the balance of payments and other macroeconomic variablesj with
the various exchange-rate regimes and exchange-rate determinationj with
international financial markets and the problems of the international mon-
etary systems such as currency crises, debt problems, international policy
coordinationj with international monetary integration such as the European
Monetary Union.
The theory 0/ international trade (which has an essentially microeco-
nomic nature) deals with the causes, the structure and the volume of in-
ternational trade (that is, which goods are exported, which are imported,
and why, by each country, and what is their amount)j with the gains from
international trade and how these gains are distributedj with the determina-
tion of the relative prices of goods in the world economyj with international
specializationj with the effects of tariffs, quotas and other impediments to
tradej with the effects of international trade on the domestic structure of
production and consumptionj with the effects of domestic economic growth
on international trade and vice versaj and so on. The distinctive feature of
the theory of international trade is the assumption that trade takes place in
the form of barter (or that money, if present, is only a veil having no influ-
ence on the underlying real variables but serving only as a reference unit,
1.2. Structure of the Book 3
9
10 Chapter 2. The Foreign Exchange Market
exchange rate as the number of units of foreign currency per unit of domestic
currency and is, obviously, the reciprocal of the previous one (again taking
the USA as the horne country, we would have 0.61690 British pounds per US
dollar, 115.20737 Japanese yens per US dollar, 0.93545 euros per US dollar,
etc.); with this definition the exchange rate is the price of domestic currency
in terms of foreign currency.
We shall adopt the price quotation system, and a good piece of advice
to the reader is always to ascertain which definition is being (explicitly or
implicitly) adopted, to avoid confusion. The same concept, in fact, will be ex-
pressed in opposite ways according to the definition used. Let us consider for
example the concept of "depreciation of currency x" (an equivalent expres-
sion from the point of view of country x is an "exchange rate depreciation").
This means that currency x is worth less in terms of foreign currency, namely
that a greater amount of currency x is required to buy one unit of foreign
currency, and, conversely, that a lower amount of foreign currency is required
to buy one unit of currency x. Therefore the concept of depreciation of cur-
rency x is expressed as an increase in its exchange rate if we use the price
quotation system (to continue with the example of the US dollar, we have,
say, $1.65 instead of $1.621 per British pound, etc.) and as a decrease in
its exchange rate, if we use the volume quotation system (0.60606 instead of
0.61690 British pounds per US dollar, etc.). By the same token, expressions
such as "a fall in the exchange rate" or "currency x is falling" are ambiguous
if the definition used is not specified. Similar observations hold as regards
an exchange rate appreciation (currency x is worth more in terms of foreign
currency).
It should now be pointed out that, as there are various monetary instru-
ments which can be used to exchange two currencies, the respective exchange
rate may be different: the exchange rate for cash, for example, may be dif-
ferent from that for cheques and from that for bank transfer orders.
These differences depend on various elements, such as the costs of trans-
ferring funds, the carrying costs in a broad sense (if bank keeps foreign cur-
rency in the form of banknotes in its vaults rather than in the form of demand
deposits with a foreign bank, it not only loses interest but also has to bear
custody costs). These differences are however very slight, so that henceforth
we shall argue as if there were only one exchange rate for each foreign cur-
rency, thus also neglecting the bid-offer spread, that is, the spread which
exists at the same moment and for the same monetary instrument, between
the buying and selling price of the same currency in the foreign exchange
market.
To conclude this introductory section, we must explain another difference:
that between the spot and the forward exchange rate. The former is that
applied to the exchange of two currencies on the spot, that is, for immediate
delivery. In practice the currencies do not materially pass from hand to hand
except in certain cases, such as the exchange of banknotes; what usually
2.1. Introduction: The Exchange Rate 11
unit, or the nominal exchange rate adjusted for relative prices between the
countries under consideration:
(2.1)
or
(2.2)
where Ph, PI are the domestic and foreign price levels in the respective curren-
eies. Prom the economic point of view, it is easy to see that in (2.1) domestic
and foreign prices have been made homogeneous by expressing the latter in
domestic currency before taking their ratio, whilst in (2.2) they have been
made homogeneous by expressing the former in foreign currencyj the ratio is
of course the same.
According to another definition, the real exchange rate is the (domestic)
relative price of tradable and nontradable goods,
pT
rR = NT' (2.3)
P
The rationale of this definition is that, in a two-sector (tradables-nontradables)
model, the balance of trade depends on PT/PNT because this relative price
measures the opportunity cost of domestically producing tradable goods, and
the balance of trade depends on the excess supply of tradables.
A widely held opinion is that the real exchange rate ShOlUd give a measure
of the external competitiveness of a country's goods (if non-traded goods are
also present, only tradables should be considered), but even if we so restrict
the definition, it is by no means obvious which index should be taken. In the
simple exportables-importables model of trade the real exchange rate reduces
to the notion of terms 0/ trade defined in the theory of international trade,
namely
Px
rR = 1f = - ,
rpm
(2.4)
1
-Po;
7f=~ (2.5)
Pm'
since the two formulae are mathematically equivalent.
The terms of trade 7f can serve both the domestic and the foreign con-
sumer (country) for the relevant price-comparison, because in (2.4) the prices
of domestic and foreign goods are expressed in domestic currency, while in
(2.5) they are expressed in foreign currency.
It is dear that, since exports are part of domestic output, Po; = Ph, and
similarly Pm = Pj, so that (2.4) and (2.1) coincide.
Another definition of real exchange rate uses the ratio of unit labour costs
at home (Wh) to unit labour costs abroad (Wj ), expressed in a common
monetary unit through the nominal exchange rate (r),
(2.6)
effective exchange rate. Let us begin with the nominal effective exchange
rate, which is given by the formula
n n
rei = L Wjrji, L Wj = 1, (2.7)
j=l,#i j=l,#i
where
r ei = nominal effective exchange rate of currency i,
rji = nominal exchange rate of currency i with respect to currency j,
Wj = weight given to currency j in the construction of the index; by
definition, the sum of the weights equals one.
Usually the effective exchange rate is given as an index number with a
base of 100 and presented in such a way that an increase (decrease) in it
means an appreciation (depreciation) of the currency under consideration
with respect to the other currencies as a whole. This implies that rji is
defined using the volume quotation system.
Unfortunately it is not possible to determine the weights unambiguously:
this is an ambiguity inherent in the very concept of index number. Many
effective exchange rates thus exist in theory; usually, however, the weights
are related to the share of the foreign trade of country i with country j in
the total foreign trade of country i. Effective exchange rates are computed
and published by the IMF, central banks and private institutions.
If we carry out the same operation defined by Eq. (2.7) using real rather
than nominal bilateral exchange rates we shall of course obtain areal effec-
tive exchange rate. This will give a measure of the overall competitiveness
of domestic goods on world markets rather than with respect to another
country's goods.
the exchange rates from n(n - 1) to n(n - 1)/2. If we denote by Tji the exchange
rate of currency i (i = 1,2, ... , n) with respect to currency j (j = 1,2, ... , njj f. i)
in the ith financial centre, that is, given the definition adopted, the number of
UIutS of currency i exchanged for one unit of currency j (price of currency j in
terms of currency i)2, the consistency condition requires that
(2.8)
where k and s are any two currencies. In fact, the consistency condition (also called
neutmlity condition) means that by starting with any given quantity of currency k,
exchanging it for currency S and then exchanging the resulting amount of currency
S for currency k, one must end up with the same initial quantity of currency k.
More precisely, starting with x units of currency k and selling it in financial centre
s for currency s we obtain XTks units of currency Sj if we then seIl this amount of
currency S in financial cent re k for currency k we end up with (XTks) Tsk units of
currency k. The consistency condition x = (XTks) Tsk must therefore holdj if we
divide through by x and rearrange terms we obtain Eq. (2.8). From this equation
it immediately follows that
(2.9)
2The reader should note that the order of the subscripts is merely conventional, so
that many authors (as here) use Tji to denote the price of currency j in terms of currency
i, whereas others follow the reverse order and use Tij to denote the same concept. It is
therefore important for the reader to carefully check which convention is adopted.
16 Chapter 2. The Foreign Exchange Market
the yen. This continues as long as arbitrage is no longer profitable, that is, when
the exchange rates between the two currencies in the two financial centres have
been brought to the point where they satisfy the condition of neutrality.
In practice this condition is never exactly satisfied, because of possible friction
and time-Iags (such as, for example, transaction costs, different business hours,
different time zones, etc.), but in normal times the discrepancies are so small as
to be negligible for all purposes.
After exarnining the relations between the bilateral exchanges rates, we must
now introduce the notion of indirect or cross (exchange) rate. The cross rate
of currency i with respect to currency j indicates how many units of currency
exchange indirectly (this is, through the purchase and sale of a third currency, m)
for one unit of currency j. More precisely, with one unit of currency j one can
purchase Tjm units of currency m in financial centre jj by selling this amount of
currency m for currency i in financial cent re i at the exchange rate Tmi, one obtains
TjmTmi units of currency i. The indirect rate between currency i and currency j
is thus TjmTmi. The consistency (or neutrality) condition obviously requires that
the indirect and direct rates should be equal, and as the direct rate of currency i
with respect to currency j is Tji, the mathematical relation which must hold is
(2.10)
for any triplet of (different) indexes i,j,m. This condition can also be written-
recalling that, from (2.9), we have Tji = I/Tij,T mi = I/Tim-as
(2.11)
If, for example, the US dollar/euro rate in new York is 1.069 dollars per one
euro, and the yen/dollar exchange rate in Tokyo is 115.20737 yen per one dollar in
Tokyo, then the euro/yen cross rate in Tokyo is 115.20737 x 1.069 = 123.1566785
yen per one euro. It is still arbitrage, this time in the form of three-point or
triangulaT aTbitrage (as th!ee currencies are involved), which equalizes the direct
and indirect exchange rate.
The considerations made above on the almost instantaneousness and negligible
cost of the various operations also explain why these will continue until the direct
and indirect exchange rates are brought into line, so as to cause the profit to
disappear. This will, of course, occur when, and only when, the direct exchange
rate between any two currencies coincides with all the possible cross rates between
them. In practice this equalization is never perfect, for the same reasons as in the
case of two-point arbitrage, but here too we can ignore these discrepancies.
It can readily be checked that the cross rates between any pair of currencies
(i,j) are n-2: in fact, as there are n currencies, it is possible to exchange currencies
i and j indirectly through any one of the other (n - 2) currencies. And, since all
these cross rates must equal the only direct rate between currencies i and j, it can
easily be shown that it is sufficient to know the n - 1 direct rates of one currency
vis-a-vis all the others to be able to determine the fuH set of (direct) exchange
rates among the n currencies. Let us in fact assume that we know the n - 1 direct
rates of one currency, say currency 1, vis-a-vis all the others: that is, we know the
2.3. The Forward Exchange Market 17
Now, since the rates rjl and rml are known by assumption, from Eqs. (2.13) and
(2.14) it is possible to determine all the direct exchange rates between all pairs
of currencies (m, j) and therefore the full set of bilateral exchange rates. This
completes the proof of the statement made at the beginning of this section.
3Some writers distinguish between covering and hedging. Covering (by means of for-
ward exchange) is an arrangement to safeguard against the exchange risk on a payment
of adefinite amount to be made or received on a definite date in connection with a self-
liquidating commercial or financial transaction. Hedging (by means of forward exchange)
is an arrangement to safeguard against an indefinite and indirect exchange risk arising
from the existence of assets or liabilities, whose value is liable to be affected by changes in
spot rates. More often, however, no distinction is made and hedging (in the broad sense)
is taken to include all operations to safeguard against the exchange risk, however it arises.
2.3. The Forward Exchange Market 19
between the forward exchange rate and the (current) spot exchange rate.
For this we need to define the concept of forward premium and discount.
A forward premium denotes that the currency under consideration is more
expensive (of course in terms of foreign currency) for future delivery than
for immediate delivery, that is, it is more expensive forward than spot. A
forward discount denotes the opposite situation, i.e. the currency is cheaper
forward than spot. The higher or lower value of the currency forward than
spot is usually measured in terms of the (absolute or proportional) deviation
of the forward exchange rate with respect to the spot exchange rate.
We observe, incidentally, that in the foreign exchange quotations the for-
ward exchange rates are usually quoted implicitly, that is, by quoting the
premium or discount, either absolute or proportional. When the forward ex-
change rate is quoted explicitly as a price, it is sometimes called an outright
forward exchange rate. We also observe, as a matter of terminology, that
when the spot price of an asset exceeds (falls short of) its forward price, a
backwardation (contango, respectively) is said to occur.
This is one of the cases where it is most important to have a dear idea of
how exchange rates are quoted (see Sect. 2.1). If the price quotation system
is used, the higher value of the currency forward than spot means that the
forward exchange rate is lower than the spot exchange rate, and the lower
value of a currency forward than spot means that the forward exchange rate
is higher than the spot rate. But if the volume quotation system is used the
opposite is true: the higher (lower) value of a currency on the forward than
on the spot foreign exchange market means that the forward exchange rate is
higher (lower, respectively) than the spot rate. If, say, the $ in New York is
more expensive forward than spot with respect to the euro, this means that
fewer dollars are required to buy the same amount of euros (or, to put it the
other way round, that more euros can be bought with the same amount of
dollars) on the forward than on the spot exchange market, so that if the USA
uses the price quotation system, in New York the $/euro forward exchange
rate will be lower than the spot rate, whereas if the USA used the other
system, the opposite will be true.
Therefore in the case 'of the prke quotation system the forward premium
will be measured by a negative nillnber (the difference forward minus spot
exchange rate is, in fact, negative) and the forward discount by a positive
number. This apparently counterintuitive numerical definition (intuitively it
would seem more natural to associate premium with a positive number and
discount with a negative one) is presumably due to the fact that this ter-
minology seems to have originated in England, where the volume quotation
system is used, so that by subtracting the spot from the forward exchange
rate one obtains a positive (negative) number in the case of a premium (dis-
count). Be this as it may, having adopted the price quotation system and
letting r denote the generic spot exchange rate and r F the corresponding
2.3. The Forward Exchange Market 21
(2.15)
gives a measure of the forward premium (if negative) and discount (if posi-
tive). As there are different maturities for forward contracts, in practiee the
proportional difference (2.15) is given on aper annum basis by multiplying it
by a suitable factor (if, for example, we are considering the 3-month forward
rate, the approximate factor is 4) and as a percentage by multiplying by 100.
The reason why the forward margin (a margin is a premium or a discount)
is expressed in this proportional form is that, in this way, we give it the
dimension 0/ an interest rate and can use it to make comparisons with the
(domestic and foreign) interest ratesj expression (2.15) is, in fact, sometimes
called an implicit interest rate in the forward transaction.
So equipped, we can go back to compare the various alternatives of the
agent who has to make a future payment (the case of the agent who has
to receive a future payment is perfectly symmetrie). We first show that
alternatives (b) and (c) are equivalent. We have already seen that the costs
are equivalentj as regards the benefits, we can assurne that the discount made
by the foreign creditor for advance payment (case b) is percentually equal to
the interest rate that our debtor might earn on foreign currency invested in
the creditors country (case c). More precisely, let ih and i f be the horne and
the foreign interest rate respectively, referring to the period considered in the
transaction (if, for example, the delay in payment is three months, these rates
will refer to a quarter), and x the amount of the debt in foreign currency.
With alternative (b), thanks to the discount allowed by the foreign creditor,
it is sufficient to purchase an amount x / (1 + i f) of foreign currency now. The
same is true with alternative (c), because by purchasing an amount x / (1 + i f )
of foreign currency now and investing it in the creditor's country for the given
period at the interest rate i f, the amount [x / (1 + i f )]( 1 + i f) = x will be
obtained at the maturity of the debt. The purchase of this amount of foreign
currency spot requires the immediate outlay of an amount r[x/(l + if)] of
domestic currency.
Therefore, if we consider the opportunity cost of domestie funds (interest
foregone on owned funds, or paid on borrowed funds), referring to the period
considered, the total net cast of the operation in cases (b) and (c), referring
to the maturity date of the debt, is obtained by adding this opportunity cost
to the sum calculated above. Thus we have
(2.16)
Let us now consider case (a): the agent under consideration will have to
pay out the sum r F x in domestic currency when the debt falls due. It is
then obvious that alternative (a) will be better than, the same as, or worse
22 Chapter 2. The Foreign Exchange Market
than the other one [since (b) and (c) are equivalent, there are actually two
alternatives] according as
rx (1 + Zh
< --.-
rF x:> . ). (2.17)
1 + zf
(2.18)
On the left-hand side we meet our old friend, the forward margin; the
numerator of the fraction on the right-hand side is the interest (rate) dif-
ferential between the domestic and the foreign economy. Formula (2.19) is
often simplified by letting 1 + if ~ 1, but this is legitimate only when if
is very small (for a precise determination of the degree of approximation,
see Sect. 4.1). The condition of indifference between the alternatives then
occurs when the forward margin equals the interest rate differential.
It is interesting to observe that an absolutely identical condition holds in
the case of covered interest arbitrage, that will be treated in Chap. 4, Sect.
4.l.
We conclude the section by observing that in the forward exchange market
the same type of arbitrage operations on foreign exchange takes place as
described in relation to the spot market (see Sect. 2.2), so that the direct
and indirect (or cross) forward rates come to coincide.
2.4.1 Speculators
In general, speculation can be defined as the purchase (sale) of goods, assets,
etc. with a view to re-sale (re-purchase) them at a later date, where the
motive behind such action is the expectation of a gain deriving from a change
in the relevant prices relatively to the ruling price and not a gain accruing
through their use, transformation, transfer between different markets, etc.
In general, the agent who expects an increase in the price of an asset is
called a bult, whereas a bear is one who expects a decrease in the price of an
asset. Therefore, if we denote by r the expected future spot exchange rate,
a bull in foreign currency (r > r) will normally buy foreign currency (have
a long position) and a bear (r < r) will normally sell foreign currency (have
a short position). Both deliberately incur an exchange risk to profit from
the expected variation in the exchange rate. This risk is usually accounted
for by introducing a risk premium, which will be the greater, the greater
the dispersion of expectations and the size of commitments. More precisely,
consider for example a bull, whose expected speculative capital gain in per-
centage terms is given by (r - r)jr. Although the interest rate gain (from
placing the funds abroad) and loss (forgone domestic interest) are negligible
in speculative activity, they have to be taken into account for a precise eval-
uation, hence the bull will speculate if (f - r)/r + 8 + i f > i h , where 8 is
the risk premium. Similar considerations show that the bear will speculate if
(r - r)jr + 0 + ij < ih. No incentive to speculate in either direction will exist
when
r-r
- - + 0 + ij = i h. (2.20)
r
This is speculation on spot foreign exchange, besides which a forward ex-
change speculation also exists. The latter derives from a divergence between
the current forward rate and the expected spot rate of a currency. If the ex-
pected spot rate is higher than the current forward rate, it is advantageous
for the speculator to buy foreign currency forward, as he expects that, when
the forward contract matures, he will be able to seIl the foreign currency spot
at a price (the expected spot rate) higher than the price that he knows he
will pay for it (the current forward rate). In the opposite case, namely if the
expected spot rate is lower than the current forward rate, it is advantageous
for the speculator to sen foreign currency forward, in the expectation of being
able to buy it, at delivery time, at a price (the expected spot rate) lower than
24 Chapter 2. The Foreign Exchange Market
the price that he knows he will be paid for delivering it (the current forward
rate).
We have talked of delivery etcetera. In practice, the parties of a forward
exchange speculative transaction settle the difference between the forward
exchange rate and the spot exchange rate existing at maturity, multiplied by
the amount of currency contemplated in the forward contract. It should also
be noted that, in principle, forward speculation does not require the availabil-
ity offunds (neither command over cash nor access to credit facilities) at the
moment the contract is stipulated, by the very nature of the forward contract
(both payment and delivery are to be made at a future date). In practice
banks often require the transactor in forward exchange to put down a given
percentage of the contract as collateral; this percentage depends, amongst
other things, on the efficiency and development of the forward market, and
on possible binding instructions of central banks.
2.4.2 Non-Speculators
A second functional category is that of non-speculators. This category in-
cludes exporters and importers of goods and services, businesses which carry
out investment abroad, individual or institutional savers who wish to di-
versify their portfolios between national and foreign assets on the basis of
considerations of risk and yield (excluding speculative gains), arbitrageurs,
etcetera. Non-speculators are more precisely defined by exclusion, i.e., those
agents who are neither speculators nor monetary authorities.
2.5.1 Futures
2.5.2 Options
A currency option is a contract that gives the purchaser the right (but not
the obligation) to buy or seIl a currency at a predetermined price (exchange
rate) sometime in the future. Hence options are a much more complicated
instrument than forwards and futures. They also have a precise terminology,
which is the following.
The party selling the option is called the writer, while the purchaser is
the holder. A call option gives the holder the right to purchase the currency
involved, while a put option gives the right to sell the currency. The currency
in which the option is granted is called the underlying currency, while the
currency in which the price will be paid is the counter currency. For example,
if the contract specifies the right to seIl euro1, 000, 000 at $1.05/euro1, the
euro is the underlying currency while the dollar is the counter currency. The
price at which the underlying currency can be bought or sold is the strike (or
exercise) price. The price that the holder pays to the writer for an option
is known as the option premium. The date at which the contract expires is
called the expiry date or maturity date. Finally, a distinction is made between
the American option (the right to buy or sell the currency at the given price
can be exercised any time up to the maturity date) and the European option
(the right can be exercised only on the maturity date).
There are two main differences between options and forward or futures
contracts, which both derive from the fact that the option gives the holder a
right but not an obligation.
The first is that the option provides the agent interested in hedging with
a more flexible instrument, because it enables him to fix a maximum payable
price (the sum of the option premium plus the exercise price) while leaving
him free to take advantage of favourable movements in the exchange rate.
With a forward or futures contract the hedger is obliged to respect the con-
tract in any case, also when the spot exchange rate at maturity is more
favourable than the forward rate agreed upon when the contract was signed.
On the contrary, with an option the holder can decide not to exercise the
right if the spot exchange rate at the expiry date is more favourable than
that the exercise price, account being taken of the option premium.
Suppose, for example, that a US company has to make a payment of
.LI million in sixth months' time, and that the forward/futures exchange
rate is $1.50/1:1. Alternatively, the company can buy a call option with an
exercise price of $1.50/1:1 for 8 cents per pound (the option premium). At
maturity, if the spot exchange rate is higher than $1.50/.LI, the US firm will
exercise the option. This is in any case cheaper than buying pounds spot,
but of course more expensive than would have been with the forward/futures
contract, given the option premium. If the spot exchange rate is lower than
$1.50/1:1, the firm will not exercise the option and buy pounds spot. The
cost will again be higher than with a forward/futures contract, but only if
2.5. Currency Derivatives 27
the spot exchange rate is higher than $1.42/.LI, because adding the option
premium (8 cents per pound) the price paid will be higher than $1.50/ .n. If
the spot exchange rate is lower than $1.42/ f.l, the option will have provided
a cheaper means of hedging than a forward/futures contract.
The second difference concerns the asymmetry in the risk-return char-
acteristics of the contract. With a forward/futures contract, for every cent
the spot exchange rate at the date of expiry is above (below) the exchange
rate established in the forward/futures contract, the buyer makes (loses) a
cent and the seller loses (makes) acent. This means a perfect symmetry.
On the contrary, with an options contract, the maximum loss of the option
holder equals the option premium, which is also the maximum gain for the
option writer, but there is unlimited potential gain for the option holder and,
correspondingly, unlimited potential loss for the option writer. This feature
makes options very attractive for speculators, because speculative holders can
combine limited losses (the premium paid) with unlimited potential profit.
open simultaneously only for a short time in the day, so that Europeans who had
to borrow or lend dollars found it convenient to do this directly in London rather
than in New York through a London bank.
Be this as it may, the enormous growth of the Xeno-currency markets has
complicated the international financial market: let it suffice to think of the greater
complexity of interest arbitrage operations and of the birth of new types of interna-
tional banking transactions. As regards these, they can be classified in four main
types: onshore-foreign, offshore-foreign, offshore-internal, and offshore-onshore.
The first word of each pair refers to the currency in which the bank is transact-
ing: if it is that of the country in which the bank is resident the transaction is
onshore, whilst if it is the currency of another country the transaction is offshore.
The second word refers to the residence of the customer (borrower or lender): the
customer is internal if resident in the same country as the bank, foreign if resident
in a country different from that where the bank is resident and also different from
the country which issues the currency being transacted; in fact, the customer is
onshore if resident in the country issuing the currency.
Before the birth of Xeno-currencies, international banking transactions were
entirely of the onshore-foreign type: an example of an onshore-foreign deposit is a
deposit in dollars placed with ehase Manhattan, New York, by a non-US resident.
The growth of offshore deposits related to Xeno-currencies has given rise to the
multiplication of the three other types of international banking transactions.
An example of an oiJshore-foreign deposit is a deposit in euros placed with a
Swiss bank by a Japanese resident.
An example of an oiJshore-internal deposit is a deposit in US dollars placed
with a Dutch bank by a Dutch resident.
Finally, an example of an oJJshore-onshore deposit is a deposit in US dollars
placed by a US resident with a Japanese bank.
When international capital flows where to some extent subject to controls
(this was the normal situation during the Bretton Woods system, and also after
its collapse several countries maintained capital controls), a specific analysis of
Xeno-markets, by their very nature exempt from national controls (a situation
that worried central bankers very much), was very important. But in the early
1990s completely free international mobility of capital became the rule rather than
the exception, hence this importance no longer exists.
Exchange rate quotations can be found in all financial newspapers, such as The
Wall Street Journal and The Financial Times.
Chapter 3
31
32 Chapter 3. Exchange-Rate Regimes and the International Monetary System
in which, without itself buying and selling gold, a country stands ready to buy
or sell a particular foreign currency which is fully convertible into gold. This
system enables the international economy to economize gold with respect to
the gold standard, because the ultimate requests for conversion into gold of
the convertible foreign currency are normally only a fraction of the latter.
It must be emphasized that, for this system to be a true gold exchange
standard, the convertibility of the foreign currency must be free and full, so
that it can be demanded and obtained by any agent. In this case the system
is equivalent to the gold standard.
If, on the contrary, the convertibility is restricted, for example solely to
the requests from central banks, we are in the presence of a limping gold
exchange standard, in which case the automatic mechanisms governing the
gold standard no longer operate, and the concept itself of convertibility has
to be redefined: now convertibility simply means that private agents have the
right to freely exchange the various currencies between each other at fixes
rates. When convertibility into gold is completely eliminated, even between
central banks, we have a pure exchange standard, in which a country buys
and sells foreign exchange (or a stipulated foreign currency) at fixed rates.
Other fixed exchange rate arrangements include a) situations where coun-
tries have no national currency, either because they belong to a currency
union or because they have formally adopted the currency of some other
country (so called dollarization, but the foreign currency may be other than
the dollar), and b) currency boards. The main characteristics of a currency
board arrangement are: 1) the board stands ready to exchange domestic cur-
rency for the foreign reserve currency at a fixed rate; 2) to ensure the rigidity
of this rate, the board is required to hold liquid financial assets in the reserve
currency at least equal to the value of the domestic currency in circulation.
Hence in this system there can be no fiduciary issue of domestic money.
Arrangements a) and b) are usually described as "hard pegs".
of dollars into gold was restrieted to the requests from central banks.
The member countries were required to stand ready to maintain the de-
clared parity in the foreign exchange market by buying and selling foreign ex-
change (usually dollars, which thus became the main intervention currency);
more precisely, the actual exchange rate could vary only within the so-called
support (or intervention) points, which were initially set at 1 percent above
or below parity.
The modifications consisted in the fact that parity, notwithstanding the
obligation to defend it, was not immutable, but could be changed in the case
of "fundamental disequilibrium" in accordance with certain rules: changes
up to 10% could be made at the discretion of the country, whilst for greater
changes the country had first to notify the IMF (the International Monetary
Fund, which is one of the international organizations set up by the Bretton
Woods agreement) and obtain its assent.
The obligation to maintain the declared parity together with the possibil-
ity of changing it gave the system the name of adjustable peg. The idea behind
it was a compromise between rigidly fixed and freely flexible exchange rates,
and it is clear that the greater or lesser extent to which it approached either
system depended essentiallyon the interpretation of the rules for changing
parity. The prevailing interpretation was restrietive, in the sense that parity
was to be defended at all costs and changed only when it was unavoidable.
these demands and supplies are not necessarily weH behaved, and depend on
a lot of other factors, which will determine shifts in the schedules, but we
shaH ignore these complications. We further assume that the market behaves
as aH other markets, i.e., the price (in our case the price of foreign currency is
the exchange rate) tends to increase (decrease) if there is an excess demand
(excess supply) in the market.
S(r)
D(r)
foreign exchange
Let us now suppose that the exchange rate has to be pegged at T' whilst
the market is in equilibrium at Te. In the absence of official intervention, the
exchange rate would move towards Te, driven by the excess supply of foreign
exchange. To prevent this from happening, the monetary authorities must
absorb, as residual buyers, the excess supply A' B' (providing the market
with the corresponding amount of domestic currency). If, on the contrary,
the exchange rate were to be pegged at T", to prevent it from depreciating
towards Tein response to the pressure of excess demand for foreign exchange,
the monetary authorities would have to meet (as residual seHers) the excess
demand, by supplying an amount A" B" of foreign currency to the market
(absorbing from the market the corresponding amount of domestic currency).
It should be pointed out that, as the schedules in question represent
fiows, the monetary authorities must (ceteris paribus ) go on absorbing A' B',
or supplying A" B", of foreign exchange peT unit of time. This may weH give
rise to problems, especially in the case T", because by continuously giving
up foreign exchange the monetary authorities run out of reserves. These
3.3. Other Limited-Flexibility Systems 35
the situation at the moment of going to the press, is that each country can
choose the exchange-rate regime that it prefers and notify its choice to the
IMF, so that various regimes coexist. Some countries peg their exchange
rate to a reference currency (usually thedollar, but also the euro and other
currendes) with zero or very narrow margins; naturally they will follow the
reference currency's regime with respect to the other countries.
EXCHANGE RATE ARRANGEMENTS
(in percent of IMF membership)
Hard Pegs Regimes 25.8, of which: Dollarization 4.3, Currency union 17.2,
Currency board 4.3
Soft Pegs Regimes 30.1, of which: Fixed peg to a currency 16.7, Fixed peg to a
basket 5.4, Horizontal band 2.7, Crawling peg 2.2, Crawling band 3.1
Floating Regimes 44.1, of which: Tightly managed float 8.6, Other managed
float 14.0, Independently floating 21.5
The classification system is based on the members' actual, de facto regimes that may
differ from their officially announced arrangements. The scheme ranks exchange rate
regimes on the basis of the degree of flexibility of the arrangements.
Source: IMF, Annual Report on Exchange Arrangements and Exchange Restric-
tions, 2003.
Then there are other countries which peg their currency to a compos-
ite currency such as, for example, the IMF's Special Drawing Right (SDR).
A composite currency, also called a "basket-currency" is an artifidal cur-
rency consisting of predeterminOO amounts of various currencies. The Special
Drawing Right issued by the International Monetary FUnd (see below) is a
basket-currency whose composition (revised in 2001; the basket composition
is revisOO every five years) is 45% US dollar, 29% euro, 15% Japanese yen,
11 % pound sterling. Another example of a basket-currency was the ECU
(European Currency Unit).
Groups of countries enter into monetary agreements to form currency
areas (by maintaining fixed exchange rates among themselves), or monetary
unions with a common currency, such as the European Monetary Union.
Further, the situation is continually changing, hence the reader had bet-
ter to consult the Annual Report on Exchange Arrangements and Exchange
Restrictions publishOO by the International Monetary FUnd to know the up-
datOO situation. In the box we report the situation as of 2003.
We conclude this section by stressing again that we have deliberately ab-
stained from giving a comparative evaluation of the various exchange-rate
regimes. The reason is that such an evaluation requires familiarity with
notions (adjustment processes of the balance of payments, macroeconomic
equilibrium in an open economy, etc.), which will be dealt with in the follow-
ing chapters. Thus the comparative evaluation of the different exchange-rate
regimes, which has 100 to hot debates in the literature, has to be deferrOO
(see Sect. 9.5).
38 Chapter 3. Exchange-Rate Regimes and the International Monetary System
exchange rates was dose to a regime with a single world currency. And,
in fact, it brought about great benefits, amongst which the impetus to the
growth of international trade. We must then ask ourselves why it didn't last
longer.
The answer to this question is generally based on the so called 'friffin
dilemma. In a growing world with growing trade, there is a growing world
demand for money for trans action purposes. In the system under consider-
ation, this amounts to a growing demand for international reserves, that is
for dollars, which can be acquired only by running balance-of-payments sur-
pluses, exduding of course the USo In fact, given the international consistency
condition
the nth country (the US) has to incur a balance-of-payments deficit for the
"n-l
rest-of-the-world to acquire dollars, namely B n < 0 for L..,i=l Bi > O. But
this process poses the problem of the convertibility of dollars into gold. Since
the gold stock owned by the US cannot grow at the same rate as the growth of
dollars held by the rest-of the-world central banks, there is a loss of confidence
in the ability of the US to guarantee convertibility. If, on the other hand,
the United States brought their balance of payments into equilibrium, the
international monetary system would suffer from a liquidity shortage, with
the possible collapse of international trade. Hence the dilemma: if the US
allow the increase in international liquidity through deficits in their balance
of payments, the international monetary system is bound to collapse for a
confidence crisis; if, on the other hand, they do not allow such an increase,
the world is condemned to deflation.
In addition to the 'friffin dilemma, two other concurring causes are brought
into play: the rigidity of the system and the "seignorage" problem. The rigid-
ity 0/ the system was due to the fact that the idea of "fixed but adjustable"
exchange rates (adjustable peg) was interpreted, as observed in Sect. 3.2, in
the restrictive sense. That is to say, the parity was to be defended by all
means, and was to be changed only when any further defence turned out to
be impossible. Hence the external adjustment could not come about through
reasonably frequent exchange-rate variations, but through deflationary poli-
eies.
The seignorage problem was related to the reserve-currency role of the
dollar. This enabled the US to acquire long-term assets to carry out direct
investment abroad in exchange for short-term assets (the'dollars). These dol-
lars were usually invested by the rest-of-the-world central banks in short-term
US 'freasury bills, that carried a relatively low interest and whose purchas-
ing power was slowly but steadily eroded by US inflation. This problem was
much feit by several countries (especially by France under De Gaulle) and,
though not being an essential cause, certainly weakened the will to save the
40 Chapter 3. Exchange-Rate Regimes and the International Monetary System
Now, if we apply this law to the gold-dollar system, it can be seen that the
increase in dollars was not matched by an increase in the gold stock, and that
the official price was losing credibility. In fact, while the purchasing power of
dollars (in terms of goods and services) was decreasing because of inflation,
it remained fixed in terms of gold. Thus the so-called "gold pool" , made up
of the central banks of the most industrialised countries, that acted buying
and selling gold in the private market with the aim of stabilizing its price
at the official level, was faced with increasing gold demand from the private
sector at the given official price ($ 35/oz.). The ensuing gold loss compelled
the gold pool to discontinue any intervention and leave the private price free,
maintaining the official price for transactions between central banks only.
The August 1971 official declaration of inconvertibility of the dollar was just
the de jure acknowledgment of a de facto situation, namely the functioning
of the system as a dollar standard.
If one accepts this analysis, the corollary follows that, even if the Triffin
dilemma were hypothetically solved, the working of Gresham's law would not
make it advisable a "return to gold" (as some still advocate).
3.5.2 Petrodollars
The repeated increases in oil prices 1 charged since October 1973 by the oil
producing and exporting countries united in the OPEC cartel (Organization
of Petroleum Exporting Countries), gave rise to serious balance-of-payments
problems in the importing countries and to the accumulation of huge dollar
balances (also called petrodollars) by these countries, as oil was paid in dol-
lars. In 1974 the (flow) financial surplus of the OPEC countries was about $
56 billion, which mirrored an equal overall deficit of the oil importing coun-
tries vis-a-vis OPEC countries. About two-thirds of this deficit concerned
industrial countries, the remaining third the non-oil-producing developing
countries (also called the "Fourth World"). The problem of financing the
oil deficits was particularly acute given the suddenness and the amount of
the price increases, and was solved by means of the so-called recycling pro-
cess, through which the oil surpluses of the OPEC countries were lent back
(indirectly) to the deficit countries. The recycling process operated mostly
through market mechanisms, both international (the Eurodollar market) and
national (the US and UK financial markets), and for the rest through ad hoc
mechanisms. Since the OPEC countries invested their petrodollar surpluses
in the Eurodollar market and in the US and UK financial markets, the deficit
countries could borrow the dollars that they needed by applying to the Eu-
rodollar market and to the US and UK financial markets. The ad hoc mech-
anisms concerned both bilateral agreements between an oil importing and
an oil exporting country, and agreements brought into being by international
organizations. The latter included the IMF's "oil facility" brought into being
in 1974 (and discontinued in 1976) and financed by borrowing agreements be-
tween the IMF and other countries (mostly oil producing countries) with the
aim of granting loans, with certain conditions, to countries facing balance-
of-payments disequilibria due to oil deficits. The agreements also included
borrowing facilities created within the EEC and OECD organisations.
The recycling process was of course a short-run solution as it did not solve
the problem of the elimination of the oil surpluses and deficits. This can be
examined in the context of the theory of cartels (see Chap. 14, Sect. 14.4.2).
ITo set the problem into proper perspective it should be pointed out that these increases
occurred after a long period of low oll prices, which, though stable in nominal terms, had
actually been decreasing in real terms. Many people wonder therefore whether a more far
sighted policy (on the part of all those concerned), of gradual price increases in the period
before 1973, might not have avoided the problems created by the huge and sudden 1973
price increase.
42 Chapter 3. Exchange-Rate Regimes and the International Monetary System
mIed the demonetization of gold, thus removing the privileged status that
gold had previously enjoyed, with the ultimate aim of making it like any
other commodity.
The Bretton Woods Agreement gave gold a central place in the inter-
national monetary system, as was elear from the obligation on the part of
members to pay out twenty-five per cent of their IMF quota in gold and to
declare the par value of their currencies in terms of gold or, alternatively,
in terms of the US dollar (which was the same thing, since the dollar was
convertible into gold at the irrevocably fixed price of $ 35 per ounce). The
maintenance of this official price of gold was an easy matter until the end
of the fifties. But in 1960 problems began to arise because of the fact that
the free market price of gold, quoted in some financial centres (mainly Lon-
don), began to diverge from the official price, mostly because of speculative
hoarding. To counter this, the central banks of eight countries, by the Basle
Agreements of 1961 and 1962, constituted the so-called Gold Pool, which-
by using the gold provided by the central banks themselves-had the task of
intervening on the free market to stabilize the gold price, by buying (selling)
gold when its price fell (rose) beyond certain limits with respect to the official
price. In this way a single price of gold prevailed. But in 1968 the Gold Pool
was discontinued, because of huge losses of gold due to increasing speculative
pressures on the market, and the two-tier market for gold was established:
the official market for the transactions between central banks, at the agreed
price of $ 35 per ounce, and the free market, where the price was formed
by the free interplay of supply and demand; the central banks agreed not to
intervene on the free market.
The increase in the official price of gold to $ 38 and then to $ 42.44
did not solve the problem of the gold-reserve freeze, due to the fact that no
central bank was willing to meet its international payments by releasing gold
at the official price when the market price was much higher. Partial solutions
were found within the EEC (1974), with the settlement in gold of payment
imbalances at an official gold price related to the market price. However, the
general problem of the function of gold remained, and there were two main
schools of thought.
The first aimed at maintaining the monetary function of gold as the
paramount international means of payment and suggested the revaluation
of its official price to bring it elose to the market price. The second aimed
at eliminating this function so aB to allow the emergence of international
fiduciary means of payments, as had happened in the individual national
economic systems (where the link with gold hadlong been eliminated and re-
placed by fiduciary money). The second school prevailed, and this is reflected
in the Kingston (Jamaica) Agreement of January 1976, which contemplated,
amongst other,
(1) The elimination of the function of gold as numeraire of the system,
i.e. as common denominator of par values of currencies.
3.6. International Organisations 43
called a quota subscription, which is determined by the IMF itself on the ba-
sis of the country's wealth and econornie performance. Quotas are reviewed
every five years and can be lowered or raised on the basis of the needs of the
IMF and the econornic situation of the member. In 1946, the 39 members
paid in the equivalent of $7.6 billion; by 2001, the 183 members had paid in
the equivalent of about $269 billion, and there is a proposal to raise quotas
to a still higher value.
Quotas serve various purposes:
1) they eonstitute the resources from which the IMF draws to make loans
to members in financial diffieulty;
2) they are the basis for deterrnining how much a member can borrow
from the IMF, or receives from the IMF in periodic allocations of special
assets known as SDRs (Special Drawing Rights; see above, Sect. 3.4);
3) they deterrnine the voting power of the member.
The highest link of the chain of eommand in the Fund is the Board of
Governors and Alternate Governors (one Governor and one Alternate per
member). These persons are ministers of finance or heads of central banks,
and meet onee a year. The day-to-day management of the Fund is delegated
to the Executive Board chaired by the Managing Direetor. The Executive
Board at the moment consists of 24 persons, eight of whom represent in-
dividual countries (China, France, Germany, Japan, Russia, Saudi Arabia,
the United Kingdom, the United States), while the remaining sixteen repre-
sent groups of countries. By tradition, the Managing Director is a non-US
national, while the President of the World Bank is a US national.
The IMF performs three main functions in the interest of an orderly
functioning of the international monetary system:
1) Surveillance. After the dernise of the Bretton Woods system it seemed
that the preerninent role of the IMF would disappear. This has not hap-
pened, as under the current system the IMF has been entrusted with the
exarnination of all aspects of any member's economy that are relevant for
that member's exchange rate and with the evaluation of the economy's per-
formance for the entire membership. This entails more scope for the IMF
to monitor members' policies. The activity we are describing is called by
the IMF "surveillance" over members' exchange policies, and is carried out
through periodie consultations conducted in the member country.
2) Financial assistance. This is perhaps the most visible activity to the
general public: as of July 1998, for example, the IMF had committed about
$35 billion to Indonesia, Korea and Thailand to help them with their finan-
cial crisis (the Asian crisis) and around $21 billion to Russia to support its
econornic program. The general rules for obtaining financial assistance from
the fund are the following:
i) a member country with a payments problem can immediately withdraw
from the IMF the 25 percent of its quota. Once this was called the gold
tranche position, because it had to be paid in gold to the Fund. After the
3.6. International Organisations 45
individual countries (France, Germany, Japan, the United Kingdom and the
United States) while the remaining 19 represent groups of countries.
While the task of the IMF is to promote a weH functioning and orderly
international monetary system, the main task of the World Bank is to pro-
mote growth of poorer countries. Contrary to the IMF, whose resources are
the members' quotas, the World Bank raises almost all its funds in finan-
cial markets by selling bonds and other assets. Its average annualloans are
around $30 billion.
Over the years the World Bank has become a group, consisting of five
institutions: the IBRD proper, the IDA (International Development Asso-
ciation, whose assistance is focused on the poorest countries), IFC (Inter-
national Finance Corporation, which provides finance for business ventures
in developing countries), MIGA (Multilateral Investment Guarantee Agency,
which covers foreign investors in developing countries against non-commercial
risks) , and ICSID (International Center for Settlement of Investment Dis-
putes, which arbitrates disputes between foreign investors and the country
where they have invested).
The IMF in providing financial assistance, and the World Bank in pro-
viding development assistance, have in mind a model. The model on which
the IMF-WB growth-oriented adjustment programs are based will be treated
in Chap. 12, Sect. 12.4.
have involved not only tariffs, but also subsidies and countervailing measures,
anti-dumping, teehnieal barriers to trade, government procurement, and so
on.
The original agreement (ealled GATT 1947) was amended and updated
in 1994 (GATT 1994). GATT 1994 is an integral part of WTO, which was
established on 1st January 1995.
As the names say, WTO is an organization, while GATT 1947 was an
agreement. This is not only a semantie differenee or a juridical subtlety: an
agreement is simply a set of mIes with no legal institutional foundation; a
(permanent) organization is an institution with legal personality and its own
secretariat and powers. This implies, amongst other, that the WTO dispute
settlement system is faster and more automatie, and the implementation of
its deeisions on disputes is more easily assured. From the economie point of
view, WTO has a greater seope than GATT, for GATT mIes applied solely
to trade in merchandise, while WTO in addition to goods also covers trade in
services (GATS, General Agreement on Trade in Services) as weIl as trade-
related aspects of intelleetual property (TRIPS, Trade-Related aspects of
Intellectual Property rights). Member countries are requested to make their
trade policies transparent by notifying the WTO ab out laws in force and
measures adopted, and the secretariat issues regular reports on countries'
trade policies.
GATT, and now WTO, are sometimes deseribed as free-trade institutions.
This is not entirely eorrect, if only because tariffs (and, in limited circum-
stances, other forms of protection) are permitted. The basic aim of GATT
and WTO mIes is to secure open, fair and undistorted competition in inter-
national trade. Rules on non-diserimination (such as the MFN dause and the
national treatment principle, whieh eondemns diserimination between foreign
and national goods in terms of taxation and regulation, onee the applicable
border measures have been satisfied), as weIl as those on dumping and subsi-
dies (governments are allowed to impose compensating duties on these forms
of unfair eompetition), are designed to bring about fair conditions of trade.
1) The BIS assists central banks and other financial authorities in their
efforts to promote greater monetary and financial stability. This assistance
takes two forms:
1a) Direct contributions to international cooperation. The BIS provides
an institutional framework for cooperation in the monetary and financial area
and serves as a meeting place mainly for central banks, but also for other
financial and regulatory authorities.
1b) Services to committees established by central bank Governors over the
course of the past decades, and support to a number of other groupings with
secretariats at the BIS. The most important committees are, in chronolog-
ical order of establishment, the Markets Committee (1962), the Committee
on the Global Financial System (1971), the Basle Committee on Banking
Supervision (1974) and the Committee on Payment and Settlement Systems
(1990).
2) It acts as a bank, almost exclusively for central banks, providing ser-
vices related to their financial operations. Trustee and collateral agency
functions are also part of these services. In addition to its central bank
customers, the BIS also acts as a banker to, and manages funds for, a num-
ber of international financial institutions. The Bank's Statutes do not allow
the Bank to open current accounts in the name of, or make advances to,
governments.
Three decision-making bodies are relevant within the Bank's governance
structure: the General Meeting of member central banks (currently 51), the
Board of Directors and the Management of the Bank, headed by a General
Manager. Decisions taken at each of these levels concern the running of
the Bank and as such are mainly of an administrative and financial nature,
related to its banking operations, the policies governing internal management
of the BIS and the allocation of budgetary resources to the different business
areas.
International Interest-Rate
Parity Conditions
The relations between interest rates (domestic and foreign) and exchange
rates (spot and forward) that were already mentioned in Chap. 2, are very
important and frequently used in international finance. Hence, we give here a
general overview, with additional important considerations on the efficiency
of the foreign exchange market and on capital mobility.
49
50 Chapter 4. International Interest-Rate Parity Conditions
Note that this inequality coincides with the inequallty concerning the various
covering alternatives of commercial traders, see Eq. (2.19), of course when
both refer to the same period of time.
The condition of equallty in (4.2) or in (4.3), that iswhen funds have
no incentive to move from where they are placed, is called the neutrality
condition and the forward rate is said to be at interest parity or simply
that covered interest parity (CIP) prevails, and the corresponding forward
exchange rate is called the parity forward rate. When there is a difference
between the forward margin and the interest rate differential such that funds
tend to flow in (out), we say that there is an intrinsie premium (discount)
for the domestic currency.
The equations that define CIP
rF 1 + ih rF - r ih - i1
(4.4)
r 1+i/ --r- = 1 + i1 '
can be written in alternative specifications, that are often used in the litera-
ture, which are obtained by an approximation!
(4.5)
lAs can easily be checked, i;~; = (ih -i,) - (ih -i,)~. Hence the approximation
error of using ih - i, in the place of i;:;ii/
is very small, since it is given by - (i h - i,) ~,
which is of the second order of magnitude.
4.2. Uncovered Interest Parity (UIP) 51
i.e. the interest differential equals the forward margin, or the domestic inter-
est rate equals the foreign interest rate plus the (positive or negative) forward
margin. These equations are also referred to as the covered interest parity
conditions.
(4.6)
where the interest rates and expectations are referred to the same time hori-
zon. If the two sides of Eq. (4.6) are not equal, the agent can earn a profit
by shifting funds in or out of the country according as the left-hand side of
Eq. (4.6) is greater or smaller than the right-hand side.
If we divide both members of (4.6) by (1 + if), we get
1 + ih r
--=-
1 + if r'
~h - ~f r- r
--=-- (4.7)
1 + if r
(4.10)
where iRh is the real interest rate and Wh - Ph)/Ph is the expected inflation
rate. A similar definition holds for the rest-of-the-world real interest rate,
namely
(4.11)
4.5. Effideney of the Foreign Exchange Market 53
Let us now consider the uncovered interest parity condition (4.8), that we
report here for the reader's convenience,
. . r-r
~h = ~i + --,
r
and assume that the expected variation in the exchange rate equals the dif-
ference between the expected inflation rates in the two countries, namely
Pi -Pi
(4.12)
r Ph Pi
have to consider the possible discrepancy between the two sides of Eq. (4.9).
This last observation also holds for speculators, as we have seen above.
If we assurne the interest rates as given, in the foreign exchange market
the variables that must reflect the available information are the spot (both
current and expected) and forward exchange rate. Hence, since both CIP
and UIP hold if the market is efficient, we have
whence
(4.14)
from which
(4.15)
namely the forward exchange rate and the expected spot exchange rate (both
referred to the same time horizon) coincide. In a stochastic context it turns
out that the forward exchange rate is an unbiased and efficient predictor of
the future spot exchange rate.
In the case ofrisk premium, using (4.5) and (4.9) we get
r F -r - -r
r
--=--+<5, (4.16)
r r
namely the forward margin is equal to the sum of the expected variation in
the spot exchange rate and the risk coefficient. In terms of levels we have
rF = r+RP, (4.17)
The distinction between perfect capital mobility and perfect asset sub-
stitutability (advocated by Dornbusch and Krugman, 1976) is not generally
accepted. Other authors (for example Mundell, 1963, p. 475), in fact, in-
clude in the notion of perfect capital mobility not only instantaneous portfolio
adjustment, but also perfect asset substitutability. We believe that the dis-
tinction between the two notions is important and useful, therefore we shall
keep it.
57
58 Chapter 5. The Balance of Payments
(a) The first basic principle is that, as all accounting documents, the bal-
ance of payments is kept under standard double-entry bookkeeping. Therefore
each international trans action of the residents of a country will result in two
entries that have exactly equal values but opposite signs: a credit (+) and
adebit (-) entry in that country's balance of payments. The result of this
accounting principle is that the total value of debit entries necessarily equals
the total value of credit entries (so that the net balance of all the entries is
necessarily zero), that is, the balance of payments always balances.
Naturally, the credit and debit entries are not arbitrary but must follow
precise rules. "Under the conventions of the system, a compiling economy
records credit entries (i) for real resources denoting exports and (ii) for fi-
nancial items reflecting reductions in an economy's foreign assets or increases
in an economy's foreign liabilities. Conversely, a compiling economy records
debit entries (i) for real resources denoting imports and (ii) for financial items
reflecting increases in assets or decreases in liabilities. In other words, for
assets-whether real or financial-a positive figure (credit) represents a de-
crease in holdings, and a negative figure (debit) represents an increase. In
contrast, for liabilities, a positive figure shows an increase, and a negative
figure shows a decrease. Transfers are shown as credits when the entries to
which they provide the offsets are debits and as debits when those entries
are credits" (from the Manual, p. 7).
Note that under this rule an increase in foreign liabilities of the country
(the sale of domestic securities to non residents gives rise to such an increase)
gives rise to a credit entry, while an increase in foreign assets (the purchase
of foreign securities by residents gives rise to such an increase) gives rise to a
debit entry. A decrease in foreign assets is treated like an increase in foreign
liabilities, and a decrease in foreign liabilities is treated like an increase in
foreign assets.
The convention concerning the recording of finandal items implies that
a capital outftow is a debit in the capital account, as it gives rise to an
5.1. Balance-of-Payments Accounting and Presentation 59
(b) The second basic principle concerns the timing of recording, that is,
the time at which transactions are deemed to have taken place. In general,
various rules are possible; the principle adopted by the Fund's Manual is
the change of ownership principle. By convention, the time of change of
ownership is normally taken to be the time that the parties concerned record
the transaction in their books.
portfolio investment, reserves, other capital. The main feature of direct in-
vestment is taken to be the fact that the direct investor seeks to have, on
a lasting basis, an effective voice in the management of a nonresident en-
terprise; this gives rise to accounting conventions which will be examined in
Sect. 16.1. Portfolio investment covers investment in financial assets (bonds,
corporate equities other than direct investment, etc.). On reserves see below.
"Other capital" is a residual category.
2) The length of the operation: long-term and short-term capital. The
convention adopted is based on the original contractual maturity of more
than one year (long term) and one year or less (short-term). Assets with
no stated maturity (e.g. corporate equities, property rights) are also consid-
ered as long-term capital. The initial maturity convention may give rise to
problems: for example the purchase of, say, a foreign bond with an original
maturity of three years, but only 6 months to maturity when the purchase
is made, is nonetheless recorded as a long-term capital movement. The in-
evitable convention derives from the fact that usually no data are available
on the time to maturity of securities when international transactions occur.
3) The nature of the operator: private and official, the latter possibly
divided in general government, central monetary institutions (central banks
etc.) and other official institutions.
The Manual adopts the first criterion, and categorizes between: 1. Direct
investment, 2. Portfolio investment, 3. Other investment, 4. Reserve assets.
Within these categories further subdivisions are present, which use the other
criteria.
"Reserves" is a junctional category comprising all those assets available
for use by the central authorities of a country in meeting balance of payments
needsj this availability is not linked in principle to criteria of ownership or
nature of the assets. In practice, however, the term reserves is taken to
include monetary gold, special drawing rights (SDRs) in the FUnd, reserve
position in the FUnd, foreign exchange and other claims available to the
central authorities for the use described above.
To conclude this section we point out that, in addition to the balance of
payments, there also exists the balance oj indebtedness (also called by the
Manual the international investment position). In general, the balance of
indebtedness records the outstanding claims of residents on nonresidents and
the outstanding claims of nonresidents on residents at a given point in time,
such as the end of the year. Therefore this balance is concerned with stocks,
unlike the balance of payments, which refers to flows.
Various balances of indebtedness can be reported according to the claims
considered. One is that which considers only the stock of foreign assets and
liabilities of the central authorities; another considers the direct investment
position, etcetera.
62 Chapter 5. The Balance of Payments
Merchandise trade can serve as a channel for clandestine outward capital move-
ments. To achieve this, imports will be overinvoiced, or exports underinvoiced. In
the former case the foreign exporter, who receives a greater amount than the true
value of the goods, will credit the difference to, say, a secret account that the
domestic importer holds abroad. In the latter case the foreign importer, who has
to pay a lower amount than the true value of the goods, will credit the difference
to, say, the account that the domestic exporter secretly holds abroad. These dif-
ferences are (clandestine) capital movements, which of course are not recorded as
such in the capital account, since they are hidden in merchandise trade.
As regards labour income and workers' remittances, the best known device is
that of direct compensation through an illegal organisation. Let us consider Mr.
Y, a worker who has (temporarily or definitively) come to work in country 2 from
country 1, and who wishes to send (part of) his earnings to his relatives in country
1. Mr. Y can either use the official banking, post al and other authorised channels
(in which case there will be arecord in the balance of payments), or give the
money to the organisation's representative in country 2, who offers hirn perhaps
more favourable conditions, or immediate delivery, or hidden delivery (Mr. Y may
be an illegal migrant and doesn't want to be discovered). The money remains in
country 2 to be used by the organisation, and the organisation's representative
in country 2 only has to instruct his counterpart in country 1 to pay the agreed
amount to Mr. Y's relatives out of the organisation's funds held in country 1.
(5.1)
namely the balance of payments of all countries (of course the same type
of balance, and expressed in a common monetary unit) must algebraically
sum up to zero. This follows from the fact that the surplus of one country
must necessarily be matched by a corresponding deficit of one or more other
countries, provided that no reserve creation takes place (more generally, the
sum of the balances of payments of all countries, Le., the world balance of
payments, equals the increase in net world reserves: see Mundell, 1968).
From Eq. (5.1) it follows that only n - 1 balances of payments can be
independently determined. For example, a maximum of n - 1 countries can
independently set balance-of-payments targets: the nth country has to accept
the outcome or the system will be inconsistent.
5.3. Some Important Accounting Relations 65
Domestic
monetary
base ßHp ßHb ßHc 0
Domestic
bank
deposits ßDp ßDb ßDf 0
Domestic
securities ßNp ßN9 ßNb ßNc ßNf 0
Foreign
money ßRp ßRJ, ßRc ßRf 0
Foreign
securities ßFp ßFb ßFc ßFf 0
COLUMN
TOTALS 0 0 0 0 0
Generally the monetary base is the liability of the central bank and consists of
coin, banknotes and the balances which the banks keep with the central bank. As
regards bank deposits, it is outside the scope of the present work to discuss whether
only demand deposits or also time deposits (and other types of deposits or financial
assets) are to be considered as money. In any case what is not included as money
here, comes under the heading of the national (or domestic) securities item, which
includes, besides securities proper, any form of marketable debt instrument. As
regards joreign money, we do not make the distinction between monetary base
and bank deposits, since all the foreign sectors (including the central bank and the
banking sector) have been consolidated into a single sector (the rest-of-the-world
or foreign sector), and for residents foreign exchange in either of its forms (cash or
deposits) is foreign money. There are, finally, joreign securities, for which similar
considerations hold as for domestic ones.
The table can be read along either the rows or the columns: in any case, as
this is an accounting presentation, the algebraic sum of the magnitudes in any row
is zero, as is the algebraic sum of the magnitudes in each column. It should be
5.3. Some Important Accounting Relations 67
which states that the algebraic sum of the private sector's excess demand for goods
and services (measured by the difference between investment and saving) plus the
public sector's excess demand (measured by the difference between government
expenditure and revenue, the latter being measured by taxes net of transfer pay-
ments) plus the foreign sector's excess demand (measured by the difference between
exports and imports of goods and services plus unilateral transfers) must be zero.
If we recall that C + S = Yd by definition, where C, Yd are the private sector's
consumption and disposable income, it follows that 1 - S = (1 +C) - Yd. Therefore,
(I - S) is indeed the private sector's excess demand for goods and services, while
(G - T) is the excess demand of the government (budget deficit), and CA the
excess demand of the rest of the world (if we neglect unilateral transfers, this is
net exports). These excess demands can be either positive or negative: a negative
excess demand of the foreign sector, for example, means that the current account
balance shows a deficit.
One of the implications of (5.2) is that it is not possible for the three excess
demands to have the same sign: for example, it is not possible simultaneously
to have a positive excess demand of the private sector (1 > S), a budget deficit
(G > T) and a current account surplus (CA> 0).
If we remember that the private sector's disposable income is given by national
income minus taxes net of transfer payments, i.e. Yd = Y - T, whence S = Y - T-
68 Chapter 5. The Balance of Payments
(5.5)
which states that the difference between private saving and investment gives rise
to a change (ß) in the stock of private net financial wealth where Wp (money,
bonds, etc.). An excess of saving over investment means that the private sector
spends less than its disposable income, hence an increase in its wealth (ßWp > 0),
while ß Wp < 0 means that the private sector spends more than its disposable
income. In fact, given that (see above) S = Yd - C = Y - T - C, and defining a
new aggregate called "absorption" of the private sector Ap as the sum C + I, so
that S - I = Yd - A p , we can rewrite (5.5) as
(5.6)
namely the change in the private sector's stock of wealth equals the difference
between disposable income and absorption. This also called the net acquisition
of jinancial assets (NAFA) or the jinancial surplus (that of course can be either
positive or negative) of the private sector.
The government budget constraint tells us that
G-T=-ßN9, (5.7)
which states that the government budget deficit (excess of expenditure over re-
ceipts) is financed by issuing government bonds (a negative excess demand, that
is an excess supply, equal to -ßN9). This embodies the assumption that the
monetary financing of the public deficit is forbidden.
The various row and column constraints can be combined to derive other mean-
ingful constraints. Taking for example the last column (the rest-of-the-world bud-
get constraint), using the fifth and sixth rows, and rearranging terms, we get
5.3. Some Important Accounting Relations 69
Equation (5.8) is simply the expression of the ovemll balance 01 payments (B)
already examined in Sect. 5.2. In fact, CA is the current account balance and the
expression in braces is the autonomous-capital balance, consisting of the change in
domestic assets (deposits, !:l.D j , and securities, !:l.Nj) owned by non residents plus
the change in foreign assets (money, !:l.R, and securities, !:l.F) owned by residents,
who are subdivided into private sector (hence !:l.Rp + !:l.Fp ) and banking sector
(hence !:l.~ + !:l.Fb); note that the minus sign before the square bracket reflects
the accounting convention illustrated in Sect. 5.1. The offsetting item is given
by the change in official international reserves (!:l.R) in the wide sense, subdivided
into liquid assets (foreign money, !:l.R,;) and medium/long term assets (foreign
securities, !:l.Fe) owned by the central bank. Hence the identity
B=!:l.R, (5.9)
which states that the overall balance of payments coincides with the change in
international reserves.
Another derived identity is obtained aggregating the columns "Banking" and
"Central Bank" , which gives the budget constraint of the aggregate banking sector
Let us now observe that from the second row of Table 5.1 we have
Therefore, if we define
namely the overall balance of payments, implies a net change in the domestic
quantity of money. lf the monetary authorities wish to offset such a change (i.e.,
to "sterilize" the external component of liquidity, namely the effect of the balance
of payments on the money supply), they can act on the internal component of
liquidity, for example by open market operations. In terms of accounting identities,
this means that the monetary authorities can act on ßQ in such a way that
ßR+ßQ=O.
We would like to emphasize, in conclusion, that all these relations, being mere
accounting relations, are always valid ex post, but cannot tell us anything on the
causal relations between the variables considered, which require the consideration
of behavioural functions of the various agents.
The data contained in the publications of the IMF are also available on line in
the IMF's site (http://imf.org) against payment of a fee. Other sources of online
data at the international level are:
-the World Bank (http://worldbank.org), free
-the World Trade Organization (http://wto.org), free
-the Organization for Economic Cooperation and Development
(http://oecd.org), against payment of a fee
-the Bank for International Settlements (http://bis.org), free
-the statistical office of the European Union (http://europa.eu.int/eurostat),
partly free
-the European Central Bank (http://ecb.int), free
Part 11
In this chapter we shall deal with the adjustment processes of the balance of
payments based on flows. We shall first examine the balance on goods and
services, then introducing capital movements. In Sections 3.1 through 3.5,
"balance of payments" is synonomous with "balance on goods and services" .
Px
71"=-, (6.1)
rpm
where Px represents export prices (in terms of domestic currency), Pm import
prices (in terms of foreign currency), and r the nominal exchange rate of the
country under consideration. The meaning of 71" has already been examined
in Eq. (2.4).
The idea behind this adjustment process is that a change in the relative
price of goods, ceteris paribus, brings about a change in the demands for
the various goods by both domestic and foreign consumers, thus inducing
changes in the fiows of exports and imports which will hopefully adjust a
disequilibrium in the payments balance considered.
The terms of trade may vary both because of a change in the prices Px and
Pm expressed in the respective national currencies, and because r changes.
The analysis with which we are concerned in this chapter focuses on the
73
74 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
_ b.m/m b.m r
TJm = - b.r/r = - b.r m' (6.2)
where b. as usual denotes a change, and the minus sign before the second
fraction serves to make it a positive number (b.m and b.r have, in fact,
opposite signs because of what we said at the beginning, so that the fraction
by itself is negative).
Since each country normally records its balance of payments in terms of
domestic currency, we consider the payments balance in domestic currency
(6.3)
where the value of imports in terms of foreign currency (Pm, we remember,
is expressed in foreign currency) has to be multiplied by the exchange rate
to transform it into domestic currency unitsj as Px is expressed in terms of
domestic currency, the value of exports, PxX, is already in domestic currency
units.
To examine the effects of a variation in the exchange rate, let us consider
a depreciation by a small amount, say b.r. Correspondingly, exports and
imports change by b.x > 0 and by b.m < O. The new value of the balance of
payments is then
(6.4)
and by subtracting the previous value B as given by Eq. (6.3) we obtain the
change in the balance of payments
Since b.r, b.m are small magnitudes, their product is of the second order of
smalls and can be neglected. Collecting Pmmb.r we obtain
We now multiply and divide the first fraction in square brackets by r/x,
whence
It follows that
Since Ilr > 0, the sign of IlB will depend on the sign of the expression in
square brackets, hence an exchange-rate depreciation (Ilr > 0) will improve
the balance of payments (IlB > 0) when
Pxx
--""x+""m> 1. (6.6)
rpmm
If we consider a situation near equilibrium, such that PxX '::= rpmm, the
condition becomes
""X +""m > 1, (6.7)
namely a depreciation will improve the balance of payments if the sum of
the export and import elasticities is greater than unity. Inequality (6.7) is
variously called the Marshall-Lerner condition 1 , or the Bickerdicke-Robinson
condition, or the critical elasticities condition.
Let us note that a depreciation is unlikely to occur when the balance
of payments is near equilibrium. On the contrary, it will normally occur
when the balance of payments shows a deficit (which is the case in which an
exchange-rate depreciation normally comes ab out , either spontaneously or
through discretionary intervention by the monetary authorities). A deficit
means rpmm > PxX, and we should apply the more general condition (6.6),
which is more stringent than (6.7). In fact, since pxx/rpmm < 1, the value of
""X will be multiplied by a factor smaller than one, hence it may well happen,
if the sum of the elasticities is just above unity and the deficit is huge, that
condition (6.6) is not satisfied.
In order to translate this analysis into practical policy recommendations
(i.e., that a way of coping with a trade-balance deficit is to devalue the ex-
change rate, or let it depreciate) we must be sure that the critical elasticities
condition is satisfied. In the past there was a heated debate between elasticity
IThe most frequent denomination is, by far, Marshall-Lerner condition. However, some
authors maintain that such adenomination is wrong: see, e.g., MundeIl (2001), Sect. VI;
Gandolfo, 2002, p. 24.
6.1. The Elasticity Approach 77
pessimism (elasticities are too low) and optimism (elasticities are sufficiently
high). Recent studies (Hooper et al., 2000; Gagnon, 2003) show that price
elasticities for imports and exports generally satisfy the Marshall-Lerner con-
dition.
BOX 6.1 Does a devaluation help? The J curve
In November 1967, the pound sterling was devalued (from 2.80 dollars per pound
to 2.50 dollars per pound) due to balance-of-payments difficulties. According to the
estimates of the time, the elasticities were sufficiently high to satisfy the Marshall-
Lerner condition. However, the devaluation was followed bya trade deficit which
lasted untill970. The terminology J curve was coined in relation to this phenomenon
(NIESR 1968, p. 11). The J curve can be explained by introducing adjustment lags,
and, more precisely, by distinguishing various periods following the devaluation in
which the effects of the devaluation itself take place. These are, in the terminology
of Magee (1973), the currency-contmct period, the pass-through period, and the
quantity-adjustment period.
The currency-contmct period is defined as that short period of time immediately
following the exchange-rate variation in which the contracts stipulated before the
variation mature. During this period both prices and quantities are predetermined.
Normally both the export and import contracts stipulated before the devaluation
are expressed in foreign currency. In fact, in the expectation of a possible devalua-
tion, both domestic and foreign exporters will try to avoid an exchange-rate loss by
stipulating contracts in foreign currency. Now, as a consequence of the devaluation,
the domestic-currency value of both imports and exports will increase by the same
percentage as the devaluation, so that as the pre-devaluation value of imports was
higher than that of exports, the deficit will increase.
The pass-through period is defined as that short period of time following the
exchange-rate variation in which prices can change (as they refer to contracts agreed
upon after the exchange rate has varied), but quantities remain unchanged due to
rigidities in the demand for and/or supply of imports and exports. Consider, for
example, the case of a devaluation with a demand for imports by residents of the de-
valuing country and a demand for the devaluing country's exports by the rest of the
world which are both inelastic in the short-run. The domestic-currency price of im-
ports increases as a consequence of the devaluation but the demand does not change,
so that the outlay for imports increases. The foreign-currency price of exports de-
creases as a consequence of the devaluation, but the demand does not change, so
that the foreign-currency receipts will decrease and their domestic-currency value
will not change. Therefore the domestic-currency balance deteriorates (again a per-
verse effect).
Finally, we come to the quantity-adjustment period, in which both quantities and
prices can change. Now, if the suitable conditions on the elasticities are fulfilled, the
balance of payments ought to improve. This is undoubtedly true from the viewpoint
of comparative statics, but from the dynamic point of view it may happen that
quantities do not adjust as quickly as prices, owing to reaction lags, frictions etc.,
so that even if the stability conditions occur the balance of payments may again
deteriorate before improving towards the new equilibrium point.
In addition to these lags, there may be other elements which cause the devaluation
not to be fully passed through to prices. For example, in an imperfectly competitive
setting, part of the devaluation may be absorbed (in the short run) by foreign
producers and domestic importers (in order to avoid losing market shares), so that
the domestic price of imports rises by less than the percentage of the devaluation.
The J-curve phenomenon is by now a weil established fact: see, e.g., Bahmani-
Oskooee et a1. (1999, 2003), Hacker and Hatemi (2003), LaI (2002).
78 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
The equations represent, in this order: the consumption function (Co is the
autonomous component, b is the marginal propensity to consume, and y is
national income), the investment function (the autonomous cornponent is [0,
and h is the marginal propensity to invest), the import function (mo is the
6.2. The Multiplier Approach 79
S+m=1+x, (6.13)
which is the extension to an open economy of the well-known S = 1 closed-
economy condition. From (6.13) we obtain
that is, the excess of exports over imports is equal to the excess of saving
over investment, namely the excess of imports over exports is equal to the
excess of investment over saving.
Equations (6.8)-(6.12) form a complete system by means of which the
foreign multiplier can be analyzed. Since, however, we are interested in
balance-of-payments adjustment, we add the equation which defines the bal-
ance of payments B (since prices and the exchange rate are rigid, they can
be normalized to one):
B=x-m. (6.16)
The problem we are concerned with is to ascertain whether, and to
what extent, balance-of-payments disequilibria can be corrected by income
changes. Suppose, for example, that a situation of equilibrium is altered by
an increase in exports, so that B shifts to a surplus situation. What are
the (automatie) corrective forces that tend to re-equilibrate the balance of
payments? The answer is simple: via the multiplier the increase in exports
brings about an increase in income, which in turn determines an induced
increase in imports via the marginal propensity to import. This increase in
80 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
imports tends to offset the initial increase in exports, and we must ascertain
whether the former increase exactly matches the latter (so that the balance of
payments returns to an equilibrium situation) or not. In the second case the
situation usually depicted is that the balance of payments will show a sur-
plus, although smal1er than the initial increase in exports: in other words, the
induced change in imports will not be sufficient to re-equilibrate the balance
of payments. However, we shal1 see presently that the contrary case (that is,
when induced imports increase more than the initial increase in exports) as
well as the borderline case cannot be excluded on apriori grounds.
To rigorously analyse these and similar problems the first step is to find
the formula for the multiplier. If we substitute from Eqs. (6.8)-(6.11) into
Eq. (6.12) and solve for y, we obtain
1
y= 1 b h (Co+lo-mo+xo), (6.17)
- - +J.L
where of course
1-b-h+J.L>0 (6.18)
for the solution to be economically meaningful. If we then consider the
variations (denoted by 6.), we get
1
6.y = 1- b - +J.L
h(6.Co + 6.10 - 6.mo + 6. xo). (6.19)
contains both national and foreign goods. The part pertaining to the latter
is thus apart of the total increment b + h, and coincides, in our pure flow
model, with J.L.
Algebraically, let the subscripts d and f denote domestic and foreign
goods (and services) respectively; then
b = bd + bf , (6.22)
h hd + hf , (6.23)
(6.24)
Therefore
(6.25)
(6.26)
which states that the change in the balance of payments is equal to the
exogenous change in exports minus the change in imports, the latter being
partly exogenous (~mo) and partly induced (J.L~Y, where ~y is given by the
multiplier formula (6.19) found above).
Here we have all that is needed to analyse the balance-of-payments ad-
justment problem. Consider for example the case of an exogenous increase
in exports.
By assumption, no other exogenous change occurs, so that ~Co = ~Io =
~mo = 0, and the equations of change become
1
~B = ~xo - J.L~Y, ~y = 1- b h
- +J.L
ßXo,
whence
1
ßB ~xo - J.L b h ~xo
1- - +J.L
I-b-h
1 - b - h +J.L ~xo, (6.27)
82 Chapter 6. The Basic Models: Elasticities, Multiplier, MundeIl-Fleming
x-mt
S-1
Figure 6.1: Exogenous increase in exports, the multiplier, and the balance of
payments
It can be seen that the various macroeconomic variables are functions of both
national income and the exchange rate (except for exports, which depend,
in addition to the exchange rate, on foreign income, here taken as exogenous
thanks to the SOE assumption).
The main result of this model is that the critical elasticities condition,
although necessary, is not sufficient to ensure the adjustment of the balance
of payments: for an exchange rate depreciation to improve the balance of
payments the sum of the elasticities must be greater than a critical value
which is greater than one.
The economic reason for this result is intuitive. Let us assume, for exam-
pIe, that the balance of payments is in deficit. The exchange rate depreciates
and, assuming that the traditional critical elasticities condition occurs, the
deficit is reduced. However, the depreciation increases total demand for da-
mestic output; this causes an increase in income and so imports increase,
thus opposing the initial favourable effect of the depreciation on the balance
of payments: this effect must therefore be more intense than it had to be in
the absence of income effects, Le. the sum of the elasticities must be higher
than in the traditional case.
The model under consideration still remains in the context of current
account adjustment. Given the ever increasing importance of capital move-
6.4. The Mundell-Fleming Model 85
ments, we now turn to a model in which capital flows are explicitly intro-
duced.
M = L(y,i), (6.30)
where M indicates the money stock and L the demand for money, that
depends positivelyon income and negativelyon the interest rate.
The balance of payments includes not only imports and exports of goods,
but also capital movements. Net capital flows (inflows less outflows) are
expressed as an increasing function of the interest differential. It is in fact
clear that the greater the domestic interest rate with respect to the foreign
rate, the greater, ceteris paribus, will be the incentive to capital inflows and
the less to outflows, as we have seen in Chap. 4. Sinee this is a one-country
model, the foreign interest rate is exogenous, so that the movement of capital
is ultimately a function of the domestic interest rate). We can therefore write
the following equation for the equilibrium in the balance of payments
°
where K(i) ~ indicates the net private capital inflow (outflow). Let us
note, in passing, that the condition that the overall balance of payments is
in equilibrium (external equilibrium) is equivalent to the condition that the
stock of international reserves is stationary, as shown in Sect. 5.3.
The system, composed of the three equations studied so far, is determined
if the unknowns are also three in number: it is therefore necessary to consider
M also as an unknown, in addition to y and i. We shall discuss this fact at
length in Sect. 6.4.1.2.
It is convenient at this point to pass to a graph of the equilibrium condi-
tions, which will be of considerable help in the subsequent analysis.
s
y
Figure 6.2: Mundell-Fleming under fixed exchange rates: the real equilibrium
schedule
Furthermore, the I S curve has the property that at all points above it
there will be a negative excess demand, while at all points below it the excess
demand for goods will be positive.
Consider in fact any point A' above the I S curve. Here the rate of interest
is greater than at point A, while income is the same. Point A is a point of
real equilibrium, being on the I S schedule. If, income being equal, the rate
of interest is greater, domestic demand will be less, so that at A' there will
be negative excess demand (excess supply). In the same way, it can be
demonstrated that at A" there is positive excess demand.
Let us now examine the monetary equilibrium schedule. Given that the
demand for money is an (increasing) function of income and a (decreasing)
function of the interest rate, there will be a locus of the combinations of these
two variables which make the total demand for money equal to the supply,
which is represented by the familiar schedule, LM in Fig. 6.3.
The LM curve is increasingj in fact, given a certain supply of money, if
income is higher the demand for money will also be higher: in consequence,
it is necessary to have a higher value for the interest rate, so as to reduce
the demand for money itself, to maintain the equality between demand and
supply. Furthermore the LM schedule has the property that, at all points
above it, there is negative excess demand for moneyj while at all points below
it, there is positive excess demand. Consider for example any point above
LM, for example, A'j there the rate of interest is higher, income being equal,
than at point A on LM. At A', therefore, there is a lower demand for money
than at A. Since at A the demand for money equals the supply and given
that at A' the demand for money is lower than at A, it follows that at A'
the demand for money is less than the supply. In the same way, it is possible
to show that at any point below the LM schedule (for example at A") the
88 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
-------------,A'
I
Figure 6.3: Mundell-Flerning under fixed exchange rates: the monetary equi-
librium schedule
B'
Figure 6.5: Mundell-Fleming under fixed exchange rates: the external equi-
librium schedule
balance of payments occurs when the algebraic sum of the current account
balance and the capital movements balance is nil. As exports have been
assumed exogenous and imports a function of incomeand the interest rate,
and the capital movements balance as a function of the interest rate, it is
possible to show in a diagram all the combinations of the interest rate and
income which ensure balance-of-payments equilibrium, thus obtaining the
BB schedule (Mundell called it the F F schedule, but there is no standardized
denomination: BB, BP, FX, NX are all used in the literature).
This schedule is upward sloping: in fact, as exports are given, greater
imports correspond to greater income and therefore it is necessary to have
a higher interest rate (which tends on the one hand to put a brake on the
increase in imports and on the other to improve the capital account) in order
to maintain balance-of-payments equilibrium.
We observe that the slope of BB depends, ceteris paribus, on the respon-
siveness of capital fiows to the interest rate, Le, on the degree of international
capital mobility. The higher the mobility of capital, the fiatter the BB sched-
ule. In Fig. 6.5, the degree of capital mobility underlying schedule B' B' is
higher than that underlying schedule BB. In fact, if consider for example
the external equilibrium point H, a higher income (for example, Y2 instead of
Yd will mean a balance-of-payments deficit. This requires-as we have just
seen-a higher value of the interest rate to maintain external equilibrium.
Now, the more reactive capital fiows to the interest rate, the lower the re-
quired interest rate increase. Given Y2 the interest rate will have to be i 2 in
the case of BB, and only i3 in the case of B' B', In the limit, namely in the
case of perfect capital mobility, the BB schedule will become a horizontal
straight line parallel to the Y axis.
Furthermore, the BB schedule has the property that at all points above it
there is a surplus in the balance of payments, while at all points below it there
is a deficit. In fact, consider any point A' above the line BB. At A', while
90 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
income is the same, the interest rate is greater than at A, where the balance
of payments is in equilibrium. Thus, as imports are a decreasing function of
the interest rate and as the capital account balance is an increasing function
of the rate itself, at A' imports will be lower and the capital account balance
will be higher-ceteris paribus-than at A, so that if at A the balance of
payments is in equilibrium, at A' there must be a surplus.
In the same way, it is possible to show that at all points below BB (see,
for example, point A") there will be a balance-of-payments deficit.
y y
a) b)
(ii) if, on the other hand, the quantity of money is considered to be vari-
able, then, in principle, it is always possible to find a value for the quantity of
money itself such that the corresponding LM schedule will also pass through
point E. In this case-see Fig. 6.6b-we have the simultaneous occurrence
of real, balance of payments, and monetary equilibrium.
But, one might ask, do any forces exist which tend to cause the necessary
shifts in the LM schedule? The answer to this question cannot be given in
isolation, but requires a general analysis of the dynamics of the disequilibria
in the system, that is, of the behaviour of the system itself when one or more
of the equilibrium conditions are not satisfied. For this purpose it is necessary
to make certain assumptions regarding the dynamic behaviour of the relevant
variables: money supply, income and interest rate. The assumptions are as
follows:
(a) the money supply varies in relation to the surplus or deficit in the
balance of payments and, precisely, increases (decreases) if there is a surplus
(deficit). This assumption implies that the monetary authorities do not in-
tervene to sterilize (see Sect. 5.3) the variations in the quantity of money
determined by disequilibria in the balance of payments: in fact, given Eq.
(5.10), we have b.M = B when the other items are set to zero.
(b) Income varies in relation to the excess demand for goods and, to be
exact, it increases (decreases) according to whether this excess demand is
positive (negative). This is the assumption usually made in the context of
Keynesian-type models.
(c) The rate of interest varies in relation to the excess demand for money
and, more precisely, it increases (decreases) if this excess demand is positive
(negative). This is a plausible hypothesis within the context of an analysis
of the spontaneous behaviour of the system. In fact, if the interest rate is,
in a broad sense, the price of liquidity, an excess demand for liquidity causes
an increase on the market in this price and vice versa. The mechanism,
commonly described in textbooks, is the following: an excess demand for
money-that is, a scarcity of liquidity-induces holders of bonds to offer
them in exchange for money: this causes a fall in the price of bonds, and
thus an increase in the interest rate (which is inversely related to the price
of bonds).
Having made these behavioural hypotheses, it will be seen that the system
is stable and will therefore tend to eliminate disequilibria, that is to say, it
will tend to reach the point of simultaneous real, monetary and balance-
of-payments equilibrium, if the marginal propensity to domestic expenditure
is less than one (as already previously assumed) and ij, in addition, the
marginal propensity to import is below a certain critical value.
A simple case of disequilibrium and the related adjustment process can
be analysed intuitivelyon the basis of Fig. 6.7. Assume, for example, that
the system is initially at point A. At that point there is real and monetary
equilibrium, but not equilibrium of the balance of payments: more exactly,
92 Chapter 6. The Basic Models: Elasticities, Multiplier, Mundell-Fleming
i
B
o y
for money remains, with a further tendency for the interest rate to increase
(vertical arrow rising above Ad. We thus have a situation in which y and i
approach their respective equilibrium values and also a shift of LM toward
the position LEME.
We can now ask ourselves what the economic meaning of the conditions
of stability might be. As far as concerns the condition that the marginal
propensity to domestic expenditure is less than one, the meaning is the usual
one: an increase in income, due to a positive excess demand, causes a further
increase in domestic demand, but the process is certainly convergent, if the
increases in demand are below the increases in income, that is to say, if the
marginal propensity to domestic expenditure is less than one. In the oppo-
site case, the process could be divergent, unIess other conditions of stability
intervene.
With regard to the condition that concerns the marginal propensity to
import, the meaning is as folIows: if this propensity is too great, it may hap-
pen, in the course of the adjustment process, that the reduction in imports
(induced by the reduction in y and the increase in i) is such as to bring the
balance of payments into surplus (that is to say, point Eis passed). An 00-
justment in the opposite direction is then set into motion: the money supply
increases, the rate of interest drops, domestic demand, income and imports
all increase (both because y has increased and i decreased). And if the
marginal propensity to import is too high, then it may be that the increase
in imports is such as to produce a new deficit in the balance of payments. At
this point, a new process comes into being, working in the opposite direction
and so on, with continual fluctuations around the point of equilibrium which
each time may take the system further away from equilibrium itself.
coupled with an increase in the interest rate that causes a capital outflow
will do the job.
Since nowadays capital mobility is practically perfect in developed coun-
tries, the model's results on the effectiveness of the various policies under
fixed and flexible exchange rates with perfeet capital mobility are particu-
larly relevant.
It should at this point be recalled from Sect. 4.6 that Munden implicitly
assumed perfect capital mobility to imply perfect asset substitutability, so
that VIP holds, ih = if + (f - r)/r. Prom this it is possible to obtain
the condition, used by Munden, that the domestic and foreign interest rate
coincide, i h = if> assuming that, if we are under fixed exchange rates, the
given exchange rate is supposed to be credible (so that agents expect no
change), while in the case of flexible exchange rates static expectations are
held by the representative agent. In the case of perfeet capital mobility it
was shown by Munden that fiscal policy becomes completely ineffective under
flexible exchange rates while monetary policy is fully effective. This result
is symmetrie to that of the complete ineffectiveness of monetary policy and
fuH effectiveness of fiscal policy under fixed exchange rates.
M
I ",M'
/
/
if
B B
YE y' y
Figure 6.8: Perfect capital mobility and fiscal and monetary policy under
fixed and flexible exchange rates
Let us examine Fig. 6.8, where we have drawn the 18, BB, LM sched-
ules corresponding to the equilibrium value (rE) of the exchange rate. This
diagram is similar to Figs. 6.6(b) and 6.7, except for the fact that the BB
schedule has been drawn parallel to the y axis with an ordinate equal to i f, to
denote that the domestic interest rate cannot deviate from the given foreign
interest rate (if) owing to the assumption of perfect capital mobility. In the
6.4. The Mundell-Fleming Model 97
with it the burden of interest payments (a debit item in the current account)
will also increase.
If we also consider the previous point, it follows that the capital inflow can
be maintained only through an increasing interest differential, which makes
the burden of interest payments still more serious. This casts serious doubts
on the validity of the use of an expansionary fiscal policy coupled with a
restrictive monetary policy to solve the dilemma cases under fixed exchange
rates. Such a use can be at best considered as a short-run measure.
In the next chapter we will tackle the problems deriving from the fact that
capital movements are actually a stock adjustment (portfolio approach).
101
102 Chapter 7. The Monetary and Portfolio Approaches
in relative prices.
Be it as it may, we now turn to summarize the MABP in a few basic
propositions, from which certain implications for economic policy can be
derived.
CA=Y-A.
in the stock of assets owned by the public: in fact, this divergence is the
equivalent of one between saving and investment and therefore, given the
balance constraint of the private sector (see Sect. 5.3), it is translated into a
variation in the stock of assets held by this sector. If, for the sake of simplicity,
we introduce the hypothesis that the only asset is money, a variation in the
money stock comes ab out , which in turn, given (7.1), coincides with the
overall balance of payments.
Ultimately what has happened, through this sequence of effects (for which
we shall give a model below) is that an excess demand or supply of money, by
causing an excess or deficiency of absorption with respect to national income
(product), has been unloaded onto the balance of payments: an excess of
absorption means a balance-of-payments deficit (the only way of absorbing
more than one pro duces is to receive from foreign countries more than one
supplies to them) and a deficiency in absorption means a balance-of-payments
surplus. In other words, if the public has an excess supply of money it gets
rid of it by increasing absorption and, ultimately, by passing this excess on
to foreign countries in exchange for goods and services (balance-of-payments
deficit). If on the other hand the public desires more money than it has
in stock, it procures it by reducing absorption and, ultimately, it passes
goods and services on to foreign countries in exchange for money (balance-
of-payments surplus).
What is implicit in our reasoning is the hypothesis that the level of prices
and the level of income are a datum (otherwise the variations in absorption
with respect to income could generate variations in prices and/or income)
and in fact this is what, among other things, the next propositions refer to.
Proposition II
There exists an efiicient world market for goods, services and assets and
that implies, as far as goods are concerned (as usual, goods refers to both
goods and services), that the goods themselves must have the same price
everywhere (law of one price), naturally account being taken of the rate of
exchange (which is, by hypothesis, fixed) and therefore that the levels of
prices must be connected-if we ignore the cost of transport-by the rela-
tionship
(7.2)
where PI is the foreign price level expressed in foreign currency, pis the da-
mestic price level in domestic currency, and r is the exchange rate. Equation
(7.2), which is also called the equation of purchasing power parityl, has the
effect that, given r and PI, P will be fixed.
Proposition III
Production is given at the level of fuH employment. This proposition
implies that the MABP is a long-term theory, in which it is assumed that
lWe shall deal more fully with the purchasing power parity later, among other things
so as to examine the reasons for deviations from it. in Sect. 9.1.
104 Chapter 7. The Monetary and Portfolio Approaches
production tends toward the level of fuH employment thanks to price and
wage adjustments.
or
(7.3)
The parameter a is a coefficient which denotes the intensity of the effect on
absorption of a divergence between M and L. It is assumed positive but
smaller than unity because the divergence is not entirely eliminated within
one period, due to lags, frictions and other elements.
Equation (7.3) is the crucial assumption of the MABP, and represents
expenditure as coming out of a stock adjustment rather than deriving from a
relation between ßows as in the functions in the Keynesian tradition encoun-
tered in the previous chapter. Here the representative agent does not directIy
decide the ßow of expenditure in relation to the current ßow of income, but
does instead decide the ßow of (positive or negative) saving out of current
income in order to bring the current stock of wealth (here represented by
money) to its desired stock (here represented by money demand).
We then have the accounting equation (see Sect. 5.3)
(7.4)
which expresses the fact that the excess of income over expenditure coincides
with an increase in the money stock held by the public, if we assume that
money is the sole asset in existence. This same assumption allows us to write
the accounting relation (7.1) as
which says that the variation in reserves (coinciding with the balance of
payments) depends, through coefficient a, on the divergence between the
demand and supply of money.
The self-correcting nature of the disequilibria can be clearly seen from Bq.
(7.6): if, for example, the existing money stock is excessive, that is if M > L,
there is a balance-of-payments deficit (B = 6R is negative) and therefor~
via Eq. (7.5)-M decreases. This reduction restores the equilibrium between
monetary stocks (M ~ L) and hence brings the balance of payments back
to equilibrium.
the system to its own devices (Implication II) for everything to be automat-
ically adjusted.
It is not surprising therefore that the MABP gave rise to a large num-
ber of criticisms and rejoinders. Most of this debate, however, did not get
the gist of the matter, which is the validity of the behavioural assumption
(7.3). If it is valid, the MABP is perfectly consistent, since other contro-
versial points (such as the neglect of financial assets other than money, the
constancy of price levels, etcetera) are really not essential for its conclusions.
If, on the contrary, the representative agent behaves according to a pure flow
expenditure function A = A(y, ... ), then the MABP conclusions are clearly
unwarranted.
Be it as it may, the MABP should be given due acknowledgement for
one fundamental merit: that of having directed attention to the fact that
in the case of balance-of-payments disequilibria and the related adjustment
processes, stock equilibria and disequilibria must be taken into account. Nat-
urally, this must not be taken in the sense-typical of the cruder versions
of the MABP-that only stock equilibria and disequilibria matter, but that
they also matter in addition to pure ftow equilibria and disequilibria.
simplicity that the risk element and the utility function do not undergo any
variations, we have the functions
(7.8)
where ih and i f indicate as usual the horne and foreign interest rates respec-
tively. The three functions (7.8) are not independent of each other insofar as
once two of them are known the third is also determined given the balance
constraint (7.7).
It is then assumed that these functions have certain plausible properties.
The signs under the explanatory variables express the eifect of each such
variable on the dependent variable. First of all, the fraction of wealth held in
the form of money is a decreasing function of the yields of both national and
foreign bonds: an increase in the interest rates ih and i f has, other things
being equal, a depressive eifect on the demand for money and, obviously,
an expansionary eifect on the demand for bonds (see below). Also, f is an
increasing function of y and this means that, on the whole, the demand for
bonds is a decreasing function of y.
The fraction of wealth held in the form of domestic bonds, on account of
what has just been said, is an increasing function of i h ; it is, furthermore, a
decreasing function of i f insofar as an increase in the foreign interest rate will
induce the holders of wealth to prefer foreign bonds, ceteris paribus. Similarly
the fraction of wealth held in the form of foreign bonds is an increasing
function of if and a decreasing function of i h . Finally the fraction of wealth
held totally in the form of bonds, (N + F) /W, is-for the reasons given
above--a decreasing function of y.
We could at this point introduce similar equations for the rest of the
wor1d, but so as to simplify the analysis we shal1 make the assumption of
a smal1 country and thus use a one-country model. This implies that the
foreign interest rate is exogenous and that the variations in the demand for
foreign bonds on the part of residents do not influence the foreign market
for these bonds, so that the (foreign) supply of forE:lign bonds to residents is
perfect1y e1astic. Another implication of the small-co~try hypothesis is that
non-residents have no interest in holding bonds from this country, so that
capital flows are due to the fact that residents buy foreign bonds (capital
outflow) or sell them (capital inflow).
Having made this assumption, we now pass to the description of the asset
market equilibrium and introduce, alongside the demand functions for the
various aSsets, the respective supply functions, which we shall indicate by
M for money and NB for domestic bonds; for foreign bonds no symbol is
needed. as the hypothesis that their supply is perfectly elastic has the eifect
that the supply is a1ways equal to the demand on the part of residents. The
equilibrium under consideration is described as usual by the condition that
108 Chapter 7. The Monetary and Portfolio Approaches
M+NS+F=W. (7.10)
(7.11)
which is the formal statement of Walras' law. From Eq. (7.12) we see that,
if any two of the expressions in square brackets are zero (namely, if any two
of Eqs. (7.9) are satisfied), the third is also.
Equations (7.9) therefore provide us with two independent equations
which, together with (7.10) make it possible to determine the three un-
knowns, which were the horne interest rate (i h ), the stock of foreign bonds
held by residents (F), and the stock of domestic money (M): the equilib-
rium values of these three variables will thus result from the solution of the
problem of portfolio equilibrium, while the stock of domestic bonds (N S ) is
given as are i h , y and W.
N------~~------N
L F
F. F
i.
N--~~~~~--------N
~.
1', Fo p
Capital Movements,
Speculation, and Currency
Crises
113
114 Chapter 8. Capital Movements, Speculation, and Currency Crises
1 A portfolio, that is a given allocation of funds among the various assets, is efficient if
a greater return can be achieved only by accepting a greater risk (or a lower risk can be
obtained only by accepting a lower return).
8.1. Long-Term Capital Movements 115
a market with a high product differentiation, and, for this corporation, direct
investment is often an alternative to exporting its products, as the ownership
of plants in foreign countries facilitates the penetration of foreign markets.
From this point of view it is clear that the theory of direct investment belongs
to the theory of multinational firms (which has had an enormous development
in recent times). Abrief account of this theory will now be given.
consumption) and hence can be considered as a public input for the firm
that owns it.
b) Location advantages. With production plants located near the final
consumers, MNE save on transport costs and may hire cheap local factors of
production (for example, labour in developing countries). Besides, they can
circumvent possible barriers to trade such as tarifIs imposed by the foreign
country (tanff-jump argument). Vertical multinationals may find it optimal
to export intermediate inputs and the services of its knowledge capital to
a foreign affiliate for final assembly and shipment back to the MNE's home
country (jragmentation).
c) Internalization advantages. Ownership and location advantages could
in principle be reaped also through agreements with foreign licensees. How-
ever, the property of knowledge capital that makes it easily transferred also
makes it easily dissipated: licensees can absorb the knowledge capital and
then defect, or ruin the firm's reputation for greed. Thus MNE transfer
knowledge internally in order to maintain the value of their knowledge capi-
tal and avoid its dissipation.
Several other theoretical developments have taken place in recent years,
see Markusen (2002).
the market, but will be replaced by someone else who wishes to have a go.
The idea that professional speculators might on the average make profits and
destabilise the exchange rate while achanging body of "amateurs" regularly
loses large sums is not far from reality.
It should be stressed that it would not be correct to argue, from what we
have said, that under flexible exchange rates speculation is always destabil-
ising. It is, in fact, quite possible for speculators to behave as described by
Friedman, in which case their stabilizing effect is self-evident.
Thus we have seen that, whilst under an adjustable peg regime specula-
tion is generally destabilising, under flexible exchange rates it may have either
effect, so that the question we started from has no unambiguous answer.
(more similar to the western standard) financial system. But something went
wrong. Stabilization policy failed and did not prevent the materialization of
a deep recession.
According to Krugman and others, in any case the plan could not have
ben successful, given the features of the crisis. To abandon the parity would
have undoubtedly caused the crisis, as seen above. On the other hand, when
the leverage is high, as in the countries under consideration, the economy
can stabilize the real exchange rate only at the cost of a deep depression.
Given this dilemma, the heated debate on the role of the IMF in managing
the crisis (according to some it has well performed, according to others it has
failed) might seem pointless, because both policies (to defend the parity or
to abandon it) would have been unsuccessful.
We conc1ude by mentioning two related problems, that of indicators and
that of contagion.
The former proposes to forecast a crisis by using several macroeconomic
indicators such that, when they exceed certain threshold values, this can
be taken as a signal that a crisis is approaching. The suggested indicators
inc1ude financial indicators (the M2 multiplier, the ratio of domestic credit
to nominal GDP, the real interest rate on deposits, the ratio of lending-
to-deposit interest rates, etc.), real indicators (industrial production, equity
prices, etc.), fiscal indicators (the overall budget deficit as apercent of GDP,
the public debt/GDP ratio, etc.). It is not yet certain whether the use of
indicators can indeed help to forecast a crisis.
The term contagion means that speculative attacks and the ensuing cur-
rency crises are like infectious diseases: they tend to spread contagiously.
The channels of contagion are both commercial (dose commercial relations)
and financial. However, diseases do not only spread to disease-prone persons
hut also to healthy people: can we carry the similitude as far as to state that
speculative attacks against a misaligned currency tend to spread not only to
other misaligned currencies hut also to apparently sound currencies? This
question has enormous policy implications, because an affirmative answer
would warrant the bailout (by international organisations, other governments
or groups of governments like the G-7) of any country under speculative at-
tack, so as to prevent contagion to other (sound) countries. On this point no
definite answer yet exists.
Woods system (see Sect. 3.5.1) had already shown the unsustainability of an
adjustable peg regime. However, the corner solution has been criticized, and
a less extreme view is now gaining ground, according to which, excluding
soft pegs, a variety of exchange rate arrangements (in addition to hard pegs
and free floats) of the managed float type remains viable (Fischer, 2001).
Exchange-Rate Determination
The problem of the forces that determine the exchange rate and, in particu-
lar, its equilibrium value, is self-evident under flexible exchange rates, but is
also important under limited flexibility and even under fixed exchange rates
(if the fixed rate is not an equilibrium one, the market will put continu-
ing pressure on it and compel the monetary authorities to intervene in the
exchange market, as we have seen in Sect. 3.2.1).
A related problem is the age-old debate on fixed versus flexible exchange
rates, that will also be examined in this chapter.
125
126 Chapter 9. Exchange-Rate Determination
country, between the prices of traded and non-traded goods: the inexistence,
even in the long run, of these conditions, is a well-known fact. Besides,
the law of one price presupposes that traded goods are highly homogeneous,
another assumption often contradicted by fact and by the new theories of
international trade (see below, Chap. 17), which stress the role of product
differentiation.
The same idea of free markets, of both goods and capital, lies at the basis
of the other proposals, which run into trouble because this freedom does not
actually exist. Cassel himself, it should be noted, had alreadY singled out
these problems and stated that they were responsible for the deviations of
the exchange-rate fram PPP.
These deviations, which make the PPP theory useless to explain the be-
haviour of exchange rates in the short-run, were one of the reasons which
induced most economists to abandon it in favour of other approaches. It
should however be pointed out that the PPP theory has been taken up again
by the monetary approach (which has been dealt with in Sect. 7.1; see also
below, Sect. 9.3.2), and used as an indicator of the long-run trend in the
exchange rate.
for and supplies of foreign exchange, on the other, it does not affect the fact
that it is the interaction between these demands and supplies which actually
determines the exchange rate.
(9.2)
the new equilibrium point. This path implies an exchange rate temporarily
higher than its new long-run equilibrium value. Since the exchange rate is a
jump variable, the economy can jump from the initial equilibrium point on
the (unique) stable trajectory; after this overshooting, the exchange rate will
gradually appreciate towards its new long-run equilibrium.
ih = if
r-T
+ -T- + 15, (9.4)
(9.5)
where the demands are expressed, in accordance with portfolio selection the-
ory,as
'
h( zh-zf r --T)W,
. - -
_T (9.6)
T-T
g(i h - if - --)W,
T
(9.7)
132 Chapter 9. Exchange-Rate Determination
and so, by substituting into (9.6) and dividing the second by the first equation
there, we obtain
rFS .. r-r
Ns = <p(Zh - zf - -r-)' (9.8)
where <p( ... ) denotes the ratio between the g( ... ) and h( ... ) functions. From
Eq. (9.8) we can express the exchange rate as a function of the other variables
(9.9)
Equation (9.9) shows that the exchange rate can be considered as the
relative price of two stocks of assets, since it is determined-given the interest
differential corrected for the expectations of exchange-rate variations-by the
relative quantities of NS and PS.
The basic idea behind all this is that the exchange rate is the variable
which instantaneously adjusts so as to keep the (international) asset markets
in equilibrium. Let us assurne, for example, that an increase occurs in the
supply of foreign bonds from abroad to domestic residents (in exchange for
domestic currency) and that (to simplify to the utmost) expectations are
static (r - r = 0). This increase, ceteris paribus, causes an instantaneous
appreciation in the exchange rate.
To understand this apparently counter-intuitive result, let us begin with
the observation that, given the foreign-currency price of foreign bonds, their
domestic-currency price will be determined by the exchange rate. Now, res-
idents will be willing to hold (demand) a higher amount of foreign bonds,
ceteris paribus, only if the domestic price that they have to pay for these
bonds (i.e., the exchange rate) is lower. In this way the value of rF d = rFs
remains unchanged, as it should remain, since all the magnitudes present on
the right-hand side of Eqs. (9.6) are unchanged, and the market for foreign
assets remains in equilibrium (F d = PS) at a higher level of Fand a lower
level of r.
, ,
CA
'0 ,
-~,
B
I
1
--------'0 -- ,
I
I I
-+---------'E __ L_
I 1
I I
I I
I
t I
-
1 I
1 I
I 1-
0 C + 0 Fo FE F
To illustrate this mechanism we can use Fig. 9.1, where the left-hand
panel shows the current account (CA) as a nmction of the exchange rate (the
2 As we know from balance-of-payments accounting, the algebraic sum of the current
account and the capital account is necessarily zero (see Sect. 5.1). Here the implicit as-
sumption is that there are no compensatory capital movements, Le. that there is no official
intervention (the exchange rate is perfectly and freely flexible). Under this assumption all
capital movements are autonomous and originate from private agents.
134 Chapter 9. Exchange-Rate Determination
positive slope means, as we have said, that the critical elasticities condition
occurs); the right-hand panel shows the relation between r and F, which is
a rectangular hyperbola as its equation is (9.10).
The long-run equilibrium corresponds to the exchange-rate rE, where the
current account (and so the capital account) is in equilibrium. Let us assume,
for example, that the exchange rate happens to be ro, hence a current account
surplus oe; the initial stock of foreign assets (which corresponds to ro) is
Fo. The current-account surplus is matched by a capital outfiow, i.e. byan
increase in foreign assets; thus the residents stock of foreign bonds increases
(from Fo towards FE, which is the equilibrium stock) and so the exchange
rate appreciates (point A moves towards point E). This appreciation reduces
the current-account surplus; the process goes on until equilibrium is reached.
The case of an initial exchange rate lower than rE is perfectly symmetrical
(current-account deficit, decrease in the stock of domestically held foreign
assets, exchange rate depreciation, etc.).
Although this is a highly simplified model, it serves weH to highlight
the features of the interaction between the current account and the capital
account. This interaction, it should be emphasized, does not alter the fact
that in the short-run the exchange rate is always determined in the asset
market(s), even if it tends towards a long-run value (rE) which corresponds
to current-account equilibrium. All the above reasoning, in fact, is based
on the assumption that Eq. (9.10) holds instantaneously as an equilibrium
relation, so that any change in F immediately gives rise to a change in r,
which feeds back on the current account, which ''foHows'' the exchange-rate
behaviour. The assumption of instantaneous equilibrium in asset markets, or
(at any rate) of a much higher adjustment speed 0/ asset markets compared to
that 0/ goods markets, is thus essential to the approach under consideration.
Thanks to this assumption, in fact, the introduction of goods markets and
so of the current account does not alter the nature of the relative price of
two assets attributed to the exchange rate in the short run.
the determinants of the supply of and demand for steel and the mechanism
which brings these into equilibrium in the steel market, and so determines
the price of steel as weIl as the quantity exchanged.
Exactly the same considerations apply, according to Kouri, to the ex-
change rate, which is a price actually determined in the foreign exchange
market through the demand for and supply of foreign exchange. This lack
of, or insufficient, consideration of the foreign exchange market is, in Kouri's
opinion, the main shortcoming of the modern theory. In fact, no theory of
exchange-rate determination can be deemed satisfactory if it does not explain
how the variables that it considers crucial (whether they are the stocks of
money or the stocks of assets or expectations or whatever) actually translate
into supply and demand in the foreign exchange market which, together with
supplies and demands coming from other sources, determine the exchange
rate.
We fully agree with these considerations of Kouri's which, of course, are
not to be taken to mean that the modern approach is useless. Indeed, we
believe that neither the traditional nor the modern theory is by itself a satis-
factory explanation. In fact, as we have already observed, the determinants
that we are looking for are both real and financial, derive from both pure flows
and stock adjustments, with a network of reciprocal interrelationships in a
disequilibrium setting. It follows that only an eclectic approach is capable
of tackling the problem satisfactorily, and from this point of view we believe
that more complex models seem called for, because-as soon as one consid-
ers the exchange rate as one among the various endogenous variables which
are present in the model of an economic system-simple models like those
which have been described in this book are no longer suflicient. One should
move toward economy-wide macroeconom(etr)ic models. When doing this,
however, one shOuld pay much attention to the way in which exchange-rate
determination is dealt with.
In order to put this important topic into proper perspective, we first intro-
duce the distinction between models where there is a specific equation for the
exchange rate and models where the exchange rate is implicitly determined
by the balance-of-payments equation (thus the exchange rate is obtained by
solving out this equation). From the mathematical point of view the two
approaches are equivalent, as can be seen from the following considerations.
Let us consider the typical aggregate foreign sector of any economy-wide
macroeconomic model, and let CA denote the current account, N F A the
stock of net foreign assets of the private sector (its variation, 6.N FA, rep-
resents net capital flows) , R the stock of international reserves. Then the
balance-of-payments equation simply states that
where the minus signs reflect the accounting convention explained in Sect.
136 Chapter 9. Exchange-Rate Determination
Starting from an equilibrium situation in which both prices and output are at
their respective target levels, suppose that there is an unanticipated exoge-
nous rise in money demand. Given the money supply, this tends to cause an
increase in the interest rate. Because of the assumed perfect capital mobility,
there is an incipient capital inflow.
Under fixed exchange rates the incipient capital inflow causes an increase
in money supply until it becomes equal to the increased money demand.
Output does not change neither does the price level. The fixed exchange
rate regime completely stabilizes the economy.
Under flexible exchange rates the incipient capital inflow causes an ap-
preciation of the exchange rate which depresses aggregate demand. The
(negative) excess de~and for output causes both a decrease in output and a
decrease in the price level.
It is clear that in these circumstances the flexible exchange rate regime is
142 Chapter 9. Exchange-Rate Determination
exchange-rate regime will depend on the relative weights given to the two
targets.
11) In the latter case the initial excess supply for money turns into an
excess demand for money, which causes a tendency for the interest rate to
increase and hence an incipient capital inflow.
Under fixed exchange rates the incipient capital inflow causes an increase
in the money supply until it is brought in line with the increased demand. In
the final equilibrium output remains lower and the price level remains higher
than in the initial equilibrium, a result qualitatively similar to case I).
Under flexible exchange rates the incipient capital inflow causes an
exchange-rate appreciation which lowers aggregate demand and hence out-
put. In the final equilibrium output will be lower than under fixed exchange
rates, and the price increase will also be lower than in the fixed exchange
rate regime.
Thus we see that now fixed exchange rates are superior as regards output
stabilization, inferior as regards price stability. Also in this case the decision
on the better exchange-rate regime will depend on the relative weights given
to the two targets. However, given the weights, the decision will be crucially
dependent on the structural parameters of the economy. With the same
relative weights, in fact, the choice in case I will be exactly the opposite of
the choice in case 11, and vice versa.
9.5.2.4 Conclusion
The modern approach has succeeded in reducing the range of uncertainty
but has not been able to settle the debate. Notwithstanding its simplicity,
the model adopted has clearly shown the reasons for this failure. The choice
between the two regimes does, in fact depends on several factors:
a) the type of shock,
b) the structural parameters of the economy,
c) the objective function of the authorities.
More complicated models would not change this conclusion.
Costs and benefits will have to be weighted according to a social pref-
erence function, which may vary from country to country (and from period
to period in the same country). In addition, the structural parameters of
the economy are not given once-and-for-all, but are subject to change, the
more so the more dynamic is the economic system. In conclusion, "no single
currency regime is right for all countries or at all times" (Frankei, 1999).
Cushman, D.O., 2000, The Failure of the Monetary Exchange Rate Model
for the Canadian-U.S.
Dollar, Canadian Journal of Economics 33, 591-603.
De Grauwe, P., 1996, International Money: Postwar Trends and Theories,
2nd edition, Oxford: Oxford University Press.
Dornbusch, R, 1976, Expectations and Exchange Rate Dynamics, Journal
of Political Economy 84, 1161-76.
Frankei, J.A., 1983, Monetary and Portfolio Models of Exchange Rate De-
termination, in J.S. Bhandari and B.H. Putnam (eds.), Economic In-
terdependence and Flexible Exchange Rates, Cambridge (Mass.): MIT
Press, 84-115.
Frankei, J. A., 1999, No Single Currency Regime is Right for All Countries
or at All Times, Essays in International Finance No. 215, International
Finance Section, Princeton University.
Gandolfo, G., P.C. Padoan and G. Paladino, 1990, Exchange Rate Deter-
mination: Single-Equation or Economy-Wide Models? A Test Against
the Random Walk, Journal of Banking and Finance 14,965-92.
Howrey, P.E., 1994, Exchange Rate Forecasts with the Michigan Quarterly
Econometric Model of the U.S. Economy, Journal of Banking and Fi-
nance 18, 27-41.
Isard, P., 1995, Exchange Rate Economics, Cambridge (UK): Cambridge
University Press.
Kouri, P.J.K., 1983, Balance ofPayments and the Foreign Exchange Market:
ADynamie Partial Equilibrium Model, J .S. Bhandari and B.H. Put-
nam (eds.), Economic Interdependence and Flexible Exchange Rates,
Cambridge (Mass.): MIT Press.
Lanyi, A., 1969, The Case for Floating Exchange Rates Reconsidered, Es-
says in International Finance No. 72, International Finance Section,
Princeton University.
Meese, RA. and K. RogofI, 1983a, Empirical Exchange Rate Models of the
Seventies: Do They Fit Out of Sampie?, Journal of International Eco-
nomics 14, 3-24.
Meese, RA. and K. RogofI, 1983b, The Out-of-Sample Failure of Empir-
ical Exchange Rate Models: Sampling Error or Misspecification?, in
J.A. Frenkel (ed.), Exchange Rates and International Macroeconomics,
Chicago: Chicago University Press.
Meese, RA. and K. RogofI, 1988, Was It Real? The Exchange Rate-Interest
Differential Relation over the Modern Floating-Rate Period, Journalof
Finance 43, 933-48.
Pakko, M.R. and P.S. Pollard, 2003, Burgernomics: A Big Mac™ Guide to
Purchasing Power Parity, Federnl Reserve Bank of St Louis Review 85,
9-27.
Stockman, A.C., 1999, Choosing an Exchange-Rate System, Journal of Bank-
ing and Finance 23, 1483-98.
Chapter 10
y=A+B, (10.1)
whence, considering the variations and rearranging terms,
(10.2)
Equation (10.2) shows that for a devaluation to improve the balance of pay-
ments it must either cause a decrease in absorption at unchanged income, or
an increase in income at unchanged absorption or (better still) both effects,
or suitable combinations of changes in the two variables (for instance, both
145
146 Chapter 10. The Intertemporal Approach to the Balance of Payments
income and absorption· may increase, provided that the latter increases by
less, etc.). No elasticities are actually involved.
Equation (10.2) is of course an accounting identity, and, to give it a causal
interpretation, we must answer three questions:
(i) how does the devaluation affect income;
(ii) how does a change in income affect absorption;
(iii) how does the devaluation directly (i.e., at any given level of income)
affect absorption.
For this purpose we first recall from Chap. 6 that consumption and in-
vestment are functions of income, so that we can write the functional relation
where c is the sum of the marginal propensity to consume and the marginal
propensity to invest, and d denotes the direct effect of the devaluation on
absorption. By obvious substitutions we get
L -______~__-L~~_______ Co
o
takes place at point Eh-, where the (absolute value of) slope of the indifference
curve I T equals tanß = 1 + i*, hence i* = p.
The attainment of point Eh- implies the borrowing from abroad of an
amount COTCOA in the current period (which is like an import, hence a current
account deficit) and the repayment of ClACIT in the next period (this is like
an export, namely a curent account surplus). It is easy to see that EAQ =
ETQ·tanß, hence ClACIT = COTCOA · (1 +i*), or COTCOA = CIACIT /(1+i*).
In general, letting CA denote the current account,
CAI
CAo + -1-. =0, (10.6)
+ z*
where each current account has to be taken with its sign, and the second re-
lation follows from the accounting identity (I - S) +CA = 0 (see the first row
of the matrix illustrated in Sect. 5.3; here we have neglected the government
sector). Equation (10.6) is ealled the intertemporal budget constraint.
The souree of intertemporal trade is a difference between i A and i*, just as
in the statie model trade is determined by a differenee between the autarkie
and world eommodity priees.
Unlike the statie model, where commodities are exehanged for commodi-
ties (barter trade), in the intertemporal model trade involves the exchange
of commodities for assets, which are claims on future production. Thus in
period 0 the importing country will give some sort of bond to the exporting
150 Chapter 10. The Intertemporal Approach to the Balance of Payments
country; this bond states the obligation to repay in period 1 the amount of
the commodity borrowed plus interest. Using the terminology of the bal-
ance of payments (see Chap. 5), in period 0 our economy runs a current
account deficit matched by a capital account surplus (the sale of the bond
to the foreign country). This confirms what we stated at the beginning,
that the current account is determined by the saving-investment decisions of
the economy. These decisions also determine capital fiows as the necessary
counterpart to commodity fiows. Hence the overall balance of payments is
necessarily in equilibrium, and its structure is completely determined.
We have seen that in moving from E A to ET the economy attains a higher
welfare level. This, however, implies that the agent who borrows is the same
as the one who repays, and may be misleading when this is not the case. If,
for example, the actual length of the period is very long, so that the agents
living in period 1 are different from those who lived in period 0, the outcome
is that agents living in period 0 are better off at the expense of those living
in period 1, who must repay the debt without having previously enjoyed the
benefits of higher consumption. This problem is even more serious when
one extends the model to a multi-period setting, and can be dealt with in
several ways. One is to assume infinitely-lived agents. Another is to assume
overlapping generations, namely a multi-period setting in which people live
for two periods, so that in any period there are both "old" (those in their
second period of life) and ''young'' (those in their first period of life) agents.
A simpler way out is described here.
In the model so far examined there is no place for investment. Borrow-
ing and lending take place only to smooth consumption between different
time periods. Let us now assume that our homogeneous good can be both
consumed and invested (used as capital good) , so that lower consumption to-
day means higher output and consumption next period (less corn consumed
today means more corn planted and hence more corn produced tomorrow).
The intertemporal transformation curve is drawn in Fig. 10.2; it is concave
to the origin due to diminishing marginal productivity of capital.
Let us first consider autarky. The economy is endowed with OQOA, OQlA
respectively in the current and next period. By consuming less than the
current endowment and investing the amount COA QOA of saved output, the
economy can obtain the additional output QlAClA in period 1, thus being
able to consume OClA rather than just OQlA. This is actually the solution
that maximizes the country's welfare, as shown by the tangency between the
intertemporal transformation curve and the highest indifference curve I A at
EA .
The opening up of trade (with the same endowments) at the international
interest rate i*, where tan ß = 1 + i*, enables the country to attain the
optimum point E T , clearly superior to E A . But what is striking is that now
consumption is high er in both periods (hence no intergenerational confiict can
arise). In the previous case (Fig. 10.1), higher current consumption could be
10.2. Intertemporal Decisions, the Current Account, and Capital Flows 151
I
I
I
I
I
I
I
CIT ---:..-----
C --I..-----~.
lA I E
: A:
I I I
I I I
QIA _+ ______ 4Q
I I A' :
: : I I
I I I I
I I I I
L-~____~~~--------__ Co
COA r COT
QOA
Let us observe, in conclusion, that the country will run, as in the pure
consumption case, a current account deficit matched by a capital account
surplus, hence the overall balance of payments will be in equilibrium.
I/Y =0.04
(0.02)
+ 0.89
(0.07)
S/Y, R 2 = 0.91,
where the numbers below the coefficients are the standard errors. The au-
thors took this as evidence for the lack of capital mobility. Subsequent studies
by the same and other authors found similar econometric results.
This poses a puzzle, because these results contradict other evidence that
capital is quite mobile within developed countries. The main theoretical point
is then whether a high correlation between saving and investment indicates
low capital mobility.
Various explanations have been set forth to solve the puzzle, and the most
recent ones have focused on two points.
The first evaluates the appropriateness of the saving-investment correla-
tion as a measurement criterion, and argues that international parity con-
ditions are better criteria. It shows that international parity conditions,
properly viewed and estimated, provide strong evidence for capital mobility,
and hence the puzzle is resolved.
The second draws on the intertemporal approach, and starts by observ-
ing that saving and investment are related over time, as is implied by the
10.3. Intertemporal Approaches to the Real Exchange Rate 153
intertemporal budget constraint: see Eq. (10.6). The effect of this con-
straint seems strong enough to explain the high correlation between saving
and investment, without any implication on capital immobility.
where SN and IN are national saving and investment, economic agents de-
termine SN and IN according to SOFC, hence CA is determined.
The NATREX is the intercyclical equilibrium real exchange rate that
ensures balance-of-payments equilibrium in the absence of cyclical factors,
speculative capital movements and movements in international reserves. In
other words, the NATREX is the equilibrium real exchange rate that would
prevail if the above-mentioned factors could be removed and the GNP were
at capacity. Since it is an equilibrium concept, the NATREX guarantees
both the internal and the external equilibrium, the focus being on the long
run.
The study of the NATREX requires complex mathematical and econo-
metric analyses that we shall not go into; what we want to stress is that the
NATREX does not aim at tracking the actual real exchange rate, but is,
on the contrary, a measure of the long-run equilibrium real exchange rate,
the benchmark against which we can measure the misalignment of the actual
real exchange rate. Thus expressions like "the domestic currency is weak" ,
"the domestic currency is strong", "the domestic currency is undervalued" ,
"the domestic currency is overvalued", etcetera, which are often used in a
vague sense, can be given a precise meaning. Let us remember that the real
exchange rate (and hence the NATREX) is defined in such a way that an
increase means a (real) appreciation of the domestic currency. Hence when
the actual real exchange rate is lower (higher) than the NATREX, it follows
that the domestic currency is undervalued (overvalued).
International Monetary
Integration and European
Monetary Union
11.1 Introduction
There are various degrees of monetary integration, from the simple currency
area to the full monetary union (with a single currency). Thus a preliminary
conceptual and terminological clarification is called for. A good starting
point is the definition given in areport to the Council and Commission of
the European Economic Community commonly known as the Werner Report
(1970). It identifies a first set of conditions (called "necessary conditions" by
the subsequent Delors Report, 1989) to deflne a monetary union:
1) within the area of a monetary union, currencies must be fully and
irreversibly convertible into one another;
2) par values must be irrevocably fixed;
3) fluctuation margins around these parities must be eliminated;
4) capital movements must be completely free.
The second set of conditions identified in the Werner Report concerns the
centralization 0/ monetary policy. In particular, this centralization should
involve all decisions concerning liquidity, interest rates, intervention on the
exchange markets, management of reserves, and the fixing of currency parities
vis-a-vis the rest of the world.
Finally, in the Werner Report the adoption of a single currency, though
not indispensable for the creation of a monetary union, is considered prefer-
able to maintaining the various national currencies. This is so for psycholog-
ical and political factors, as the adoption of a single currency would demon-
strate the irreversible nature of the undertaking.
Some authors take the first set of elements listed in the Werner Report
and call it monetary integration. In practical usage this latter definition is
155
156 Chapter 11. International Monetary Integration and European Monetary Union
In fact, if either the fixed exchange rate regime or the flexible one could be
shown to be definitely superior, then there would be no need for a theory of
(optimum) currency areas: the optimum currency area would coincide with
the world (if fixed exchange rates were superior) or would not exist (in the
case of superiority of flexible exchange rates).
Three approaches can be distinguished in the theory of optimum currency
areas. The first is the traditional approach, which tries to single out a crucial
criterion to delimit the appropriate domain. The second is the cost-benefit
approach, which believes that the participation in a currency area has both
benefits and costs, so that optimality has to be evaluated by a cost-benefit
analysis. The third is the "new" approach.
goods one, so as to meet the increased (foreign) demand for exports and the
higher (domestic) demand for import substitutes. This implies huge distur-
bances (amongst which possible inflationary effects) in the non-traded goods
sector because of its relative smallness.
In this situation it would be more effective to adopt fixed exchange rates
and expenditure-reducing policies which reduce imports and free for expor-
tation a sufficient amount of exportables previously consumed domestically.
(c) A third criterion is that of product diversijication. A country with
a high productive diversification will also export a wide range of different
products. Now, if we exc1ude macroeconomic events which influence the
whole range of exports (for example a generalized inflation which causes
the prices of all domestically produced goods to increase), in the normal
course of events commodities with a fine or brilliant export performance
will exist beside commodities with a poor export performance. It is self-
evident that these offsetting effects will be very feeble or will not occur at
all when exports are concentrated in a very limited number of commodities.
On the average, therefore, the total exports of a country with a high product
diversification will be more stable than those of a country with a low one.
Since the variations in exports influence the balance of payments and so-
ceteris paribus-give rise to pressures on the exchange rate, it follows that
a country with high product diversification will have less need for exchange-
rate changes and so can tolerate fixed exchange rates, whilst the contrary
holds for a country having low product diversification.
(d) A fourth criterion is that of the degree of jinancial integration. It par-
tially overlaps with criterion (a), but it is especially concerned with capital
flows as an equilibrating element of external imbalances. If there is a high de-
gree of international financial integration, no need will exist for exchange-rate
changes in order to restore external equilibrium, because slight changes (in
the appropriate direction) in interest rates will give rise to sufficient equili-
brating capital flows; in this situation it is possible to maintain fixed exchange
rates within the area where financial integration exists. It goes without say-
ing that a condition for financial integration is the elimination of all kinds of
restrictions on international capital movements.
(e) A fifth criterion is that of the similarity in rates 0/ inflation. Very
different inflation rates do, in fact, cause appreciable variations in the terms of
trade and so, insofar as these influence the flows of goods, give rise to current-
account disequilibria, which may require offsetting exchange-rate variations.
When, on the contrary, the rates of inflation are identical or very similar,
there will be no effect on the terms of trade and so-ceteris paribus-an
equilibrated flow of current-account transactions will take place (with fixed
exchange rates) within the currency area.
(f) A sixth criterion is that of the degree 0/ policy integration. Policy
integration can go from the simple coordination of economic policies among
the various partner countries to a situation in which these surrender their
11.2. The Theory of Optimum Currency Areas 159
(2) The saving on exchange reserves. The members no longer need inter-
national reserves for transactions within the area, exactly as in the case of
regions within a country. This, of course, will occur when the credibility of
the fixed exchange rates is complete, whilst in the initial stages it may be
necessary to hold the same amount of pre-union reserves to ensure the agreed
exchange-rate rigidity, i.e. to enforce the fixed parities established within the
area.
(3) Monetary integration can stimulate the integration of economic poli-
eies and even economic integration. The idea is that partieipation in a cur-
rency area, and so the obligation to maintain fixed exchange rates vis-a..vis
the other members, compels all members to make their economic policies uni-
form (in particular anti-inflationary polieies) with those of the most virtuous
member, at the same time making more credible domestically the statements
of a firm intention to pursue a strong policy against inflation. This is essen-
tially the same argument of monetary discipline already set forth in general
in the debate on fixed versus flexible rates (see Sect. 9.5), strengthened by
the fact that the commitment to maintain fixed exchange rates within the
area would be feIt more strongly than the commitment to defend a certain
parity vis-a..vis the rest of the world as under the Bretton Woods regime.
Another argument, however, suggests that monetary agreements might give
rise to more inflation. This might happen when the national policy author-
ity is involved in a policy game with other institutional agents (trade unions,
etc.).
The argument of uniform inflation rates has been the subject of much
debate. Firstly, it has been noted that it curiously turns the position of
the traditional approach upside down: we have in fact seen in Sect. 11.2.1,
criterion (e), that according to the traditional approach the similarity in the
rates of inflation is a precondition for taking part in an (optimal) currency
area, whilst it now becomes a (beneficial) effect of taking part. Secondly,
many nonmonetarist writers deny that the achievement of a common rate of
inflation is ontologically a benefit, and observe that different countries may
have different propensities to inflate: some are more or less inflation-shy,
others more or less inflation-prone, and this reflects a different structure of
their national preference function.
In effect, some confusion seems to exist in relation to this argument.
Benefits (1) and (2) are fairly objective ones, in the sense that they do not
presuppose the adhesion to a particular school of thought or to a particular
preference function (practically everybody agrees that elimination of desta-
bilising speculation, and saving on international reserves, are benefits). On
the contrary, the equalization of the inflation rates is not considered univer-
sally desirable. Prom this point of view the traditional approach seems more
neutral, because it only states that if there is similarity in inflation rates,
then ground exists for participating in a currency area.
Behind the idea that monetary integration is conducive to economic inte-
11.2. The Theory of Optimum Currency Areas 161
under fixed exchange rates (see Sect. 6.4.3), but this is true for an isolated
country. In the case of a country belonging to a currency area, its fiscal
policy may be constrained by the targets of the area as a whole (for example
to maintain a certain equilibrium in the area's balance of payment vis-a-
vis the rest of the world). And since the joint management of the single
members' fiscal policies is carried out in the interest of the majority, it may
happen that some member is harmed (unless a vetoing power is given to each
member, in which case, however, there is the risk of a complete paralysis).
(3) Possible increase in unemployment. Assuming that the area includes a
country with low inflation and an external surplus, this country will probably
become dominant and compel the other members (with greater inflation and
an external deficit) to adjust, because-as there are no means to compel the
former country to inflate-the deficit countries will have to take restrictive
measures which will lead to a decrease in employment. The writers of the
monetarist school claim that in the long-run every country will be better off
thanks to the lower rate of inflation, but, even allowing this to be true, the
problem remains of determining how long is the long-run, as it is clear that
in the short-run there are costs to be borne.
4) Possible deterioration of previous regional disequilibria. "Regional" is
here used in the strict sense, i.e. referring to single regions within a mem-
ber country. Since the international mobility of capital (in the absence of
controls) is higher than the international mobility of labour, the greater pos-
sibilities of finding better-rewarded uses of capital in other count ries of the
area, together with the relatively low internationallabour mobility, may ag-
gravate the development problems of the underdeveloped regions of a coun-
try. It should be noted that this negative effect occurs insofar as what was
listed as the first criterion in the traditional approach (the high international
mobility of all factors) does not occur, but it seems in any case likely that
international labour mobility is lower than that of capital.
Having thus listed the benefits and costs, we conc1ude by pointing out
that the already mentioned problem of weighing them cannot be given a
generally valid answer, as it depends on the social welfare functions of the
different countries, that may be quite different.
the Phillips curve first by the natural rate of unemployment and then by the
NAIRU) and reality (the stagflation problem in the 1970s and early 1980s).
Thus it seems that, under this respect, the main benefit of flexible exchange
rates is only the ability of choosing a different rate of inflation from other
countries.
Recent work on the theory of optimum currency areas and monetary
integration concentrates on two issues: the effects 0/ shocks and reputation al
considerations.
As regards the effects of shocks, we have already seen in the new debate on
fixed versus flexible exchange rates (Sect. 9.5.2) that the modern approach
has succeeded in reducing the range of uncertainty but hasn't been able
to settle the debate. There are various reasons for this failure: the shock-
absorption capability of fixed and flexible exchange rates depends on several
factors, such as the type of shock, the structural parameters of the economy,
and the objective function of the authorities.
Reputational considerations start from the observation that, since policy
makers can do little more than choose an optimal rate of inflation, they should
aim at a zero (or at least very low) inflation rate. This is so because inflation
distorts relative prices through which information is usually transmitted,
thus creating uncertainty and inefIicient allocation of resources. The more
credible the anti-inflation commitment, the lower the costs associated with
a given decrease in the inflation rate, and the easier the implementation of
an anti-inflation plan. Credibility requires time consistency, namely that the
government will pursue the policy in the future because it has no incentive to
change it, and the public is convinced that the government will not change
it.
Given this, the view has been set forth that a high-inflation country in-
creases its credibility by fixing its exchange rate with respect to the currency
of a low-inflation country. But why the exchange rate rather than monetary
policy? The answer is twofold. First, the exchange rate is an easily observ-
able and still more easily understood variable by the public, while the money
supply is not (or is less) and, to the extent that it is observable, is difIicult
to control. Second (and related to the first), is the discipline argument: by
pegging the exchange rate the monetary authorities tie their hands much
more strictly than by a money supply commitment, hence gain in credibility.
Thus we must conclude that recent theoretical and empirical work has
produced ambiguous results. But, after all , the move to international inte-
gration is more an expression of political will than the outcome of purely
economical calculations.
problem, which means that the agreement to create a currency area may have
a negative efIect on non-participating countries. Consider,for example, the
formation of a currency area among countries with regionally difIerentiated
goods: a possible result is a reduction in the area's output (because of the
interaction between the common exchange rate and lower wage flexibility).
This will not only be a cost for the union's members, but also for the rest
of the world whose trade will be negatively afIected. On the contrary, the
benefits from the union are limited to the union's members.
Although these results need not be valid in general, the possibility of a
welfare-Iowering efIect (for the rest of the world) due to the creation of a
currency union should be taken into consideration.
(11.3)
With perfect asset substitut ability only the money stock matters (see
Sect. 9.3), so that we can consider only money-market equilibrium (demand
for money = supply of money) at home and abroad. Thus we have
PfLRf(Yf,if) = M f , (11.5)
where Lm., L Rf denote the real money deniand at home and abroad.
11.3. The Common Monetary Policy Prerequisite, and the Inconsistent Triad 165
Even if we take the price levels and the outputs as given, the three equa-
tions (11.3)-(11.5) form an undetermined system, which is unable to deter-
mine the four unknowns (the two money supplies and the two interest rates).
Thus, there is a fundamental indeterminacy of the money supply and the in-
terest rate in this two-country system. It follows that the two countries will
have to (implicitly or explicitly) agree on the conduct of monetary policy.
We must stress that the apparent absence of agreement can simply be due
to the presence of an implicit agreement. The typical implicit agreement is
asymmetrie, in the sense that it is based on the dominant role of one coun-
try. This means that one country sets its money supply according to its own
criteria and the other country adapts its money stock.
Suppose, for example, that the foreign country is the dominant country
and fixes M J . Then Eq. (11.5) determines i J , which sets ih by Eq. (11.3).
Finally, Eq. (11.4) determines M h . Thus the horne country must set its
money supply at this level.
If it does otherwise, the currency area will break down. In fact, suppose
for example that the horne country tries to fix a higher money supply. This
will depress i h below i J . As a consequence, immediate and disrupting capital
flows from horne to abroad will take place, unless controls on capital flows
are introduced (but such controls are hardly compatible with monetary in-
tegration). These flows will lead to expectations of a future exchange rate
adjustment and to the breakdown of the exchange rate commitment.
Explicit agreements on the conduct of the overall monetary policy are, on
the contrary, of the cooperative type. This requires that countries agree to
cooperate in setting their money stocks (or interest rates). This presents the
weIl known free-riding problem, the same kind of problem treated at length
in the theory of international price cartels (see Sect. 14.4.2). In general, this
means that once a cooperative agreement has been reached, there are usually
incentives for one partner to do something else than was agreed upon. This
follows from the fact that, by so doing, the partner will be better off, if the
other partners do not retaliate (for example by reneging on the agreement).
Thus some institutional mechanism has to be devised to avoid the free-riding
problem.
These considerations confirm the essential importance of viable agree-
ments on the conduct of monetary policy within the currency area. We have
illustrated this proposition by the simplest model possible, but the results do
not change substantially with more complicated models, such as those based
on Eq. (11.2) instead of Eq. (11.3). This would require the introduction of
the stocks of assets (money and other financial assets) and the determination
of the portfolio-balance equilibrium.
Since it is impossible for a country to simultaneously peg its exchange rate
and allow unfettered movement of international capital, while retaining any
autonomy over its monetary policy, the set of fixed exchange rates, perfect
capital mobility, and monetary independence has been called the inconsistent
166 Chapter 11. International Monetary Integration and European Monetary Union
they will evaluate the costs and benefits of the fixed exchange rate union, as
shown in Sect. 11.2.2. If the costs become overwhelming with respect to the
benefits, the government concerned may be tempted to change the parity,
even if this means breaking an international agreement. This is by no means
an impossible occurrence. The evaluation of the costs and benefits, in fact,
may vary over time, for example in relation to economic conditions and/or
to preference functions of different governments. Thus a fixed exchange rate
system does not eliminate the risk that a temporary change in this evaluation
might lead a member country to alter the parity. Rational economic agents
know this, hence the possibility of speculative capital flows and of an un-
certain climate for businesses. Actually, models of currency crises have been
built based on the possibility of the government to renege the commitment
to fixed exchange rates (an escape clause: see Sect. 8.3).
(3) The elimination of destabilising speculative capital flows within the
union, due to expectations of parity changes as detailed under (2).
(4) The elimination of the need for intra-union international reserves,
required to make the commitment credible and to offset possible speculative
capital flows.
(5) We have shown in Sect. 11.3 that a common monetary policy is
necessary for a monetary union. A single currency would greatly facilitate
the conduct of this overall monetary policy, and would eliminate free-riding
problems.
(6) A single currency would carry more international weight and enable
the union to reap the benefits of seignorage. In addition, the interventions
in the foreign exchange market vis-a..vis other currencies would be greatly
facilitated and would require a smaller amount of international reserves vis-
a-vis the rest of the world.
The main costs that have been stressed are the following:
(1) Costs for the transformation of the system of payments. These include
the costs of changing existing monetary values into the new currency, the
costs of changing coin machines, etc.
(2) The psychological cost to the public of introducing the new currency
and their getting used to it. It is not sufficient to declare a currency legal
tender for this to be used willingly in a country in the place of the existing
national currency. A new currency cannot be merely imposed by legislative
act, but must gain social consensus and be accepted by the market. Thus the
authorities will have to ensure that the new currency performs the functions
of money at least as efficiently as the existing national currencies. This
process of convincing the public is not without costs.
(3) This point is usually presented as an advantage of the fixed exchange
rate system, rather than as a cost of the single-currency system. It is called
the currency competition argument. Several currencies in competition stim-
ulate each national monetary authority in the group to pursue a lower rate of
inflation and, more generally, a stable value of its respective currency. How-
168 Chapter 11. International Monetary Integration and European Monetary Union
way for the European System of Central Banks (to come into being in the
stage 111). The EMI will dissolve on the starting day of stage three.
(11.3) obligation on the part of the member countries to conform domestic
legislative provisions concerning their central banks to the principles of the
Union.
(11.4) elimination of any automatie solidarity commitment to aid member
countries faced with problems.
111) The third stage was to begin on 1st January 1997 or 1st January 1999
at the latest. More precisely, the proviso was that at the end of 1996 the
European Council would meet and decide whether the majority of member
countries satisfied certain convergence criteria: in the affirmative case, the
third stage would begin on 1st January 1997. In the negative case, the
third stage would be postponed but not later than 1st January 1999, when
it would in any case begin with the participation of those countries that
met the convergence criteria. The other countries would obtain a temporary
derogation and enter when they will satisfy the criteria.
The third stage actually began on 1st January 1999.
Let us now examine the convergence criteria, which are the following:
(a) an inflation rate (as measured by the rate of increase of the consumer
price index) that does not exceed by more than 1.5 percentage points the
rate of inflation of the three best performing countries (i.e., those having the
three lowest inflation rates);
(b) a long-term nominal interest rate (measured on the basis of long-term
government bonds) that does not exceed by more than 2 percentage points
the average of those same three countries;
(c) an exchange rate that has respected the normal fluctuation margins
in the last two years;
(d) a public deficit and debt that satisfy the criteria detailed under ILl
above.
The measures contemplated in the third stage, that gave rise to the Eu-
ropean Union proper, are the following:
(IlI.1) creation of the European System of Central Banks (ESCB), which
consists of the national central banks plus the European Central Bank (ECB).
The ESCB has the task of taking all decisions concerning monetary policy,
including the control of the money supply, with the primary objective of
maintaining price stability and the subordinate (i.e., without prejudice to
the objective of price stability) objective of supporting the general economic
policies in the Union.
(Il1.2) the bilateral exchange rates are irrevocably fixed, as well as those
vis-a-vis the ECU, that will become a currency by full right.
(111.3) the ECU will replace the single national currencies at the earliest
possible date.
(IIl.4) the Community will be entitled to apply appropriate sanctions
against the countries which infringe the EC financial regulations after joining
172 Chapter 11. International Monetary Integration and European Monetary Union
stage three.
(III.5) the position of those countries that were granted aderogation
(and were therefore temporarily left out of the Union: see above) will be
reconsidered every two years.
These measures were subsequently integrated by
a) the Madrid meeting of the European Council (December 1995), where
it was established that the common European currency should be called euro
rather than ECU, to emphasize that it was a brand new currency and not a
basket currency like the ECU (for further details see below, Sect. 11.5.6);
b) the Dublin summit in December 1996, where the so-called stability and
growth pact was adopted on a proposal of the Commission and the Council
of the economic and financial Ministers of the European Union (Ecofin), that
in turn acted at the behest of Germany. Under this pact the 3% deficitjGDP
ratio is taken as an upper limit, since in normal circumstances the ins (i.e.,
the countries that have been admitted to stage III) should pursue a medium-
term balanced budget or even a surplus. Several institutional mechanisms
are introduced to ensure the respect of the 3% upper limit, in particular the
possibility of Ecofin to issue fines (up to 0.5% of GDP in any one year) to
members that, after having been admitted to the third phase, do not respect
the 3% deficitjGDP ratio. However, in the meeting of 25 November 2003,
Ecofin decided not to issue fines to France and Germany which had exceeded
the 3% limit.
b.Ng b.Y
-Ng- < -
- Y'
b.Ng b.Y Ng
-Y- <
- Y- -
y·
9 b.Y Ng
-
Y <
- Y- -
y· (11.6)
11.5. The European Monetary Union 177
where ~pjp is the inflation rate and ~yjy is the rate of growth ofreal GDP.
In a target-instrument policy framework, it is perfectly reasonable that
policy makers assign "desired" values to real growth and inflation. For ex-
ample 3% and 2% respectively, are plausible desired values, whence ~YjY =
5%.
As regards b, oral tradition says that b = 60% simply came out of the fact
that 60% happened to be the EEC average when the Maastricht treaty was
drafted. Given these values of b and ~YjY, the value of gjY = 3% foilowed.
Another oral tradition says that a public deficit was to be allowed only for
public investment expenditures, whose value could be taken as 3% of GDP,
hence gjY = 3%. Given ~YjY = 5%, the value of b = 60% followed.
We can however argue in favour of a less casual explanation. We know
that equilibrium on the real market consistent with current-account equilib-
rium requires
g+ (1 - 8) = 0, (11.9)
which follows from Chap. 5, Sect. 5.3, letting CA = o. Thus we have
9 8-1
Y = y-, (11.10)
178 Chapter 11. International Monetary Integration and European Monetary Union
where i is the nominal interest rate and gp the primary deficit (let us re-
member that according to our convention, a negative value of gp means a
primary surplus). Substitution of (11.11) into (11.6) yields
gp <
Y-
(ßY
Y
_.)• N9y· (11.12)
Using (11.8) and the definition ofreal interest rate iR = i-b.p/p, Eq. (11.12)
can be rewritten as
(11.13)
that the basic criteria for an optimum currency area, in particular high factor
mobility and high flexibility of wages and prices (see Sect. 11.2.1) are not
satisfied.
This is undoubtedly true, but these criteria should be seen in the context
of the degree of similarity in economic structure. In general, given that
each member of a single-currency area cannot use the exchange-rate tool
to cope with asymmetrie shocks, the main possibilities are migration from
low to high growth countries, wage flexibility, unilateral transfers from high
to low income countries. None of these seems to exist in EMU. However,
it should be pointed out that predominantly asymmetrie shocks do require
high factor mobility and high pricejwage flexibility for being absorbed in
areas with pronouneed regional disparities (such as the United States). These
requirements are however much less important when similarity in eeonomie
strueture reduees the likelihood of asymmetrie shoeks.
In effect, the analysis of the consequences of exogenous disturbances on
the participating countries is one of the main areas of research in the so-called
"new" theory of optimum currency areas (see Sect. 11.2.3, and Tavlas, 1993).
Unfortunately in the context of cost-benefit analysis it becomes diflicult to
make definitive statements, and we fully agree with Wyplosz when he writes
"Assessing the costs and benefits of a monetary union quantitatively is both
frustrating and useless. It is frustrating because, frankly, as economists we
are unable to compute them with any precision, and we owe it to the pro-
fession to admit so in public. Our understanding of monetary and exchange
rate policy is regrettably limited, and the lack of a precedent leaves us with
more conjectures than certainties. Moreover, quantitative estimates are use-
less unless they are sized up against the costs and benefits of the relevant
alternatives, which is equally beyond our current ability. The best that can
be done in this situation is to gain an understanding of where the costs and
benefits are likely to reside" (Wyplosz, 1997, pp. 18-19).
In this respect, it is interesting to point out a usually neglected cost-
benefit, due to the distribution of seignorage in EMU. As noted by Sinn and
Feist (1997, p. 666), some countries joining the EMU "will win more than
others, because they will receive a better currency than the one they lose. A
good currency is highly demanded as a medium of transactions and a store
of value and its wide usage creates a substantial seignorage wealth for the
issuing country. With the introduction of the euro, national currencies will
disappear and seignorage wealth will be socialized". Some countries stand to
gain, others to lose, depending on different scenarios as regards membership
in the euro.
The European Commission's study (1990) diseusses at some length the
advantages of EMU: macroeconomic stability, price stability (both low in-
flation and low variability), reduction in transactions costs, elimination of
currency risk. According to the Commission the direct costs of foreign trans-
actions in the EU were estimated at between one half and one percent of its
180 Chapter 11. International Monetary Integration and European Monetary Union
1999 the official ECU basket would cease to exist, and the euro would become
a non-circulating currency in its own right, that it would begin to circulate
together with national currencies as legal tender from 1st January 2002,
totally replacing national currencies from 1st July 2002 at the latest. In the
case of contracts denominated in ECUs substitution by the euro has been at
the rate one to one.
At the Dublin Summit (December 1996) the Heads of State or Govern-
ment ascertained that the Maastricht criteria were not satisfied by a majority
of the fifteen Member States, hence the third step (see Sect. 11.5.2) could
not begin on 1st January 1997.
On 1st May 1998 the Council of the economic and financial Ministers of
the European Union (Ecofin) reviewed the situation of the member states as
regards the Maastricht criteria and identified the member states that satis-
fied the conditions for the introduction of the euro; these countries (called
the "participating countries" ) were, so to speak, the "founding members" of
the euro. On 2nd May the recommendation of Ecofin was ratified by the
Council of the Heads of State and Government of the EU and by the Euro-
pean Parliament. The eleven founding members turned out to be Belgium,
Germany, Spain, France, Ireland, Italy, Luxembourg, the Netherlands, Aus-
tria, Portugal and Finland (on 1st January 2001 Greece has been admitted).
It is important to note that, as regards the two fiscal criteria, the one con-
cerning the deficitjGDP ratio was applied strictly, while the criterion of the
debtjGDP ratio was interpreted "dynamically", in the sense that a country
to qualify should have shown a consistently decreasing trend towards the
60% reference value, even if the current debtjGDP ratio was actually higher
(this was the case, for example, of Italy, where in the last few years this ratio
had been on the decrease, but in 1997 was still 121.6%).
On 3rd May 1998 Ecofin also established the fixed bilateral exchange
rates among the participating countries, valid in the transitional period May-
December 1998, when the euro did not yet exist. This gave rise to some fears
concerning possible destabilising speculation before the conversion rates of
the various currencies into the euro after this transitional period were estab-
lished, or possible withdrawals by prospective EMU members, but nothing
happened.
At the moment the main international currency remains the dollar, but
the importance of the euro is on the increase. According to some writers
it is possible that in about ten years the dollar and the euro will be con-
sidered equally important. If so, the countries in the rest of the world will
presumably wish to include in their international reserves an equal amount
of dollars and euros. Since the current composition of international reserves
sees the predominance of the dollar, there will probably be an increase in the
demand for euros on the part of the rest-of-the-world countries. To satisfy
this increased demand deficits in the EU balance of payments will be called
for.
182 Chapter 11. International Monetary Integration and European Monetary Union
exchange-rate variability.
The final problem remains of how to determine the dollar/euro (real)
equilibrium exchange rate. In fact, expressions like "the dollar is weak",
"the euro is strong", or "the euro is over/under-valued" are meaningless
without a benchmark. A reliable theoretical framework for the calculation of
a fundamental equilibrium exchange rate is of invaluable importance for the
EU, not only as regards monetary matters (which are delegated to the ECB),
but also as regards real matters such as trade and growth, which are the main
responsibility of the Council, which may formulate general orientations for
exchange rate policy (article 109(2) of the Maastricht Treaty on European
Union). Trade and growth are clearly related to the EU's welfare.
There is no reason to believe that the exchange rate theories examined
in Chap. 9 would perform better with the dollar/euro exchange rate than
with the dollar/Single currencies. A better research program might be to
concentrate on the real exchange rate in the context of the intertemporal
approach according to the NATREX model (see Chap. 10, Sect. 10.3).
Kenen, P. B., 2002, The Euro Versus the Dollar: Will There Be a Struggle
for Dominanee?, Journal of Policy Modeling 24,347-354.
MeKinnon, R, 2002, The Euro Versus the Dollar: Resolving a Historical
Puzzle, Journal of Policy Modeling 24, 355-359.
Mundell, RA., 1961, A Theory of Optimum Currency Areas, American Eco-
nomic Review 51, 509-17.
Mundell, RA., 1973, Uneommon Arguments for Common Currendes, in H.J.
Johnson and A.K. Swoboda (eds.), The Economics of Common Cur-
rencies: Proceedings of the Madrid Conference on Optimum Currency
Areas, London: Allen&Unwin.
Mundell, RA., 1998, What the Euro Means for the Dollar and the Interna-
tional Monetary System, Atlantic Economic Journal 26, 227-37.
Parsley, D. and S. Wei, 2001, Limiting Currency Volatility to Stimulate Goods
Market Integration: A Price Based Approach, NBER WP No. 8468.
Prati, A. and G.J. Schinasi, 1999, Financial Stability in European Economic
and Monetary Union, Princeton Studies in International Finance No.
86, International Finance Section, Princeton University.
Ried, L., 1995, Exchange Rate Regimes and Location, International Eco-
nomics Working Papers No. 291, University of Konstanz (SFB178),
December; revised version circulated as International Monetary FUnd
Working Paper No. 97/69, June 1997.
Rogers, J. et al., 2001, Price Level Convergence and Inflation in Europe,
W.P. 01-1, Institute for International Economics (Washington DC).
Rose, A., 2000, One Money, One Market: The Effect of Common Currendes
on TI:ade, Economic Policy 30, 7-45.
Rose, A., 2002, The Effect of Common Currencies on International Trade:
A Meta-Analysis, mimeo.
Sinn, H.-W. and H. Feist, 1997, Eurowinners and Eurolosers: The Distri-
bution of Seignorage Wealth in EMU, European Journal of Political
Economy 13, 665-89.
Tavlas, G.S., 1993, The 'New' Theory of Optimum Currency Areas, World
Economy 16, 663-85.
Thya, J. and L. Zamalloa, 1994, Issues on Placing Bank Supervision in the
Central Bank, in T. Balino and C. Cottarelli (eds.), Frameworks for
Monetary Stability, Washington (DC): IMF.
Werner Report, 1970, Report to the Council and the Commission regard-
ing the Step-by-Step Establishment of the Community's Economic and
Monetary Union, Brussels: Commission of the European Communities,
80ctober.
Wyplosz, C., 1997, EMU: Why and How it Might Happen, Journal of Eco-
nomic Perspectives 11, 3-22.
Chapter 12
12.1 Introduction
In this final chapter of Part II we examine some current problems of the
international monetary system. Any choice is inevitably arbitrary, and what
is a problem today might cease to be such tomorrow. However, the problems
that we are going to consider are likely to present a theoretical interest to
students of international finance and open-economy macroeconomics also in
the future. In any case we owe an explanation for the omission of the standard
topic called "plans for reform of the international monetary system". We
have already mentioned the current international monetary ''non system"
(see Chap. 3, Sect. 3.4). Proposals for the creation of a new international
monetary system have been made from time to time, but the amount of
international agreement that this creation would require (see the section on
international policy cooperation, below) makes it unlikely. Hence we prefer to
concentrate on a much more modest but viable alternative, the international
management of exchange rates.
Thus the topics we are going to deal with are:
1) international policy coordination;
2) the debt problem;
3) the Asian and other currency crises;
4) the management of exchange rates.
The third problem has already been examined elsewhere (see Sect. 8.3).
When dealing with the second problem we shall integrate our treatment with
the exposition of the basic model that the IMF and the World Bank had in
mind when dealing with crises: the so-called growth-oriented adjustment
programs.
185
186 Chapter 12. Problems of the International Monetary System
the current account (the current account of the contracting country will of
course improve). In terms of game theory we are in the so-called 'prisoner's
dilemma' situation!. If one country expands and the other does not, the
latter will gain (in terms of current account balance) at the expense of the
former. Without coordination, it is impossible to reach the optimal situation
in which both countries expand.
In reality no economic policy consists of two alternatives only (expan-
sion/ contraction in our example), but can vary in more or less continuous
manner from restriction to expansion. Furthermore, even a partial fulfilment
of a target (maybe at the expense of another target) has a value. This means
that the government of a country does not carry out its economic policy in
a fixed-target context (in our example this would be a given level of employ-
ment and equilibrium in the current account balance), but in a flexible-target
context. The flexible-target approach means that there is a trade-off among
the various targets, in the sense that a higher fulfilment of a target can com-
pensate for a lower fulfilment of another, given a social welfare function. The
policy maker aims at maximizing the social welfare function, which depends
on the degree of fulfilment of the objectives, given the constraints (repre-
sented by the economic system and the other countries' actions). This more
general framework can be represented by a diagram due to Hamada.
If we assurne that the policy configuration of each country can be rep-
resented by a synthetic and continuous variable, we can show the policy
configurations of the two countries on the axes of a diagram like Fig. 12.1.
The policy configuration of country 1 is shown on the horizontal axis, while
the policy configuration of country 2 is shown on the vertical axis. A point
in the diagram thus represents a combination of the two countries policy
Country2
p,t:
2
p'
2
its optimal welfare independently of the policy pursued by the other coun-
try. Hence the case for coordination is based on the presence of international
interdependence.
In the case of interdependence, the first step of the constrained-optimum
problem that each country has to solve is to determine its welfare-maximizing
policy configuration for any given policy configuration of the other country.
Let us consider, for example, country 1, and let us assume that country 2
adopts the policy configuration represented by point p~. If we draw from
this point a straight line parallel to country l's axis, we see that the highest
indifference curve that country 1 can reach is U~, tangent to the aforesaid
straight line. U~, in fact, is the indifference curve nearest to Ei compatibly
with the given policy choice of country 2 (the constraint). Hence, given p~,
the optimal choice for country 1 is policy p{E. Going on in like manner,
we obtain a set of points that give rise to the RiRi curve, called the policy
reaction junction of country 1 (for graphical simplicity we have drawn it
linear).
In a similar way we obtain country 2's policy reaction curve. Given for
example country 1's policy configuration P{, the indifference curve of country
2 which is nearest to the bliss point E 2 is the curve tangent to the straight
line originating from P{ and parallel to country 2's axis.
In the diagram we also have drawn a segment joining the two ideal points
Ei, E 2 , which is the locus of all points where the two countries' indifference
curves are tangential to one another. This locus is a Pareto-optimal or con-
tract curve, since in each of these points the property holds that it is not
possible to increase the welfare of one country without decreasing the other
country's welfare.
We can now use the Hamada diagram to illustrate what happens without
coordination, and the advantages of coordination. In Fig. 12.2 we have
drawn the two policy reaction curves obtained as explained above, and we
now want to know what will be the behaviour of the two countries.
Let us begin by considering a non-co operative behaviour, that can take
on various forms. The most commonly used are the Cournot-Nash and Stack-
elberg scenarios.
In the former, each country maximizes its welfare by choosing its own
optimal policy taking as given the policy configuration of the other country,
on the assumption that this configuration is beyond its influence. Given for
example country l's policy p?, country 2 will choose P~ on its own reaction
function. In its turn country 1, taking as given country 2's P~, will change
its policy to P{ on its own reaction curve; country 2 will then react to P{ by
changing its policy to P~ and so forth, until point N is reached. This is the
Ur
ur
Cournot-Nash equilibrium, where the welfare achieved is for country 1
and for country 2.
The Stackelberg or leader-follower solution is obtained when one country
(for example, country 2) is dominant (the leader) and takes account of the
190 Chapter 12. Problems of the International Monetary System
Country 2
P'2
pO
2
pO P'I Country 1
I
fact that its actions infiuence the other country's decisions, while the follower
country behaves like in the previous case. The leader knows the reaction
curve of the follower and hence country 2 knows that country 1 will react to
the leader's policy choices by choosing a policy along the R1R1 curve. Thus
country 2 maximizes its welfare function taking account of this curve as a
constraint. This means that country 2 will choose the highest indifference
curve compatible with the constraint. This curve is ul which is tangential to
R1R1 at point S. In fact, ul is country 2's indifference curve that is nearest
to E 2 compatibly with R1R1.
It can be seen that in the Stackelberg equilibrium point S, country 2 (the
leader) obtains a welfare level clearly higher than in the Cournot-Nash equi-
librium, while the follower may or may not be better off. Both equilibria are
however inefficient, since they do not lie on the contract curve E 1E 2 • Both
countries could be better off if they agreed to cooperate: if they coordinate
their policies they can reach a point on the segment at of the contract curve.
Any such point is clearly superior to both the Cournot-Nash and the Stack-
elberg equilibrium. The precise point where the two countries will end up
will of course depend on the relative bargaining power.
12.2. International Policy Coordination 191
problems. It is the lack 0/ agreement on the model that explains the trans-
mission of the effects of economic policies (both domestically and from one
country to another) that determines the impossibility of international policy
coordination to fight unemployment, low (or negative) growth, inflation as
the case may be. In Fig. 12.1 we have taken for granted the existence of
a model-no matter which-accepted by all countries. In the contrary case
there is no ground at all for carrying out the analysis.
A further problem is model uncerlainty. Assuming that all countries
agree on a model, it might happen that this model turns out to be wrong.
Cooperation based on an incorrect model might be negative rather than
positive.
In addition to these theoretical problems, there are practical problems,
the most important of which are the free-rider problem and the third-country
problem.
The Jree-rider problem is a typical problem of all cooperative equilibria
(inc1uding all kinds of agreements) in games of the prisoner's dilemma type.
Even if the various countries happen to be in a situation in which cooperation
is beneficial to all, the problem remains of how to ensure that all countries
participating in the cooperative equilibrium respect the agreement. In terms
of Table 12.1, each country-taking the other country's expansionary policy
for granted-has the incentive not to respect the agreement and adopt a re-
strictive policy, leaving the burden of expansion on the other country. Each
country tries to move from the last cell in the payoff matrix to a cell out-
side the main diagonal. Such a behaviour causes retaliation from the other
country, and we are back in the non-cooperative inferior situation. Hence
institutional mechanisms have to be devised and introduced to enforce the
co operative agreement.
The third-country problem is related to the fact that the agreement to
co ordinate usually inc1udes only a subset of the countries of the world. It
may then happen that the coordinated policies undertaken by the partici-
pating countries have a negative effect on non-participating countries. These
third countries could react by carrying out policies that negatively affect the
cooperating countries, which might then ultimately be worse off. For ex-
ample, suppose that country 1 and 2 cooperate and decide to deflate their
economies. The adverse effects on country 3's exports may lead this country
to deflate as weH. This aggravates the recession in country 1 and 2 above
what they had anticipated in deciding their coordinated policy, making them
worse off than if they had not jointly deflated.
These problems have given rise to a copious literature, that has amongst
others tackled the question of the consequences of uncertainty on the ''true''
model. The conc1usion has been that internationally coordinated policies
based on an invalid model can give rise to a lower welfare than the case
of no coordination. However, if we introduce the assumption that policy
makers have a learning ability (namely the ability of adjusting the actual
12.3. The International Debt Crisis 193
get the foreign exchange to service the debt did not materialize, partly be-
cause of the unfavourable international economic situation.
BOX 12.1 The causes of Mexico's debt default
Mexico was an oi! producing nation and therefore benefited from oi! shocks in the
1970s. Non-oi! exports, however, deteriorated rapidly during this period while im-
ports increased, contributing to a current account deficit. Other factors that led to
the current account deficit include the liberalization of trade and the revaluation of
the real exchange rate between 1978-8l.
To finance the widening current account deficit, Mexico's public and private sectors
relied on foreign debt, increasing interest payment abroad. When the second oi!
shock occurred in 1979, the US and the OECD countries responded by dramatically
increasing interest rates and imposing tight monetary controls. But the Mexican
government continued increased its public spending even as debt service became
more expensive. The Mexican private sector , however, began to shift its assets
abroad in the wake of the changes in the international economy. By August, Central
Bank reserves were almost exhausted, international banks refused to lend further
and the government announced its impossibility to service its debt. In the wake of
Mexico's default, most commercial banks reduced or stopped new lending to Latin
America. Because of the staggering accumulation of debt over the previous 10 years,
and the recent increase in the real and nominal interest rates, the interest payment
on debt were enormous. In 1984, interest due on debt as a ratio of gross national
product of the region reached 5%. To pay off their debts, Latin American countries
went through a long and painful process of adjustment. The IMF coordinated the
international credit aid and the debt renegotiations, which lasted unti! the late
1980s.
Among the causes of the debt crisis was the combination of high interest rates,
which exacerbated debt-service costs for Mexico and the other debtor nations, and
the sharp decline of oi! prices in 1982 that triggered the overall crisis. This was
coupled with the slowdown in world growth and the drop in commodity prices that
left exports stagnant and debt-service commitments hard to meet. Finally also the
private sector played an important role in causing the debt crisis. The majority of
the debt was held in non-guaranteed private sector loans which were nationalized
to meet obligations.
But to create a debt situation there must be two parties, the debtor and
the lender who supplies the funds. Hence we must examine the reasons why
the main international banks had so easily granted huge amounts of credit
to the countries under consideration. A widely shared thesis starts from the
oil shocks (see above). These had generated huge financial surpluses in the
OPEC countries, that had deposited them with the main international banks
in London and in the United States. The enormous funds received by these
banks raised the issue of how they could profitably utilise the money. A
good outlet was found to be the lending to less developed countries in the
course ofindustrializing their economies (the Newly Industrializing Countries
or NICs). The fact that the debt was incurred by governments or guaran-
teed by governments made the risk of default look fairly low. Paradoxically,
it were the increasing prices of the export goods of these countries (hence
increasing proceeds in foreign currency) rather than the industrialization to
give confidence to the lending banks, so much so that ab out one half of the
stock of debt outstanding in 1992 had been contracted in the previous two
12.3. The International Debt Crisis 195
years.
BOX 12.2 The Paris Club
The Paris Club is an informal group of official creditors with the role of finding coor-
dinated and sustainable solutions to the payment difficulties experienced by debtor
nations. Paris Club creditors agree to rescheduling debts due to them. Rescheduling
is a means of providing a country with debt relief through a postponement and, in
the case of concessional rescheduling, a reduction in debt service obligations. Since
the first meeting with a debtor country in 1956, the Paris Club or ad hoc groups
of Paris Club creditors have reached more than 360 agreements concerning over 70
debtor countries with a total amount of debt covered over $400 billion since 1983.
Although the Paris Club has no legal basis nor status, agreements are reached fol-
lowing a number of rules and principles settled by creditor countries, which help a
coordinated agreement to be reached efficiently. The basic principles layout that:
(1) decisions within the Paris Club have to be made on a case by case basis in
order to take into account the individuality of each debtor country and (2) with
the consensus of participating creditor countriesj (3) conditionality applies to debt
treatments, i.e. only for countries that need a rescheduling and that implement re-
forms to resolve their payment difficultiesj (4) solidarity ensures that creditors agree
to implement the terms agreed in the context of the Paris Club. Finally, (5) the
comparability of treatment between different creditors is preserved (in other words,
the debtor country cannot grant to another creditor a treatment less favorable for
the debtor than the consensus reached in the Paris Club). Among the different
types of debt, Paris Club agreements generally only apply to debts of the public
sector, medium and long term debts and credits granted before the "cutoff date",
i.e. when a debtor country first meets with Paris Club creditors, and not changed
thereafter. Prom the creditor side, the debts treated are credits and loans granted,
or commercial credits guaranteed by the Governments or appropriate institutions of
Paris Club creditors. Debt that was already treated in a previous Paris Club agree-
ment is normally not treated again, except for those countries where the financing
gap is large or where all pre-cutoff-date debt was already rescheduled. But with
the aim of producing agreements leading to sustainable levels of payments, longer
repayment periods have been considered over time, especially for poorer countries
(ranging from 23 to 40 years) and debt cancellation has been increasingly used.
The international debt crisis has given rise to serious problems, not only
for the creditor banks but also for the governments of the countries to which
these banks belong, and for the international monetary system. The pos-
sible bankruptcy of these banks, in fact, would create serious dangers for
both the national monetary system and the international monetary system.
Most cross-border debt contracts are denominated in the lender's currency,
so international debt and currency crises often coincide.
Although several general proposals and plans have been put forward (the
Baker plan, the Brady plan, and so on), in practice the international debt
problem has been tackled on a case by case approach, through a combination
of measures that can be summarized into three categories:
(a) modifications in the structure and nature of the debt, for example
by rescheduling, i.e. lengthening the time horizon of the debt (so that the
repayment instalments become smaller) or postponing the payment of inter-
est andj or principal to some date in the future. Another measure in this
category consists of debt-equity swaps, namely part of the debt is exchanged,
196 Chapter 12. Problems of the International Monetary System
Let us start by observing that the basic equation of the monetary ap-
proach to the balance of payments (see Sect. 7.1) can be written as
where TO is the given initial exchange rate. This equation lies at the basis
of the International Monetary FUnd 's (IMF) adjustment programs, whose
adoption is the condition for obtaining the FUnd's credit by countries in
balance-of-payments trouble that have applied for the Fund's support.
There are several reason why the FUnd adopts the monetary approach,
but the most compelling one is that data on monetary variables, in addition
to containing important macroeconomic information, are relatively more ac-
curate, easily available, and timely than other data, especially in developing
countries. The framework und~r consideration is a small open economy under
fixed (but adjustable) exchange rates.
FUnd-supported adjustment programs are aimed at the short run. Growth
problems are dealt with by the World Bank, whose growth model is of the
neoclassical type. Its basic equation is
(12.2)
because higher domestic prices, given the nominal exchange rate, imply a
competitiveness loss and hence an increase in the trade deficit (on the as-
sumption that the relevant critical elasticity conditions are satisfied: see
Chap. 6); the associated increase in foreign saving results in an increase in
investment (given the standard neoclassical assumption that all saving gets
automatically invested) and hence in real growth.
The intersection between the two schedules determines the simultaneous
monetary-real growth equilibrium, since from the equilibrium values tl PjJ ,
tly·, the equilibrium values of the other variables follow.
Given the targets of real growth, balance of payments improvement, re-
duction of inflation, this model suggests a number of policies. These include
demand management policies, exchange rate policies, structural policies, ex-
ternal financial support.
For example, demand management policies include a reduction in gov-
ernment current spending. This reduction, at the same level of fiscal revenue
and domestic credit expansion, implies an increase in saving and hence in
investment (remember that in the neoclassical context all saving gets auto-
matically invested). Since domestic credit expansion does not change, the
market for money is not influenced. The increase in output brought about by
the increase in investment has a depressive effect on domestic prices, hence
the new equilibrium will be characterized by higher growth and lower infla-
tion. Finally, if the substitution efIect prevails on the income efIect, and the
price efIects are sufficiently strong, the balance of payments improves. The
efIects on growth and inflation can easily be determined by Fig. 12.3, where
12.5. Proposals for the International Management of Exchange Rates 199
case of a sale) in the national money supplies. Thus the currency substi-
tution desires of the international agents, which are the cause of the excess
demands and supplies of the various currencies, give rise to changes in the
national money supplies while leaving the world money supply unchanged
and the exchange rates fixed. Hence currency substitution will have no effect
on the national economies.
This proposal has been criticized for various reasons. The first and fore-
most concerns the foundation itself of the proposal: currency substitution
seems to be neither the main cause of exchange-rate volatility nor the main
determinant of exchanges rates. Rather , it is asset substitution concerning
assets denominated in the various currencies that appears to have a much
greater role. Besides-the critics continue-by fixing nominal exchange rates
no room is left for real exchange-rate adjustments. These adjustments might
be required not so much because of differences in inflation rates (these could
not occur according to the proposal), but to offset different productivity
changes in the various countries.
(see Sect. 11.5.1), which before August 1993 could be considered as a target
zone with narrower margins, shows that the monetary authorities, even when
there are monetary cooperation agreements like in the ERM, are helpless
when credibility lacks.
The credibility problem has been theoretically studied in several mod-
els that can be divided into first-generation and second-generation mod-
els. First-generation models are based on simplified assumptions: economic
agents are convinced that the exchange rate will not go beyond the margins
and that the central parity will not be changed. Second-generation mod-
els are based on more general assumptions: economic agents assign non-zero
probabilities to both events (the exchange rate going beyond the margins and
the central parity being changed). For a survey of these models see Kempa
and Nelles (1999).
ability and effects. Let us for a moment suppose that it is enforceable and
examine its effects from the theoretical point of view.
The Tobin tax, when seen from the point of view of the agent engaged
in international capital movements, is a tax on the relevant foreign exchange
transactions, but can be translated into an equivalent tax on interest income.
It is, in fact, equivalent either to a tax on foreign interest income at a rate
which is an increasing function of the Tobin tax rate e, or to a negative tax
(i.e., a positive subsidy) on domestic interest income at a rate which is an
increasing function of e (see Gandolfo, 2002, Sect. R.3). Hence it acts by
suitably modifying the interest-rate differential which enters into the CIP
and UIP calculations. Here the opinions become divergent: on the one hand
there are those who maintain that its effects would be negligible, on the
other those who hold the opposite view. Both views are, however, based on
purely theoretical models without any empirical testing. Two exceptions are
Gandolfo and Padoan (1992) and Jeanne (1996).
Gandolfo and Padoan (1992) simulate the introduction of a Tobin tax
in their estimated continuous-time macrodynamic econometric model of the
Italian economy by suitably modifying the interest-rate differential on which
capital flows depend. In such a way the effects of this introduction on the
macroeconomic system can be examined taking account of all the dynamic
interrelations between the relevant variables. The results of the simulations
show that the introduction of a Tobin tax provides a crucial contribution to
the stabilization of the system with fuH capitalliberalization and speculative
capital flows. However there is more to it than that. A Tobin tax allows the
system to operate with a lower level of the domestic interest rate as it makes
the constraint represented by the foreign interest rate less stringent. This
obviously gives more room for domestic financial policy (in terms of e.g. the
financing of the domestic public debt).
Jeanne (1996) builds a target zone model in which an optimizing gov-
ernment is faced with a trade-off between its foreign exchange and domestic
objectives. The introduction of a Tobin tax would improve the credibility
of the peg by relaxing the foreign exchange constraint and reducing the cost
for the government of pegging the exchange rate. The author also applies
the model to the French franc and shows that the stabilizing effect of a 0.1 %
Tobin tax would have been quite sizeable.
Let us now come to the enforceability problem. The main practical ar-
gument against such a tax is wen known: if not an countries adopt it, then
the business would simply go to the financial centres where the tax is not
present (tax havens). Minor arguments concern the possibility of loopholes,
which could however by counteracted as soon as they are discovered.
Thus the real problem is generality of application. However, it has been
argued that it would be sufIicient that major dealing sites (the European
Union, the United States, Japan, Singapore, Switzerland, Hong Kong, Aus-
tralia, Canada and perhaps some other countries ) implemented the tax,
12.6. Suggested Further Reading 203
charging punitive tax rates for transactions crossing the border between "To-
bin countries" and ''tax havens". But the amount of international agreement
that this implementation would require makes it unlikely in the light of the
obstacles to international policy cooperation treated above, Sect. 12.2.2. In
fact, the formation of a "Tobin area" would ultimately be a manifestation
of political will, just as the formation of the European Monetary Union has
been (in spite of all its critics). And in any case a Tobin tax would be a much
less traumatic measure than the introduction of capital controls, which are
from time to time considered as a means of reducing international financial
instability.
26,529-38.
Sneddon Little J. and G.P. Olivei (eds.), 1999, Rethinking the International
Monetary System, Boston: FOOeral Reserve Bank of Boston, Conference
Series No. 43.
Tobin, J., 1974, The New Economics One Decade Older, Princeton: Prince-
ton University Press.
Tobin, J., 1978, A Proposal for International Monetary Reform, Eastern
Economic Journal 4, 153-9.
Tobin, J., 1996, Prologue, in ul Haq et al. (OOs.), ix-xviii.
ul Haq, M., 1. Kaul and 1. Grunberg (OOs.), 1996, The Tobin Tax: Coping
with Financial Instability, Oxford: Oxford University Press.
Various Authors, 1995, Policy Forum: Sand in the Wheels of International
Finance, Economic Journal 105, 161-92.
Williamson, J., 1985, The Exchange Rate System, revisOO edition, Washing-
ton (DC): Institute for International Economies.
Williamson, J., 1993, Equilibrium Exchange Rates: An Update, Washington
(DC): Institute for International Economies.
Part 111
207
208 Chapter 13. Orthodox Theory: Comparative Cost, Factor Endowments, Demand
in mind that here as in the subsequent examples, the cost of transport is as-
sumed to be absent, as its presence would complicate the treatment without
altering the substance of the theory. As we see, the unit cost of manufac-
turing cloth is lower in England than in Portugal while the opposite is true
for wine production. It is therefore advantageous for England to special-
ize in the production of cloth and to exchange it for Portuguese wine, and
for Portugal to specialize in the production of wine and to exchange it for
British cloth. Suppose, for example, that the (international) terms 0/ trade
(i.e., the ratio according to which the two commodities are exchanged for
each other between the two countries, or international relative price) equals
one, that is, international exchange takes place on the basis of one unit of
wine for one unit of cloth. Then England with 4 units of labour (the cost
of one unit of cloth) obtains one unit of wine, which otherwise-if produced
internally-would have required 8 units of labour. Similarly Portugal with 3
units oflabour (the cost of one unit of wine) obtains one unit of cloth, which
otherwise-if produced internally-would have required 6 units of labour.
In this example we have reasoned in terms of absolute costs, as one coun-
try has an absolute advantage in the production of one commodity and
the other country has an absolute advantage in the production of the other.
That in such a situation international trade will take place and benefit all
participating countries is obvious. Less so is the fact that international trade
may equally weIl take place even if one country is superior to the other in
the production of both commodities. The great contribution of the Ricardian
theory was to show the conditions under which even in this case international
trade is possible (and beneficial to both countries).
Now, this theory affirms that the necessary condition for international
trade is, in any case, that a difference in comparative costs exists. Compara-
tive cost can be defined in two ways: as the ratio between the (absolute) unit
costs of the two commodities in the same country, or as the ratio between the
(absolute) unit costs of the same commodity in the two countries. Following
common practice, we shall adopt the former, but they are totally equivalent.
In fact, if we denote the unit costs of production of a good in the two
countries by ab a2 (where the letter refers to the good and the numerical
subscript to the country: this notation will be constantly followed throughout
the book) and the unit costs of the other good by bl, b2 , then
(aI/bl = a2/~) -<==? (bI/al = ~/a2) -<==? (aI/a2 = bI/b2) -<==? (a2/al = b2/b l ),
13.1. Comparative Costs and International Trade: The Ricardian Theory 209
and similarly
(aI/bI ~ a2/b2) {::::=:? (aI/ a2 ~ bI/b2) {::::=:? (b2/a2 ~ bI/al) {::::=:? (b2/bl ~ a2/al)·
It therefore makes no difference whether the comparison is made between
aI/bI and a2/b2 or between aI/a2 and bt/b2, and so on.
The basic proposition of the theory under examination is that the con-
dition for international trade to take place is the existence of a difference
between the comparative costs. This is, however, a necessary condition only;
the sufficient condition is that the international terms of trade He between
the comparative costs without being equal to either. When both conditions
are met, it will be beneficial to each country to speciaHze in the production
of the commodity in which it has the relatively greater advantage (or the
relatively smaller disadvantage). Let us consider the following example.
In absolute value, the slope of this line equals the comparative cost in
country 1. Comparative cost and marginal rate oftransformation (or oppor-
tunity cost) are therefore one and the same thing.
In a similar way, we obtain the transformation curve of country 2. Con-
sider then Fig. 13.1, where we have brought together the transformation
curves of the two countries.
The line A'B' is the transformation curve of country 1, i.e. the diagram
of (13.2); in absolute value, tan a equals the comparative cost of country 1.
The line A" B" is the transformation curve of country 2, rotated anticlockwise
by 1800 and placed so that point B" coincides with point A"; it goes without
saying that 0" B" and 0' B' are parallel. The absolute value of tan ß equals
the comparative cost in country 2.
Let us take an arbitrary admissible value of the terms of trade, say tan {!,
and assume that international trade occurs at point E, whose coordinates are
the quantities exchanged. Country 1 specializes completely in the production
of commodity x, of which it produces the amount O'A'; of this, apart is
consumed domestically (0' D'), whilst the remaining part (D' A') is exported
in exchange for the quantity O'G' = ED' = G" B" of commodity y. Note
that, since the terms of trade are measured by tan p, and since (by considering
the right-angled triangle ED'A') we have ED' = D'A'· tan{!, it follows that
by giving D' A' of X, ED' of y can be obtained, and vice versa. This means
that the trade balance is necessarily in equilibrium. In fact, balance-of-trade
equilibrium, or value of exports=value of imports, requires
212 Chapter 13. Orthodox Theory: Comparative Cost, Factor Endowments, Demand
2The same problem would arise in the presence of many techniques, but limited in
number, of the fixed-coefficients type, such as are dealt with by activity analysis.
214 Chapter 13. Orthodox Theory: Comparative Cost, Factor Endowments, Demand
and the economic definition, we shall not claim the superiority of either one.
Here we shall use the physical definition.
In the following treatment, we assume that commodity A is capital inten-
sive relative to commodity Band that country 1 is capital abundant relative
to country 2; it goes without saying that B is labour intensive relative to A
and country 2 labour abundant relative to 1. Thus we must prove that coun-
try 1 will export commodity A whilst country 2 will export commodity B.
For this purpose we must use the transformation curves of the two countries.
Let us recall from microeconomics that the transformation curve (also
called the production possibilities curve) of a country shows the maximum
quantity of a commodity that can be produced for any given quantity pro-
duced of the other commodity, of course optimally employing all the available
factors of production. The (absolute value of) the slope of the transforma-
tion curve, called marginal rate of transformation, measures the (marginal)
opportunity cost of B in terms of A, namely the amount of A that the eco-
nomic system must forgo to obtain an additional unit of B3. It goes without
saying that the opportunity cost of A in terms of B is measured by the slope
of the transformation curve with reference to the A axis, and is the reciprocal
of the opportunity cost of B in terms of A.
It should be noted that, since the transformation curve is obtained through
an optimization procedure, the amount of A (or of B) that must be forgone
to obtain an additional unit of B (or of A) is the minimum possible given the
technology. The assumed concavity of the transformation curve implies that
its slope increases moving on the curve from left to right, namely the oppor-
tunity cost of B increases as the quantity of this commodity being produced
increases.
In Fig. 13.2 we have drawn the transformation curves of the two coun-
tries; their relative position refiects the assumption that country 1 is capital
abundant relative to country 2 and that commodity A is capital intensive
relative to commodity B.
It should be noted that it is not necessary for the two curves to intersect:
what matters is that they have a different slope along any ray through the
origin. If relative factor endowrnents were the same in both countries, then
their transformation curves would have the same slope (that is, an identical
opportunity cost) along any ray through the origin (in other words, they
would be radial blow-ups of each other); similarly, the ratio of the outputs
in the two sectors would be the same in both countries at any given common
commodity-price ratio. In such a situation, given the assumption of identical
structures of demand, there would be no scope for international trade.
The first step in our proof is to show that-at the same commodity-price
3Let us observe that this definition has a general validity, independently of the con-
text. For example, in the context of the Ricardian theory, it is possible to identify the
opportunity cost with the comparative cost.
216 Chapter 13. Orthodox Theory: Cornparative Cost, Factor Endowrnents, Dernand
o B
basic proposition on the pattern of trade rather than to determine the terms
of trade.
(a) the total amounts of the two factors existing in the economy;
(b) the distribution of these among the members of the economy,
namely the amounts of K and L owned by each member;
(c) the tastes of consumers;
(d) the state of technology, represented by well-behaved aggre-
gate production functions (a "well-behaved" production function
shows constant returns to scale and has positive but decreasing
marginal productivities).
I
I
I
I
A' ---L----
I
I
I
I
I
I
I
o B
the diagram we see that tano: > tanß, namely PB/PA is higher at H than at
E. Furthermore, OB' > OBE and OA' < OA E . This shows that the quan-
tity produced (and supplied) of B increases as its own relative price PB/PA
increases, while the quantity of A decreases. Conversely, the quantity of A
increases (and that of B decreases) as PA/PB increases.
Let us now come to the demand curves of the two commodities. In
general equilibrium the demand for each commodity depends on the price
of all commodities (in our simplified model with only two goods, on their
relative price) and on the individual's real income. This last is determined
by the marginal productivities of the factors (labour and/or capital) owned
by the individual: under perfect competition, in fact, the real unit reward
of each factor equals the factor's marginal productivity. These marginal
productivities, given the technology, ultimately depend on the relative price
of the commodities which has given rise to the outputs shown in Fig. 13.3.
It should be emphasized that these demand curves are different from the
usual Marshallian or partial equilibrium demand curves, which express the
quantity demanded of a good as a function of its (relative) price, and are
obtained on the ceteris paribus assumption, namely that everything else--
including (individual) income--is equal. On the contrary, in our derivation
income changes as PB/PA changes: in fact, when PB/PA is different, we are at
a different point on the transformation curve; therefore the marginal produc-
tivities of the factors will be different and, consequently, each individual's real
income will be different. In other words, the demand curves we are dealing
with are general equilibrium demand curves, which depend on real income
as wellas on relative prices; but, since real income depends on relative prices
220 Chapter 13. Orthodox Theory: Comparative Cost, Factor Endowments, Demand
PKKA +PLLA
PKKB+PLLB
from which
(13.4)
The left-hand side of (13.4) is the total income of all the individuals in the
ecoI,lomy (that they obtain by selling the services of the productive factors
they own). Since in this model income is entirely spent in buying commodities
A and B, we can write
(13.5)
where DA and DB are the quantities demanded of the two commodities.
Equation (13.5) is the aggregate budget constraint. From Eqs. (13.4) and
(13.5) it follows that the right-hand sides must be equal, as the left-hand
sides are equal. Therefore
(13.6)
whence
(13.7)
13.3. The Neoclassical Theory 221
which is true for any admissible value of PA and PB. The difference between
demand and supply is called excess demand and is to be taken with its sign, in
the sense that a positive excess demand means that demand is greater than
supply, while a negative excess demand (also called excess supply) means
that demand is smaller than supply.
Equation (13.7) states that the sum of the values of the excess demands
must be equal to zero, and is known as Walras' law. It is then immediately
obvious that, if a market is in equilibrium (zero excess demand), the other
must also be in equilibrium.
Walras' law teIls us something more, namely that if in a market there is an
excess demand with a certain sign, in the other market there must necessarily
be an excess demand with the opposite sign. For example, if DA - S A > 0
(positive excess demand in the market for commodity A, then D B - SB < 0
(negative excess demand, that is excess supply, in the market for commodity
B).
In general, given n markets linked by a (budget) constraint, Walras' law
implies that if n - 1 markets are in equilibrium, the nth must also be in
equilibrium. In our case there are only two markets, so that if one is in
equilibrium the other must also be: for example, if DA = SA then Eq. (13.7)
implies D B = SB, and vice versa.
the closed-economy equilibrium price ratios are different in the two countries;
without loss of generality we can assurne that this ratio is greater in country
2 than in country 1, as shown in the back-to-back diagram drawn in Fig.
13.4.
In this figure we have drawn the general-equilibrium demand and supply
curves for commodity A as functions of the relative price PA/PB in the two
countries: in the right-hand part there are the demand and supply curves for
commodity A in country 1, and in the left-hand part there are the demand
and supply curves for the same commodity in country 2. As assumed above,
the closed-economy equilibrium price-ratio in country 2 (0 RE) is greater
than in country 1 (0 PE).
It can be easily shown that when trade is opened up, commercial relations
are possible only if the international price ratio or terms of trade lies some-
where between the two internal equilibrium price ratios. We first observe
that with free trade, perfect competition and no transport costs, the same
commodity must have the same price everywhere (the law of one price), so
that the international and the national price ratios are the same. Now, for
terms of trade higher than ORE , both countries would supply commodity
PA/PB
S2A D2A
country 2 country 1
SIA
M 1A
D2A
PE
A 0 A
B B
Gi
Gz
Equivalently, this curve indicates the various terms of trade at which the
country is willing to trade.
In Fig. 13.5 we have drawn the offer curves of country 1 and country
2, DG l and DG 2 respectively, on the assumption that country 1 demands
commodity A in the international market, offering commodity B in exchange,
while the opposite is true for country 2. Each curve is normally increasing
and concave to its import axis.
Considering for example country 1, it will demand, say, OHA of commod-
ity A (a demand for imports) and offer OHB of commodity B in exchange
(a supply of exports). This gives rise to point Q of the OG l curve. At point
Q the relative price of goods (terms of trade) is measured by the slope of
the straight line connecting Q with the origin, namely by tan 0:: this relative
price, in fact, is given by the ratio of the quantities demanded and supplied,
namely by OHAjOHB = HBQjOHB = tano:.
International equilibrium is determined where the offer curves intersect,
namely at point E: country 1 demands OEA of commodity A, exactly equal
to the amount of A supplied by country 2, and supplies OEB of commodity
B, exactly equal to the amount of B demanded by country 2. International
trade will take place on the basis of OEB of B (exported by country 1 and
imported by country 2) for OEA of A (imported by country 1 and exported
by country 2); the equilibrium terms of trade are measured by tanß (slope
of the ray OE).
13.5. The Gains from Trade 225
E,
,
EA _______ IL __
, E'
I I
I I R P
o B
A A R
p
o b)
a)
Figure 13.7: Sodal indifference curves and the gains from trade: consumption
and production gains
The gains from trade can be given a more predse treatment if one is will-
ing to accept the concept of community or sodal indifference curves. The
problems raised by this concept are among the moot questions in welfare eco-
nomics. This notwithstanding, these curves are widely used in international
economics and we do not depart from general practice by using them as a
helpful expository device, though fully aware of their shortcomings.
In Fig. 13.7a the pre-trade (autarkic equilibrium) situation is depicted;
sodal welfare is maximized at point E, where a sodal indifference curve
is tangent to the transformation curve. In Fig. 13.7b, the terms-of-trade
line RR is drawn: the highest indifference curve attainable is that which is
tangent to this line, thus determining the consumption point Ec predsely,
as weH as the imported and the exported commodities and the amounts
traded (HBE~ of exports for HAEAof imports). The gains from trade are
13.6. The Four Core Theorems 227
MPK IA = MPK2A,
MPL IA = MPL 2A ,
(13.8)
MPK IB = MPK2B ,
MPL IB = MPL 2B ,
where M P K and M P L denote the marginal productivities of capital and
labour respectively, and the subscripts refer to the countries and commodities
as usual.
The importance of the assumption of incomplete specialization should
be noted here. In fact, if specialization were complete (for example, coun-
try 1 produces exclusively commodity A and country 2 commodity B), the
quantities MPK IB and MPL IB could not be defined in practice (because
commodity Bis not produced in country 1), neither could be MPK2A and
M P L2A (because commodity A is not produced in country 2); therefore Eqs.
(13.8) could not be written and the rest of the proof would fall.
Now, under perfect competition the equilibrium condition value of the
marginal product of a factor = price of the factor must hold. In symbols (re-
member that PA and PB are internationally identical) we have, with reference,
for example, to capital,
PAMPKIA = PIK,
PA MPK2A =P2K, (13.9)
PBMPKIB = PIK,
PBMPK2B =P2K,
from which---since the marginal productivities obey (13.8)-it follows that
PIK = P2K· In a similar way it can be shown that PIL = P2L. This completes
the proof of FPE.
Better to appreciate the importance of this theorem, it is sufficient to
realize that it shows that free trade in commodities is a perfect substitute for
perfect international mobility of factors. Note that, if perfect international
factor mobility existed as well, then perfect competition would necessarily
lead to international equalization of factor prices. But in our models we have
assumed absolute international immobility of factors, hence there might seem
to be no reason the factor prices to be equalized across countries.
13.6. The Four Core Theorems 229
R'
an increase in the demand for both commodities; since, as we have seen, the
output of B is lower, there will be an excess demand for this commodity
which will cause an increase in its relative price (PB/PA) and, consequently,
in its output. Therefore the new equilibrium point will be found in the
stretch Q" QIII of the curve T'T': only there, in fact, is the output of both A
and B higher than at Q. It can also be seen from the figure that at any point
included in this stretch, for example QE, the relative price of Ais lower, as
this price is measured by the (absolute value of the) slope of the RERE line
with respect to the A axis, which is smaller than the analogous slope of the
RR line.
All this concerns a closed economy. We now consider a trading open
economy, where we must distinguish the small country case from the case in
which the country is sufficiently large for its demands and supplies on the
world market to infiuence the terms of trade.
Let us assume, for example, that commodity A is the importable. We
further assume, for simplicity, that no commodity is inferior, so that, when
income increases, the demand for both A and B increases (each, of course,
increases by less than income). In the passage from Q to Q', the output
of commodity A has increased more than income whilst the output of B
has decreased. It follows that, within the country: (a) the excess demand
for A (demand for imports) decreases, as output has increased more than
demand; (b) the excess supply of B (supply of exports) decreases, as output
has decreased whilst demand has increased. Therefore, on the world market,
at the given world relative price, there will be a decrease in both the demand
for A and the supply of B.
It is at this point that the distinction between the small and the large
country case becomes relevant. In the former case the terms of trade do not
change, and the country will go on producing at Q'.
In the large country case, the excess supply of A on the world market (due
to the decrease in the country's demand for imports), and the correlative
excess demand for B (due to the decrease in the country's supply of exports)
will cause changes in world prices, since the excess supply of A will put a
downward pressure on PA and the excess demand for B an upward pressure
on PB; therefore the terms of trade PB/PA increase. This confirms the closed-
economy result. Note that, since we have assumed A to be the importable,
the terms of trade have improved.
Let us now consider the case in which the importable is commodity B,
maintaining the assumption that there are no inferior goods. When the pro-
duction point shifts from Q to Q', the consequences for the country will
be: (a) the excess supply of A (supply of exports) increases, since its out-
put (which increases by more than income) increases by more than demand
(which increases by less than income); (b) the excess demand for B (demand
for imports) increases, because output decreases whilst demand increases.
Therefore-Ieaving aside the small country case-on the world market at
232 Chapter 13. Orthodox Theory: Comparative Cost, Factor Endowments, Demand
unchanged prices there will be an increase in both the supply of A and the
demand for B and so-since the initial situation was of equilibrium-an ex-
cess supply of A and an excess demand for B. This will cause a decrease
in PA and an increase in PB, so that PB/PA will increase, again confirming
the closed-economy results. As B is the importable, the terms of trade have
moved against the country.
The relevance of Rybczynski's theorem in international trade theory lies
in its use to examine the effects of growth, when the cause of growth is an
increase in factor endowments.
speak of factor trade when we are in the presence of the mere international
mobility of primary factors. Both capital and labour can be considered under
this heading, land being immobile by its very nature.
Trade in factors, on the other hand, implies that we are dealing with
factors which are themselves produced means of production and, in addition
to being internationally mobile, can be traded as any other good. This
practically restricts the picture to (physical) capital in its various forms,
both fixed and intermediate.
Our distinction is neither semantic nor whimsical, as it has important
consequences. Suffice it to point out that, in the case of mere factor mobil-
ity, when factor prices are equalized through factor movements, factors stop
moving. On the contrary, in the case of trade in factors, when the prices of
traded factors are equalized through free trade, these factors (in their quality
of traded goods) continue to move as any other traded commodity.
The second caveat is that by factor movements we mean the physical
movement of factors from one country to another, not the mere change of
ownership of an otherwise immobile factor. Suppose that a resident of coun-
try 1 buys real estate in country 2, or that a firm residing in country 1 buys
aplant located in country 2 and operates it there with local labour and
management: these are not factor movements from our point of view. They
may imply a movement of other commodities (in the case of barter) or more
realistically a financial capital flow from country 1 to country 2 (named a
direct investment when the purchase refers to a plant), but financial capital
is not considered by the pure theory of international trade. This is why we
feel that the study of direct investment and multinational corporations is
best carried out in the context of international monetary economics.
PAMPLA = PL,
(13.10)
PBMPLB = PL,
where M P LA, M P L B are the (physical) marginal products of labour in the
two sectors, and PL is the wage rate. Letting w = PL/PA denote the real
wage rate in terms of commodity A, and P = PB/PA the commodity price
ratio, we have
MPL A =W,
(13.11)
pMPL B =W,
hence
MPLA = pMPL B, (13.12)
which determines the optimal allocation of labour between the two sectors
and hence-since the two stocks of specific capital are also fully employed-
the outputs of the two commodities for any given p.
Equation (13.12) can be given a simple graphie representation. In Fig.
13.9, the total amount of labour is measured by the segment OAOB. The
quantity of labour used in sector A is measured from the origin 0 A, while
that used in sector B is measured from OB. In the ordinate we show the
real wage rate w. Curves L~, L~ represent the demand-for-labour schedules
in the two sectors, derived from Eqs. (13.11) for a given p. The equilibrium
condition (13.12) obtains at point E. This determines the equilibrium real
wage rate WE and the optimal allocation of labour, which consists of 0 ALE
employed in sector A and OBL E employed in sector B.
Let us assume that the general-equilibrium commodity price ratio is that
corresponding to curve L~, say Pb and consider the introduction of inter-
national trade in a two-country framework. The condition for international
trade to take place is that P2, the closed-economy commodity price ratio in
country 2, is different from PI. Without loss of generality we can assume that
13.9. Suggested Further Reading 235
W w
:B :::::::::::::::~:::::::::)......... .. ...................................... WB
B LD . LD
B ~ A
L'E
P2 > PI, hence the post-trade price ratio p* will be somewhere in between.
Thus we can take E' as the post-trade equilibrium in country 1. In country
2 there will be a downward shift of the demand for labour in sector B, since
p* < P2. This shows that there will be an increase in sector B's output in
country 1 and in sector A's output in country 2.
We now consider factor rewards. From Fig. 13.9 we see that an increase
in P (the relative price of commodity B) causes more labour to be used in
sector B and less in sector A. The (specific) capital to labour ratio decreases
in sector Band increases in sector A. Since the marginal productivity of
capital (which is the real unit reward of capital) depends negativelyon the
capital to labour ratio, it follows that the marginal productivity of capital
increases in sector B and decreases in sector A.
The effect on the ubiquitous factor is however ambiguous. The wage rate
does, in fact, increase in terms of commodity A (from WE to WB)' but declines
in terms of commodity B (since the marginal productivity of capital is higher
there, the marginal productivity of labour is lower). Whether wage earners
are better or worse off depends on the composition of their expenditure, a
result that has been dubbed the neoc1assical ambiguity in trade theory.
Davis, D.R. and D.E. Weinstein, 2001, An Account of Global Factor Trade,
American Economic Review 91, 1423-53.
Heckscher, E.F., 1949, The Effect of Foreign Trade on the Distribution of
Income, in H.S. EHis and L.A. Metzler (eds.), Readings in the Theory
oi International Trade, Philadelphia: Blakiston (English translation of
the original 1919 article)
Leamer, E.E. and J. Levinsohn, 1995, International Trade Theory: The Evi-
dence, in G. M. Grossman and K. Rogoff (eds.), Handbook oi Interna-
tional Economics, Vol. III, Amsterdam: North-Holland.
Leontief, W.W., 1953, Domestic Production and Foreign Trade: The Amer-
ican Position Reexamined, Proceedings oi the American Philosophical
Society 97, 332- 349. Reprinted in: W. Leontief, 1966, Input-Output
Economics, New York: Oxford University Press, Chap. 5 (pages refer
to this reprint).
Leontief, W.W., 1956, Factor Proportions and the Structure of American
Trade: Further Theoretical and Empirical Analysis, Review oi Eco-
nomics and Statistics 38, 386-407; reprinted in: W. Leontief, 1966,
Input-Output Economics, op. cit., Chap. 6.
Negishi, T., 1982, The Labour Theory of Value in the Ricardian Theory of
International Trade, History oi Political Economy 14, 199-210 ..
Ohlin, B., 1933, Interregional and International Trade, Harvard University
Press (revised edition, 1967).
Ricardo, D., 1817, On the Principles oi Political Economy and Taxation,
London: J. Murray, Chap. VII.
Rybczynski, T.M., 1955, Factor Endowments and Relative Commodity Prices,
Economica 22, 336-41.
Samuelson, P.A., 1948, International Trade and the Equalization of Factor
Prices, Economic Journal 58, 163-84; reprinted in P.A. Samuelson,
1966, Collected Scientific Papers, Cambridge (Mass), MIT Press, Vol.
11.
Stolper, W.F. and P.A. Samuelson, 1941, Protection and Real Wages, Re-
view oi Economic Studies 9,50-73; reprinted in P.A. Samuelson, 1972,
Collected Scientific Papers, Cambridge (Mass), MIT Press, Vol. 111.
Torrens, R., 1815, An Essay on External Corn Trade, London: J. Hatchard.
Wong, K.-Y., 1995, International Trade in Goods and Factor Mobility, Cam-
bridge (Mass): MIT Press.
Chapter 14
14.1 Introduction
This chapter is concerned with what is called the theory 0/ commercial policy
in the broad sense. The traditional theory focused on tariffs, starting from
two principles generally accepted until the first world war. These were: (a)
that impediments to international trade for protectionist purposes should be
limited to tariffs, and (b) that no commercial discrimination between supplier
countries should be instituted, in the sense that, if a tariff is levied on some
imported commodity, it should be applied at the same rate and to all imports
of that commodity independently of the supplying country.
Notwithstanding the fact that in the inter-war period, and especially dur-
ing the Great Depression, these principles were systematically violated, they
were taken up again and made the foundation of the international agreement
that, it was hoped, was to rule international trade after the second world
war: GATT (the General Agreement on Tariffs and Trade). Several interna-
tional meetings for the purpose of negotiating multilateral tariff reductions
took place under the aegis of GATT (now replaced by WTO, the World
Trade Organization, on which see Sect. 3.6.3) which, however, had to take a
permissive attitude towards the violations of the above principles. The last
few decades have seen an expansion of both non-tariff barriers to trade and
discriminatory commercial policies (preferential trading agreements etc.), so
that the traditional theory has had to be broadened to make the rigorous
analysis of these phenomena possible.
237
238 Chapter 14. Tariffs, and Non-Tariff Barriers
IThe symbol generally used for the tariff rate is t. However, since the symbol t is also
used to denote time, another symbol (d, from duty) has been used to indicate the tariff
rate.
14.2. Effects of a Tariff 239
the tariff rate) and not of a specific tariff (so many dollars per unit of the
commodity).
price
which can be measured as the area under the demand curve included between
the line of the price, the price axis and the demand curve itself. For example,
in Fig. 14.1, consumers' surplus is measured-when the price is p and the
quantity q4-by the area of triangle N H R.
Now, with the increase in price from p to p(l + dd, consumers' surplus
decreases by N H H1M. This is a cost; to compute the net cost, if any, we
must calculate the benefits. These are the tarifI revenue accruing to the
government, F1F{HfHl, and the increase in producers' surplus3 MNFF1.
It is important to stress that, in order to be able to net out benefits from
costs (both are expressed in money, and so are dimensionally comparable)
we must assume that each dollar 0/ gain or loss has the same importance
independently 0/ who is gaining or losing. Without this assumption, in fact,
it would not be possible to compare the consumers' loss with the producers'
and the government's gain.
Given this assumption, it can readily be seen from the diagram that the
reduction in consumers' surplus is only partly ofIset by the tarifI revenue
and the increase in producers' surplus: we are left with the areas of the two
triangles F F{ F1 and Hf H H1 , which represent the social costs of the tarif!.
The first one, F F{ F1, measures the production cost of protection. If the
country had imported an additional amount ql q2 at the price p, its cost would
have been ql q2F{ F. Instead the country produces this amount domestically,
with an additional cost measured by the increase in the area below the supply
curve, ql q2Fl F. The difIerence F F{ F1 represents the cost of the misallocation
of resources caused by the tarifI: in fact, if the country had used an amount
of resources equal in value to qlq2F{F to increase the output of its export
industry (not shown in the diagram), with the consequent increase in exports
it could have obtained ql q2 more of the imported commodity. When instead
it increases the domestic production of this commodity, the country roust
use a greater amount of resources (equal in value to qlq2FlF) to obtain the
same additional amount (qlq2) of the commodity.
The second one, Hf H H1 , measures the consumption cost of protection,
due to the fact that the tarifI brings about an increase in the domestic price
of the total priee that consumers would be wiIling to pay rather than go without the
commodity, over that which they actually pay. The graphie measure used in the text is
only one of the measures possible and hinges on several simplifying assumptions, &mongst
which the constancy of the marginal utility of money. It should also be stressed that
consumption and consumer should be interpreted in the broad sense to mean purchase
and purchaser respectively, for whatever purpose the product is bought.
3Unlike consumers' surplus, this is a well-defined concept, as it is a synonym for the
firms' profit (difference betwen total revenue and total cost). If we neglect the fixed cost
(which has no consequence on the variations), the total cost of any given quantity, say qI,
is the area under the marginal cost (i.e. the supply) curve from the origin to the ordinate
drawn from that quantity (OVFql). As total revenue is ONFqI, producers' surplus is
V N F. If we consider an increase in output from ql to q2, the increase in producers' surplus
is VMFI - VNF = MNFF1 •
242 Chapter 14. Tariffs, and Non-Tariff Barriers
14.4.1 Quotas
An (import) quota is a quantitative restrietion (so many cars of a certain
type per unit of time) imposed by the government on the imports of a cer-
tain commodity and, therefore, belongs to the category of direct controls on
international trade. For this purpose the government usually issues import
licences (which it can distribute to importers according to various criteria)
but other forms are possible.
The effects of a quota can be analysed by means of a diagram similar
to that used in Sect. 14.2 (see Fig. 14.1) to analyse the effects of a tariff.
In Fig. 14.2, p is the world price of the commodity, of which a quantity
ql q4 is imported under free trade. The government now decides that imports
have to be reduced, for example from ql q4 to q2q3 and, accordingly, decrees
a quota. The domestic price of the commodity will rise to p', since the
(unsatisfied) excess demand by domestic consumers will drive it up from
p to the level at which the actual excess demand is exactly equal to the
given quota, F1H1 = q2q3. The effects of a quota on domestic price, output,
price
consumption, and on imports, are the same as those which would occur if a
tariff were imposed such as to cause an increase in the domestic price from
p to p': this can be readily seen by comparing Fig. 14.2 with Fig. 14.1. The
equivalent tarifj rate can be computed from the equation p' = (1 + dt}p.
There is, however, a difference between a quota and an equivalent tar-
iff: whilst in the case of a tariff the government collects a fiscal revenue
(FIF{H~Hl in Fig. 14.1), it now collects nothing and the quota gives rise to
a gain of equal size (FIF{H~Hl in Fig. 14.2) accruing to the quota holders
(this is true under assumption that the country is small and that there is
perfect competition among the foreign exporters. In the opposite case, these
could avail themselves of the occasion of the quota to raise the price charged
244 Chapter 14. Tariffs, and Non-Tariff Barriers
to domestic importers, thus depriving them of part of the gain under consid-
eration). Now, why should the government deprive itself of a fiscal revenue
if the same quantitative restriction and the same effects of a quota can be
obtained by a tariff?
Let us first observe that, in principle, the government could sell the import
licences by auction: with a perfect auction in a perfectly competitive market,
the revenue of the auction would be exactly the same as that of the equivalent
tariff. This is so because competition between importers to get hold of the
licences will induce them to make higher and higher bids until extra profits
(which are equal to FIF{H~Hl) disappear in favour ofthe government. But
this is a theoretical possibility difficult to realize in practice.
The answer to the above quest ion can be found in the fact that only a
quota gives the certainty of the desired quantitative restriction on imports,
which is lacking in the case of a tariff for various theoretical and practical
reasons, among which:
1) the equivalence of the effects on imports depends on the existence of
perfectly competitive conditions at horne and abroad: in the opposite Case, in
fact, the effects of a tariff and of a quota can be very different. For example, if
foreign exporters do not operate under perfect competition, they may reduce
the price in order not to lose market shares when the horne country imposes
a tariff, so that the increase in the domestic price will be smaller than that
required to achieve the desired reduction in imports.
2) A quota, unlike a tariff, can have important effects on the market
structure of the country which imposes it, far it can convert a potential into
an actual monopoly, that is, enable the domestic industry, fully protected
from foreign competition by the quota, to establish a monopoly. In fact, let
us assume that in the country there is a potentially monopolistic industry. In
the presence of a tariff, this industry cannot raise the price above the world
price plus tariff, for its sales would drop to zero (domestic consumers will buy
solely imported goods if the domestic importable has a price higher than the
world price plus tariff). If instead of the tariff the country decrees a quota,
the potential monopoly can become an actual one, because the domestic
industry can now raise the price without danger of its sales dropping to zero,
as imports cannot exceed the quota.
3) The computation of the tariff (d 1 in our example) which brings about
exact1y the desired reduction in imports can be made only if the curves D
and S are known exactly and do not shift unpredictably. Notwithstanding
the advances in econometrics, these curves can be determined only within
a (usually large) confidence interval. Furthermore, the possibility of (large
though predictable) shifts in these curves (because the underlying exogenous
factors change in a known way) compels the government to compute, levy
and enforce changing tariff rates.
14.4. Quotas and Other Non-Tariff Barriers 245
various producers.
This ideal allocation is not, however, easily realized in practice. In the
real world the production is apportioned on the basis of negotiations among
the members of the cartel, each of whom has its own interests and different
contraetual force. The more influential and skilful negotiators will probably
get a greater quota than the optimum corresponding to the application of the
principle of equalization of marginal costs, even if this will raise the cartel's
total cost of produetion.
We must now consider the more realistic case in which the cartel does
not include all but only part of the producers, so that besides the cartel,
also independent (i.e., not belonging to the cartel) competitive producers are
present in the world market for the commodity. These latter will have to
accept the price fixed by the cartel, but the cartel will have to take their
supply into account when fixing the price. The market form obtaining here
is quasi-monopoly. In Fig. 14.4, in addition to the world demand curve D,
we have drawn the aggregate supply curve S of the independent producers.
If we subtract, for any given price, S from D laterally, we obtain D', which
is the demand curve for the cartel's output. For example, at price ON, the
supply of the independent producers is N N': if we subtraet M M' (equal, by
construction, to NN') from world demand NM', we obtain segment NM,
that is the quantity that the cartel can sell at price ON.
Once the D' curve has been derived, the cartel can behave along it as a
monopolist and will maximize profits by the usual rule, that is, by equating
marginal cost to marginal revenue (the latter will, of course, be that con-
cerning curve D'). The cartel, therefore, will fix the price at 0PE and sell a
quantity PEH = OqE, whilst the independent producers will sell a quantity
HH' = qEq'.
One can easily check, by drawing the marginal revenue curve concerning
14.4. Quotas and Other Non-Tariff Barriers 247
curve D (which we leave as an exercise for the reader), that the price is lower
and the quantity sold greater than in the case of a monopolistic cartel. It
is also possible to check graphically that the greater the elasticity of the
supply curve of independent producers S, the smaller the cartel's markup.
More precisely, the (price) elasticity of the D' curve (denoted by 'f/c) depends
on the elasticity ofthe D curve ('f/w), the elasticity of the S curve ('f/s), and on
the cartel's share in the total consumption of the commodity (k), according
to the formula
'f/w + (1- k) 'f/s
'f/c = k . (14.2)
(14.3)
d) the members of the cartel must accept and adhere to the official deci-
sions taken by the cartel (by means of majority voting or some other way)
as regards price and output.
Condition (d) is essential for the life itself of the cartel. If, in fact, the
members begin to decartelize by selling greater amounts (than those allot-
ted to each) at lower prices, the cartel will soon break up. But why should
there be any incentive to behave in this manner? The answer is that, though
the profits of the cartel as a whole are maximized by respecting the official
decisions, the single member can obtain vastly greater profits by slightly
lowering the price below the official one, provided that the other members
adhere to the official price. In fact, buyers will be willing to buy all the
quantity demanded-previously bought from the cartel-from the single pro-
ducer who charges a slightly lower price, so that the demand curve facing
this single producer is in practice almost perfectly elastic. This producer will
therefore reallze increasing profits by increasing output, because his selling
price is greater than his marginal cost 4 , and he can seIl increasing amounts
without further reducing the price. He will therefore profit from increasing
output up to the point where his marginal cost has increased to the level of
the selling price charged by him.
Naturally, greater profits for the single producer who does not adhere to
the official price mean lower profits for the other cartel-abiding members, but
the single producer, especially if relatively small, can always hope that the
other members will not become aware of his infringement or will not react.
If, for example, his share in the cartel's output is 1%, he may think that
a 50% increase in his output (this means that his share goes up to 1.5%)
will cause so small a loss (spread out through all the other members) as to
be negligible. This is undoubtedly true, but if the same idea occurs to a
sufficient number of members and is put into practice by them, the cartel
dissolves. Therefore the cartel, to persist, must be able to put pressure (of an
economic or political or some other nature) on the single members to make
them adhere to the official decisions.
But unfaithful members are not the only cause of the dissolution of a
cartel. There are at least three other motives leading to a progressive erosion
of the markup (and so of the profits) of the cartel. They can be analysed
with reference to formula (14.3) and are:
1) The increase in "lw. Even ifworld demand is sufficientlyrigid when the
cartel is set up, the very success of the cartel, paradoxically, helps to make
this demand more elastic. As a consequence of the (usually very large) price
4We must remember that in the initial situation the official price fixed by the cartel is
higher than the marginal cost (this is true in both the monopolistic and quasi-monopolistic
cartel). From the point of view of the cartel as a whole, it is not profitable to reduce the
price (this, in fact, would lead to lower profits), whilst the single member can-for the
motives explained in the text~btain higher profits by slightly lowering his selling price
below the official one; this lower price is nevertheless higher than his marginal cost.
14.4. Quotas and Other Non-Tariff Barriers 249
increase, buyers will put their every effort into the search for substitutes
for the cartelized commodity (it suffices to mention the search for energy
sources alternative to oil and the research into energy-saving production pro-
cesses and commodities that were set into motion as a consequence of the
Organization of Petroleum Exporting Countries-OPEC--cartel) and so "'W
increases.
2) The increase in "'S' Even if independent producers' supply is rigid
when the cartel is set up, the success itself of the cartel, again, helps to make
this supply more elastic, since these producers will multiply their efforts to
increase output. If the cartel concerns an agricultural commodity, such as
sugar or coffee, the price increase yvill induce independent producers to shift
increasing amounts of resources (land, labour, capital) to the production of
the cartelized commodity. If an exhaustible natural resource is concerned,
such as oil or copper , independent producers will multiply their efforts to
find new fields. Similar efforts will also come from countries previously not
exploiting the resource. These efforts, if successful, will increase not only the
output but also the number of independent producers (think of the oil fields
found by England under the North-Sea). All this causes an increase in "'S'
3) The decrease in k. In order to increase the price without building up
excessive inventories of the commodity, the cartel must restrict output and
sales relative to the pre-cartel situation. This, coupled with the efforts of
independent producers (point 2), leads to a decrease in k.
These three forces jointly operate to erode the cartel's monopolistic power.
Also, note that as the markup is wearing away, the incentive for the single
members to decartelize (see above) becomes greater and greater.
Economic theory, therefore, predicts that, in the long run, any cartel is
bound to dissolve, even if new cartels are always being set up, so that at any
moment a certain number of cartels is in existence. Historical experience
seems to confirm this conclusion, even in the most dramatic cases. Among
these one must undoubtedly count the cartel which gathers the main oil pro-
ducing countries into OPEC. Conditions (a), (b) and (c) above certainly held
in 1973: very rigid world demand for oil, low elasticity of the supply of inde-
pendent producers, high share (above 50%) in world production controlled by
the cartel. Furthermore, for various political motives, the degree of cohesion
of the cartel was high.
The great initial success of OPEC is, therefore, not surprising. However,
forces 1), 2) and 3), slowly but steadily got down to work.
The high price of oil set into motion o'r intensified the search for alterna-
tive energy sources, for productive processes less intensive in energy, for less
energy-consuming commodities and ways of life (energy-saving cars, limits
to domestic heating, better insulation of new buildings, etc.) began or was
intensified. As a consequence, the share of oil in world energy consumption
decreased, and energy consumption per unit of real GDP fell in industrial
countries as a whole.
250 Chapter 14. Tariffs, and Non-Tariff Barriers
Another element that reinforced the drop in demand for oil was the world
depression which, by slowing down (and sometimes by causing a decrease in)
the level of activity in the various industrialized countries, reduced their
energy needs. The supply of independent producers steadily increased (the
case of England, which became a net exporter of oil, is sensational). The
cartel's share in the world market decreased weil below 50%.
As a consequence of all this, cases of members not adhering to the cartel's
official decisions were not lacking, often not because of greed, but out of sheer
necessity (many OPEC countries had set up development programs based on
estimates of an increasing-or at least not decreasing-fiow of oil revenues
in real terms, and found themselves in trouble when this fiow started to
decrease).
cost is 0.8 x 2.5% = 0.02 dollars per dollar of imports, which is equivalent to an
ad valorem tariff with a 2% rate.
(f) Government Procurement. Governments buy a large amount of goods
and services, and usually prefer to buy domestic rather than equivalent foreign
goods of the same price (in some cases they are allowed by domestic legislation
to buy domestic goods even if equivalent foreign goods have a lower price, not
below a certain percentage)j besides, governments may have recourse to aseries of
techniques aimed at limiting the opportunity for foreign producers to tender for
the supply of goods to the public sector . All this amounts to a discrimination in
favour of domestic producers, which restricts imports.
(g) Formalities of Customs Cleamnce. These are connected with the im-
position of tariffs, such as the classification and evaluation of the commodities in
transit at the customs and other bureaucratic formalities. A more rigid application
of these formalities hinders trade and involves a cost for importers.
(h) Technical, Safety, Health and Other Regulations. Countries often have
different regulations, and this is in itself an impediment to international trade, for
producers have to bear additional costs to make the commodities conform to the
different regulations, according to the country of destination. Besides, a country
may use these regulations to reduce or even stop the imports of certain commodities
from certain countries, for example, by checking with particular meticulousness
and slowness their conformity to the regulations, or even by issuing regulations
which actually prevent the acceptance of certain foreign commodities. This is
called regulatory protectionism.
(i) Border Tax Adjustments. Governments usually levy an "import equalization
tax" on imported goods equal to the indirect tax levied at home on similar goods
domestically produced and, vice versa, they give back to exporters the national
indirect tax. This may cause distortions if the import equalization tax is higher
than the national indirect tax (the difference is a covert import duty) or if the sum
returned to exporters is greater than the amount of the national indirect tax (the
difference is a covert export subsidy).
253
254 Chapter 15. Free Trade vs Protection, and Preferential Trade Cooperation
for the commodity on the world market (this implies that the tariff-imposing
country is a large country).
This reduces the cost of protectiori, and it is even possible that an im-
provement, instead of a social cost, takes place in the tariff-imposing country.
This possibility is illustrated in Fig. 15.1, which is based on Fig. 14.l.
As a consequence of the tariff, the world price decreases, far example to
p', so that the cum-tariff domestic price is p'(1 + d1 ), lower than p(1 + d1 ).
The decrease in consumers' surplus is measured by N H H1M. On the side of
benefits we count as usual the increase in producers' surplus (MNFF1) and
the increase in the government's fiscal revenue, F1F{' Hr H1. For convenience
of analysis let us break this rectangle in two parts: F1Fr Hr H1 = F1F{ H~ H 1+
F{ F{' H~' H~. The first of these, added to producers' surplus, leaves the two
triangles F F{ F 1 and H~ H H 1 (which in the previous case measured the cost
of protection) unaccounted for. But now on the side of benefits there is also
the area of the rectangle F{ F{' H~' H~, which is far greater than the sum of
price
o quantity
the areas of the two aforementioned triangles: the balance between benefits
and costs is now positive. It follows that the tarijJ has brought about a net
benefit to the country that imposes it!
It can be readily seen that the reason for this benefit lies in the decrease
in the world price, which means that foreign exporters have eventually taken
part of the burden of the tariff upon themselves. In fact, with respect to
the pre-tariff situation, domestic consumers are subjected to an increase in
15.2. The Infant Industry 255
the price of the commodity equal to MN only: the remaining part of the
absolute amount of the tariff (N NI) is indirectly paid for by foreign exporters
in the form of aprice decrease, so that it is as if the amount F{ F{' Hr Hf had
been paid out by these exporters.
By exactly knowing the situation of international demand and supply it
is then possible to calculate an optimum tariff, such that the tariff-imposing
country obtains the maximum welfare increase.
Where is the catch in this apparently impeccable reasoning? It lies in the
fact that the reaction of the rest of the world is neglected. When we bring
this reaction into the picture, results change dramatically. The exporting
country, in fact, is obviously harmed by the tariff, and will retaliate, with
the result that at the end of the trade war every country will be worse off.
On the settlement of trade disputes see Chap. 16.
15.3 Distortions
We are considering distortions in the commodity markets and distortions in
the factor markets.
By distortions in domestic goods markets we mean all those situations in
which the domestic relative price of commodities does not reflect, as it should,
the marginal rate of transformation. These distortions may be due to mo-
nopolistic elements (which make the selling price higher than the marginal
cost) or to external economies or diseconomies (which make the producer's
marginal cost different from the social marginal cost, that is, cause a diver-
gence between the private and the social marginal cost).
When the domestic relative price and the marginal rateof transformation
are unequal, free international trade may even cause a decrease in welfare
with respect to the autarkic situation, for example because the distortion has
induced the country to specialize in the wrong direction due to the wrong
signals coming from the distorted prices. Now-so the protectionist argument
runs-the introduction of a tariff would stimulate the production of the
commodity in which the true comparative advantage lies, thus increasing the
country's welfare. But the imposition of a tariff, which in this case involves
a production gain (deriving from a better allocation of resources), causes a
consumption loss, so that the net result can be, in general, either a loss or a
gain.
Let us now come to distortions in domestic factor markets. These distor-
tions imply that the equality between the price of a factor and the value of
its marginal productivity andj or the equalization of the price of a factor in
all sectors do not hold. This will lead to an inefficient allocation of resources
and, consequentIy, the country will not be on its true transformation curve,
but on a lower curve.
In other words, the distortions under consideration prevent the country
from reaching the conditions of efficiency, which require that the marginal
rate of technical substitution (given by the ratio between the marginal pro-
ductivities of the two factors) should be equal in both sectors and equal to
the (common) factor-price ratio.
The prescription is the same as that given in the previous case, namely
the introduction of a new distortion (the tariff) could offset the existing
distortion. But also in this case both an increase and a decrease in welfare
are possible.
15.4. The Theory of Second Best 257
altitude
v
possible point, our climber may have to go back quite a long way if the
second highest hill is a great distance from the very highest. In terms of Fig.
15.2, the climber arrives at B and sees that the way to V is blocked by an
insuperable gorge at D. Then, instead of staying at B or, worse, walking
towards V as far as D, the climber will have to backtrack to A to reach the
second highest point.
Now, if we apply the theory of second best to the free-trade-versus-
protection debate, it immediately follows that, in the real world, it is not
possible to ascertain apriori whether a protectionist poliey improves or wors-
ens the situation nor is it possible to state that any movement towards freer
trade automatically gives rise to an improvement.
Similarly, it is not possible to state, as the traditional theory goes, that
there exist other policies decidedly better than the imposition of a tariff. This
statement, in fact, is certainly true only if all the violations of the Pareto-
optimum conditions are eliminated.j a particular ease occurs when there is
only one violation (for example a distortion in the factor market or in the
goods market), in which case the elimination of the violation without the
introduction of others restores the optimum situation for certain (in terms
of Fig. 15.2, if our climber is at C, the last step uphill will certainly bring
this person to V). This has implicitly been the line of reasoning followed.-in
accordance with traditional theory-in the previous sections.
But if, as is true in the real world, there are numerous violations of
the Paretian conditions, it follows from the theory of second best that it is
not possible to state for certain that a policy which eliminates one of these
without introducing another violation is better than a policy which eliminates
the same violation by introducing another.
It is clear that, things being so, it becomes impossible to make statements
valid in general and deduce apriori policy prescriptions from a limited num-
15.5. Preferential Trading Cooperation 259
ber of guiding principles. In reality any outcome is possible and one must
ascertain which is the best policy (free trade or protection in its various forms)
in each actual case without being blinded by theoretical preconceptions.
union, country 1 applied a 30% tariff on all imports, whilst after the union
it keeps the tariff on imports coming from country 3 and eliminates it on
imports from country 2. Let us now consider the effects of the union with
reference to country 1.
As regards commodity A, the most efficient country is country 3, where
the unit cost is lowest. Before the union, country 1 imports commodity A
from country 3, as its price, even with the tariff, is lower than the domestic
cost of production (13 instead of 14). After the union, country 1 imports
the same commodity from country 2, because its cost is 12, lower than 13:
the lower efficiency of country 2, with respect to country 3, in producing A
is more than offset by the tariff schedule. Therefore, the union causes a less
efficient allocation of resources (trade diversion). As regards commodity B,
the most efficient country is country 1: the formation of the customs union,
therefore, does not change the fact that, for this country, it is better to
produce B domestically rather than to import it. The situation for country
1 is the same both before and after the union and the union has no effect on
its trade.
Finally, the presence of a prohibitive tariff prevented country 1 from
importing commodity C; the formation of a union with country 2, which
is the most efficient in producing C, brings about a better allocation of
resources, as country 1 now imports this commodity from country 2 (trade
creation).
This analysis considers oniy the production effects of the union, but it
should be observed that, to evaluate the consequences of the formation of
a customs union, one must also consider the consumption effects and, more
262 Chapter 15. Free Trade vs Protection, and Preferential Trade Cooperation
precisely, the efIects on consurners' surplus. Thus we also have trade creation
and diversion from the point of view of consumption. The former derives
from the fact that consurners substitute cheaper foreign goods (imported
from a member country) for more expensive domestic goods, and so benefit
from an increase in consurners' surplus. The latter derives from the fact
that consurners' surplus decreases as a consequence of consurners having to
substitute more expensive foreign goods (imported from a member country)
for formerly cheaper goods (previously imported from a country remaining
outside the union) which are now non-competitive because the union has
decided to raise tarifI rates with respect to non-member countries.
If we add the efIects on production and consurnption together, we have
trade creation and diversion in the broad sense. The surn of the two efIects
constitutes the overall efIect of the union. Unfortunately not even this more
refined analysis enables us to reach general conclusions. The fact that it is
not possible to demonstrate general propositions (except the purely negative
one that it is impossible to state any precise result, as anything can happen)
is by now obvious if one refers to the theory of second best.
It is however possible to give some indications of a probabilistic type (thus
likely to be sometimes wrang, sometimes correct). These indications are that
a customs union will be more likely to produce beneficial efIects:
(i) the greater is the degree of competitiveness among member countries,
Le. the greater the nurnber of similar goods they produce. In such a case, in
fact, due to the difIerences in productive efficiency, each country will expand
its comparatively more efficient industries and contract the comparatively
less efficient ones; thus there will be more·scope for trade creation without
much trade diversion from other countries;
(ii) the higher are the initial tarifIs between the countries forming the
customs union: in fact, the gain deriving from the elimination of these tarifIs
will be larger;
(iii) the lower are the tarifIs with the outside world: trade diversion, in
fact, will be less likely;
(iv) the wider is the union, as this increases the probability that trade
creation efIects will override trade diversion efIects (in the extreme case, if
the union includes all the world, we have free trade and no trade diversion
can occur).
So far the analysis has been of a (comparative) static type. In addition to
this, the theory of customs unions also examines the dynamic benefits of a
union; the main benefits are:
1) the increase in the size of the market made possible by the union
allows the industries producing traded goods to enjoy the fruits of economies
of scale;
2) the elimination of protection with respect to member countries brings
about an increase in competition;
3) the fact that the member countries tagether negotiate the tariffs with
15.5. Preferential 'Thading Cooperation 263
The European Economic Community (EEC) was founded with the 'Ireaty of
Rome signed in 1957. The founding countries were West Germany, France,
Italy, Holland, Belgium, Luxembourg. At the beginning it contemplated
a common external tarif!; the complete liberalization of trade in industrial
goods among the members took place only in 1968. The still existing non-
tarif! barriers were eliminated in 1986, thus realizing a true customs union.
Free factor mobility among the members was realized subsequently, first that
of capital and then (1993) that of labour, thus giving rise to a true common
market.
Over the years other countries joined the initial six, reaching the num-
ber of 15 (Austria, Belgium, Denmark, Finland, France, Germany, Greece,
Holland, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, United King-
dom); other 10 countries (Czech Republic, Estonia, Cyprus, Latvia, Lithua-
nia, Hungary, Malta, Poland, Slovenia, Slovakia) are going to join in 2004,
and negotiations are under way with other countries (Bulgaria, Romania,
Thrkey) for their admission. In the meantime the name was changed, from
EEC to EU (European Union).
The European Union is something more than a common market, because
it contemplates various measures of coordination of the members' policies
(the best known is the common agricultural policy) and other interventions
to homogenize the economies of the member countries. Details can be found
in the EU's site, http://europa.eu.int
15.6.2 NAFTA
The North Atlantic Free 'Irade Association is a free trade area formed in 1993
among the United States, Mexico, and C~ada, aimed at the elimination not
only of tariffs, but also of non-tarif! barriers to the circulation of commodities
and services. The possibility is also contemplated for each member country
to invest capital in any other member, hence NAFTA is something more
than a mere free trade association. The official site is http://www.nafta-sec-
alena.org
15.6. The Main Cases of Preferential Trading Cooperation 265
15.6.3 MERCOSUR
MERCOSUR (acronym from the Spanish Mercado Comu.n deI Sur, Common
Market of the South; or MERCOSUL, acronym from the portuguese Mercado
Comum do Sul) is an agreement signed in 1991 by some Latin-American
countries, originally Argentina, Brazil, Paraguay, Uruguay, which in 1996
were joined by Bolivia and Chile. This agreement aims at the formation of
a common market among these countries, that should be fully realized by
2005. The official site is http://www.mercosur.org.uy
15.6.4 ASEAN
The Association of Southeast Asian Nations was formed in 1967 for polit-
ical reasons, to defend the member countries against the then communist
1ndochina; the original members were Thailand, Malaysia, Singapore, 1n-
donesia, the Philippines. ASEAN was subsequently transformed into a pref-
erential trading association with the intention of moving on to a customs
union and then to a common market. The founding countries have been
joined over the years by Brunei Darussalam, Cambodia, Laos, Myanmar,
Vietnam. The official site is http://www.aseansec.org
266 Chapter 15. Free Trade VB Protection, and Preferential Trade Cooperation
15.6.5 FTAA
The Free Trade Association of the Americas (or ALCA, from the Spanish
Area de Libre Comercio de las Americas) was proposed in the Miami con-
ference held in 1994 among 34 countries of the Americas. It aims at giving
rise by 2005 to a free trade area that will eliminate all barriers to trade and
investment flows. This area will not be in competition with other existing
agreements (such as NAFTA). The official site is http://www.ftaa-alca.org
267
268 Chapter 16. The New Protectionism
price
fall from qlq4 to Q2q3' The effects on the horne country's price and output
would then be the same as under a quota.
What is completely different is the destination of the sum represented
by the area F1F{HfH1. Under a quota this is a gain accruing to importers
(unless the government auctions off the licenses). But since a VER is by
definition administered by the foreign country, the sum under consideration
accrues to this country. Thus this part of the reduction in domestic con-
sumers' surplus is not offset by aredistribution to domestic importers or
authorities, but is redistributed abroad. The fact that this "rent" from VER
protection accrues to the exporting country is clearly an important motive
for this country to prefer a VER to the imposition of a tariff or quota by the
importing country. This is an economic reason (other non-economic reasons
have been given in Sect. 16.1) for the widespread acceptance of VERs in the
place of the measures of the "old" protectionism.
A trade policy tool that can be used as an alternative to a VER is a vol-
untary import expansion(VIE) . Rather than voluntarily restricting exports
from country 2 to country 1, trade agreements between the two countries can
take the form of country 2 voluntarily increasing imports from country 1. A
VIE sets a minimum market share for imports, hence symmetrically sets a
maximum share for the domestic producer. To reduce its market share to
the level required by the VIE, the domestic firm increases its price, which
induces an increase in the foreign firm's equilibrium price.
According to some authors, VIEs are to be preferred to VERs because,
while the latter are intended to restrict trade, the former are, on the con-
trary, designed to increase trade by increasing foreign sales in countries where
structural impediments and policies restrict access to foreign suppliers. Ac-
tually, the US-Japan trade negotiations seem to have shifted from trying to
limit the access of Japanese firms to the US market to trying to increase the
270 Chapter 16. The New Protectionism
16.3 Subsidies
Subsidies can be present in both the export and the import sector. As regards
the export sector, the subsidy can be either an export subsidy (Le., given to
domestic producers only on the exported part of their output) or a production
subsidy (i.e., given to domestic producers on their whole output). Let us
begin by considering an export subsidy. In Fig. 16.2, D and S represent as
S'
usual the domestic partial equilibrium demand and supply curves. With free
trade, given the ruling international price OM, domestic price is the same,
and exports are FH = (j2q3. Suppose now that an export subsidy, say MN
per unit of output exported, is given to domestic firms. The domestic price
increases from OM to ON: domestic producers, in fact, receive ON per unit
of the commodity exported, and will not be willing to serve the domestic
market unless they receive the same price. If we exclude the possibility of re-
importing the commodity to the domestic market at the world price 0 M, the
domestic price will be driven to ON. Thus domestic producers will seIl N F1
in the domestic market at the price ON and export F1H1 at the prevailing
world price OM, but actually getting ON given the subsidy MN. The total
amount of the subsidy that they receive is thus the area FIF{H~Hl'
Benefits and costs can be calculated using the concepts explained in Sect.
14.2 as regards an import duty. Producers' surplus increases by the area
MNHIH. Consumers' surplus decreases by the area MNHF. The govern-
ment has to pay an amount HFfmHl(note that the area HFfF appears
16.3. Subsidies 271
twice among the costs). Hence the net welfare cost of the export subsidy
is the sum of the two triangles H FI F (the consumption cost) and Hl Hf H
(the production cost). These have the same interpretation as in the case of
a tariff (Sect. 14.3).
The case of a production subsidy can also be examined using Fig. 16.2.
Since this subsidy is given to domestic producers on their whole output, the
result is an equiproportional shift downwards of the supply curve (from 8 to
8') by a percentage equal to the (ad valorern) subsidy. In Fig. 16.2 we have
assumed a production subsidy of the same percentage as the export subsidy.
This is shown by the fact that at output level Oq4 the vertical distance be-
tween 8 and 8' is Hl Hf = MN, denoting that the subsidy to producers is the
same per unit as in the case of the export subsidy. The cost to the government
is now higher: since the subsidy is given on all domestic output, the total
amount is MN Hl Hf. But now there is no decrease in consumers' surplus,
since the price remains at OM. Producers' surplus increases by MNHIH, as
before; hence the net cost is now only the tri angle Hl Hf H (the production
cost). Thus it appears that a production subsidy (which creates no wedge
between the domestic and the international price) is preferable to a direct
trade intervention like an export subsidy (which creates such a wedge).
Let us finally consider a subsidy to the domestic sector producing import-
competing goods (Fig. 16.3). This is actually a production subsidy, hence the
supply curve of domestic producers (8) shifts downwards equiproportionally
by a percentage equal to the (ad valorern) subsidy (8').
The effect of the subsidy is that, for any output, the price received by
producers is greater than the price paid by consumers by the amount of
the subsidy. Let us assurne a world price ON and a subsidy such that the
price received by domestic producers on every unit of output is MN. Hence
domestic producers will be able to supply N F[ = Oq2 instead of N F = Oql.
Consumers continue to pay ON per unit, but producers receive OM. The
outcome of this protective measure is that imports fall from qlq3 to q2q3.
If we now make the usual cost-benefit analysis, we see that there is no
decrease in consumers' surplus, since they continue to pay the same price as
before. The only cost is the government's outlay for the subsidy, namely area
MN F{ F1 • On the side of benefits we have the increase in producers' surplus,
which is MN F F 1 . The balance is a net cost of the subsidy equal to the
tri angle F F{ F 1 (the production cost). It follows that, if we compare a subsidy
to the import-competing sector with a tariff (which entails both a production
and a consumption loss) , we conclude that the subsidy is preferable to the
tariff. Ceteris paribus, a tariff creates two distortions (on both the production
and the consumption side) while a subsidy only creates one distortion (on
the production side).
Let us now put this result together with the previous result, that a pro-
duction subsidy on the side of exports is preferable to an export subsidy
(which can be taken as a negative tariff). The conclusion is that produc-
272 Chapter 16. The New Protectionism
tion subsidies are to be ranked above tariffs. This conclusion lies behind the
suggestion that a subsidy to domestic production is a better policy than a
tariff.
may occur when the interest groups benefiting from protection are diHicult to
organise, and no organisation for other purposes (e.g., for socia! gatherings)
already exists that could be used for setting up the lobby, thus avoiding
the initial cost OG'. This explains why protection is not "demanded" by
everybody and why the interest groups that are already organised tend to
get additional advantages, while newcomers are in a difficult position in the
political market for protection.
Before turning to the supply side, it should be observed that, in addition
to the interest groups gaining from protection, there are groups losing from
protection. Pro-tariff groups mainly consist of firms (including the workers)
producing import-competing goods. On the contrary, anti-tariff groups are
typically the consumers and the exporters.
We now consider the supply of protection. The protectionist measures of
a country are determined by politicians (typically the government) and by
government officials (even if they are not entitled to decide the introduction
of a protectionist measure, bureauerats prepare, formulate and implement
trade bills). A government has certain ideological goals (amongst which
there may be a specific position with regard to free trade or protection), but
also has a number of other goals, amongst which the need or desire of being
re-elected. Since the interest groups demanding protection are usually better
organised than the anti-protection ones and have greater lobbying power
(which includes financial help for the election campaign), a government will
pay more attention to them. FUrthermore, a government also has constraints,
such as the budget and the balance of payments. A high balance-of-payments
deficit may induce protectionist measures, and a budget deficit may be an
additional element for a tariff (which gives a revenue to the government).
Actual tariff rates are the outcome of the interaction between the demand
and the supply in the political market for protection. Various models exist
for the analysis of this interaction.
Brock and Magee (1978), for example, consider the case of two lobbies,
one pro-tariff and the other anti-tariff, and two political parties. The pro-
tariff lobby is better organised and has more money but less votes than the
anti-tariff lobby. Hence there is a trade-off between the number of votes that
the lobby can offer to politicians and the amount of money that the lobby can
give the politicians to finance the electoral campaign. The parties want to
maximise the probability of re-election by choosing an appropriate position
on the free trade-protectionism issue. Each party knows that it can obtain
less votes but more financial resources (which in turn can be used to obtain
more votes through the electoral campaign) by taking a position in favour of
protectionism, and vice versa. The evaluation of the effects of this trade-off
on the probability of re-election is specific to each party.
It is intuitive that each party reaches the optimal position when the
marginal benefit (on the probability of re-election) of more financial re-
sources equals the negative marginal benefit of the votes lost. It can be
16.5. Administered and Contingent Protection, and Fair Trade 275
shown (Brock and Magee, 1978) that the equilibrium solution of the model
(a game-theoretic Cournot-Nash equilibrium) endogenously determines the
tariff level, the amount and distribution of the financial resources employed
in the financing of the parties, and the distribution of votes between the
two parties. In this context, tariffs can be seen as a "price" that clears the
political market for protection.
MC
q 0 Qf q 0 q
a) b) c)
This action is costly, for it entails data collection costs and legal expenses.
Let us call Co this initial (sunk) cost.
b) within a time to the institution issues a preliminary determination,
which may be interlocutory or negative. In the latter case, the procedure
ends, in the former case it continues with the next stage.
c) on the basis of the preliminary findings of the institution, the domestic
industry may decide to withdraw the petition or to pursue it. In the latter
case further ongoing legal expenses are incurred, say Cl, and the institution
continues its investigation, issuing the final decision within a time t 1 .
d) the decision may be positive or negative. In the former case, an an-
tidumping duty is levied (in the United States, the basis is usually the fair
value, see above).
Let us now examine the domestic welfare effects of a successful antidump-
ing petition. The traditional view is that the antidumping duty, as any duty,
increases producers' surplus at the expense of consumers' surplus. This view
has however been challenged on the basis of possible collusive behaviour of
the domestic and foreign industry. Let us start from the observation (Prusa,
1992) that in the United States each of the three possible outcomes of an-
tidumping cases initiated in the period 1980-85 (petition accepted, rejected,
withdrawn) accounted for approximately a third of the total. Now, since
most of the costs of a petition are sunk (Co is much greater than Cl), one
would expect few cases to be withdrawn. However, frequently a petition is
withdrawn only after the domestic industry has achieved some type of out-
of-court settlement with its foreign riYal. The settlement may involve either
a price undertaking or a quantity restriction.
Similar results hold for the European Union, where in the period 1980-
1990 the outcome was (Schucknecht, 1992, pp. 123-125) 24% rejection, 35%
acceptance (hence an antidumping duty), 41% settlement via a price under-
taking (i.e., a voluntary price increase by the foreign firm; the Commission
can negotiate undertakings with the involved parties).
Hence most if not all of the withdrawals are really out-of-court settle-
ments. This is interpreted by Prusa in terms of a game-theoretic bargaining
model which gives rise to a unique Nash solution. The result is that within
this bargaining process the domestic and foreign firms cooperate on pricing
decisions so as to achieve a collusive level of profits.
Thus antidumping cases may actually be used as a stratagem that paves
the way for collusion among (domestic and foreign) oligopolistic firms. In
these cases, as Prusa observes, the welfare conclusion is exactly the opposite
of conventional wisdom: the imposition of an antidumping duty, instead
of decreasing consumers' surplus, might actually increase it, because the
alternative is not free trade, but a collusive oligopolistic situation.
280 Chapter 16. The New Protectionism
283
284 Chapter 17. The new Theories of International Trade
The common feature of these theories is that they drop the assumption of
perfeet competition and/or of product homogeneity.
Two additional features are often stressed as peculiar to the new trade
theories: the explanation of intra-industry trade and the use of increasing
returns to scale.
The first amounts to saying that the new theories can explain intra-
industry trade while the orthodox theory cannot. Intra-industry trade (also
called horizontal trade, two-way trade, cross-hauling) is defined as the simul-
taneous import and export of commodities belonging to the same industry.
17.1. Introduction 285
country 2 country 1
Table 17.1: Orthodox theory and the new theories of international trade
PRODUCTS MARKETS
PerEect Monopolistic Oligopoly
competition competition
Homogeneous Orthodox - Brander
theory (1981)
Vertically Neo Heckscher- - Shaked and
differentiated Ohlin theories Sutton (1984)
(Falvey, 1981)
Horizontally - Demand for va- Eaton and
diJIerentiated riety (Krugman, Kierzkowsky
1979); Demand (1984)
for characteristics
(Lancaster, 1980)
good and the market form. The names of the authors are merely exempli-
ficative, given the host of contributions now existing.
About the market form it is sufficient to remark that in the "oligopoly"
heading we incIude not only duopoly but also, as a limiting case, monopoly.
About the differentiation of the product, it is instead as weIl to cIarify the
terminology.
Vertical differentiation refers to products that differ only in the quality.
For example, woollen suits that are identical except for the quality of the
wool.
Horizontal differentiation refers to products of the same quality that differ
in their (real or presumed) characteristics. For example, woollen suits made
of the same quality of wool but of different cut and colour.
In the case of vertical differentiation, it is incontrovertible that all con-
sumers prefer higher-quality to lower-quality goods. This, of course, presup-
poses the existence of universally accepted criteria for evaluating the quality.
Hence, in the absence of budget constraints, all consumers would demand
the highest-quality good (the assumption is that the price of a commodity
increases as its quality increases). It follows that the demand for different
commodities, Le. commodities of different quality, is related to different
income levels of consumers.
In the case of horizontal differentiation, the various characteristics are
valued differently by different consumers (there are those who prefer a colour
and those who prefer anotherj those who prefer a cut and those who prefer
another, etc.). In any case, consumers generally love variety (even the person
who prefers a certain colour will usually own suits of different colours rather
than all of the same colour). It follows that the demand for different com-
modities, Le. commodities having different characteristics, is related to love
for variety andjor to different subjective evaluations of the characteristics,
288 Chapter 17. The new Theories of International Trade
as we shall show in Sect. 17.5. Actually, most commodities can differ in both
quality and characteristics, but for analytical convenience we keep the two
cases distinct.
Given the greater realism of the assumptions underlying the new theories,
shouldn't we drop the orthodox theory as irrelevant? The answer is given by
Paul Krugman, one of the founders of the new theories. If one were asked to
give an actual example of the new theory of international trade with respect
to the orthodox theory, one could say that "conventional theory views world
trade as taking place entirely in goods like wheatj new trade theory sees it
as being largely in goods like aircraft. Since a good part of world trade is in
goods like wheat, and since even trade in aircraft is subject to some of the
same influences that bear on trade in wheat, traditional theory has by no
means been disposed of completely. Yet the new theory introduces a whole
range of possibilities and concerns" (Krugman, 1990, pp. 1-2).
Before examining the new theories it is as weH to mention the precursors,
namely those authors who, though not giving a rigorous analytical treatment
of the problems, set forth the basic ideas. Ideas that were later taken up,
explicitly or implicitly, by most models classified in Table 17.1. We shall
first examine these pioneering contributions, and then treat the models of
the table.
17.3 Precursors
17.3.1 Availability
According this approach, due to Kravis (1956), international trade is ex-
plained by the fact that each country imports the goods that are not avail-
able at horne. This unavailability may be due to lack of natural resources
(oil, gold, etc.: this is absolute unavailability) or to the fact that the goods
cannot be produced domestically, or could only be produced at prohibitive
costs (for technological or other reasons): this is relative unavailability. On
the other hand, each country exports the goods that are available at horne.
Now, as regards the presence or absence of natural resources, this aspect
could easily be fitted into the Heckscher-Ohlin model that, as we know,
stresses the differences in relative factor endowrnents. We only have to add
a factor natural resources and use the generalized version of the model.
The originality of this approach lies in its second aspect, that is, in the
reasons put forward to explain international differences in relative availabil-
ity. Essentially there are two reasons: technical progress and product
differentiation.
As regards the first reason, Kravis observes that the stimulus to exports
provided by technological change is not confined to the reduction in costs (in
which case we remain in the context ofthe orthodox theory) but also includes
17.3. Precursors 289
the advantages deriving from the possession of completely new products and
of the most recent improvements of existing types of goods. In such cases
the operation of the demonstration effect of Duesenberry creates an almost
instantaneous demand abroad for the products of the innovating country and
thus generates international trade.
As regards product differentiation, the idea of Kravis is to extend to in-
ternational trade the results of the theory of monopolistic competition. Dif-
ferent countries produce similar commodities or, more exactly, commodities
that are not substantially different from the point of view of their intended
purpose (clothes, automobiles, watches, cameras, cigarettes, liqueurs, etc.).
These commodities, however, due to different industrial designs, past ex-
cellence, advertising, real or imaginary secondary characteristics and so on
and so forth, are considered different by consumers. This creates, on the one
hand, a more or less limited monopolistic power of the single producing coun-
tries, and on the other a consumers' demand for foreign commodities that
they believe different from similar domestic commodities, the result being to
create international trade.
any) levied by importing countries, with the unit cost of producing in a for-
eign subsidiary. A possible triggering event that induces the firm to set up a
subsidiary abroad is the appearance in the foreign importing countries of 10-
cal producers of the good. Another important element is the danger that the
governments of importing countries, to protect their industries, may impose
rigid restrietions such as quotas on the imports of the new product.
In the second phase, therefore, it is likely that the innovating firm will set
up producing subsidiaries abroad, in developed countries. Thus, the export
from the innovating country to these count ries will dwindle away to zero,
whilst it will continue to export to developing countries.
Finally, in the third phase of the cycle, we have advanced standardiza-
tion of the good, hence the central, if not exclusive, importance of the cost of
production in determining profit ability. In this phase it may become advan-
tageous to locate production units in less-developed countries because of the
low cost of labour there. It may seem strange that this advantage makes itself
feIt also in the case of capital-intensive goods, but a less-developed country
may offer competitive advantages as a location for the production of these
goods , because the cost of capital may be less important than other factors
(e.g., the marketing of the product, or such a low cost of labour to more than
offset the greater capital intensity).
In the third phase, according to Vernon, in the country where the com-
modity originated, production dwindles whilst demand keeps increasing, so
that this country gradually becomes an importer of the commodity, from
other industrialised count ries to begin with, then from less-developed coun-
tries.
The product cycle model also implicitly offers an explanation of the local-
ization of production in different parts of the world and of the changes in this
localization, hence it can also be considered as aprecursor of the 'economic
geography' models (see Sects. 18.2.1 and 18.2.2).
which gives rise to an exchange of goods within, rather than between, indus-
tries. Empirical studies show an increasing quantitative importance of this
phenomenon.
To begin with, it should be observed that-apart from problems of phys-
ical homogeneity, which will be dealt with presently-internationally traded
goods are usually classified in categories according to the Standard Interna-
tional Trade Classification (SITC). This classification starts from a limited
number of very broad basic classes, distinguished by one digit: for example,
section 1 is "beverages and tobacco", section 8 is "miscellaneous manufac-
tured articles". Within each of these, more detailed categories are distin-
guished by two digits; each two-digit category is in turn disaggregated into
various three-digit categories, and so on up to five digits. It should be noted
that SITC, as internationally adopted, arrives at five digits; for further dis-
aggregation, the individual count ries are free to choose their own description
and coverage. In practice the maximum disaggregation used arrives at seven
digits. It is clear that the higher the number of digits of an item the more
precisely defined the set of similar goods included in that item. In Table 8.2
we give an example of the SITC classification, in which we have considered
only a few disaggregations.
Obviously, if one considers the two-digit items only, the phenomenon of
intra-industry trade is not a surprise, for we are dealing with classes so broad
as to include heterogeneous goods.
Intra-industry trade would then be a spurious phenomenon, due to sta-
tistical aggregation. But since intra-industry trade is also observed in higher-
digit items, even going as far as the seven-digit ones, it cannot be neglected
from the theoretical point of view. Grubel and Lloyd (1975) were among the
first systematically to examine the problem. From the theoretical point of
view we must distinguish between the case of identical goods and the case of
non-identical (though belonging to the same industry) goods.
In the case of identical goods the orthodox theory can supply various
294 Chapter 17. The new Theories of International Trade
p (a)
P2 (a)
PI (a)
o a
note that R;(i = 1,2) is the slope of line Pi, hence the P2 line is steeper
than Pb since R 2 > R 1 . We see from the diagram that prices are equal
in the two countries in correspondence to the "marginal" quality a o , while
country 2 has a comparative cost advantage over country 1 for lower-quality
products (a < a o ); conversely, country 1 has a comparative cost advantage
for higher-quality products (a > a o ).
If we now make the plausible assumption that in both countries there is
a demand for both lower-quality and higher-quality products, it follows that,
in the typical situation of free trade with no transport costs, there will be
international trade in the products of the industry considered: country 1 will
export higher-quality products to (and import lower-quality products from)
country 2. Since we are dealing with products of the same industry, what
has taken place is indeed intra-industrial trade.
What is more, such a trade follows the lines of the Heckscher-Ohlin
theorem, as can be easily shown. Given the assumptions made on the re-
turns to the factors of production, we have Rt/W1 < RdW2 , which means
that country 1 is capital abundant relative to country 2 according to the
price definition of relative factor abundance (see Sect. 13.2.2). Now, since
higher values of a mean both higher qualities and higher values of the capi-
tal/labour ratio, we observe that country 1, the capital-abundant country, ex-
ports capital-intensive products (conversely country 2, the labour-abundant
17.5. Monopolistic Competition and International Trade 297
1 In 1954, Samuelson assumed that only a fraction of exports reaches the country of
destination as imports, just as only a fraction of ice exported reaches its destination as
unmelted ice. The Samuelson ice similitude was subsequently called in the literat ure the
iceberg assumption.
17.6. Oligopoly and International Trade 301
q21 q22
Ru
E
B q22
I
E I I E,
q21 ---T-r---
I I I
I I
I I
I I
assumed identical, the price also will be identical in the two markets. It
follows that, due to transport costs, for each firm the f.o.b. price of exports
is lower than the domestic price of the same commodity, and therefore there
is a kind of reciprocal dumping.
When the two autarkie eeonomies are different (the diversity being mea-
sured by a different ineome distribution), a greater number of firms ean
eoexist in the eombined world eeonomy when trade is opened UPi but this
number becomes smaller as the ineome distributions get nearer.
Let us now eome to the long run , always starting from two initially au-
tarkie eeonomies. The Shaked and Sutton model shows that the number of
firms that ean survive in eaeh eountry is only two, and that other firms that
tried to enter the market would suffer losses (henee they do not enter). What
happens when international trade is opened? We must as before distinguish
two eases, that in which the two eeonomies are identical, including ineome
distribution, and that in whieh they are different as regards ineome distri-
bution. In the former ease the same result as in the two autarkie eeonomies
will eontinue to hold for the eombined world eeonomy, namely no more than
two firms producing two different qualities will survive. The model eannot
however forecast which are these firms, so that it might happen that the two
surviving firms belong to the same eountry. In this ease there would be one-
way trade, for the other eountry would have to import both eommoditiesi
of eourse there will have to exist other seetors in which sueh eountry ean
export, beeause in the eontext of the pure theory no eountry ean be only an
importer. When, on the eontrary, the two surviving firms belong to different
eountries, sinee the eonsumers in both eountries demand both eommodities,
there will be intra-industry trade with the simultaneous import and export
of different qualities of the eommodity. Finally observe that, sinee eaeh firm
will serve not only the domestic but also the foreign market, the eeonomies
of seale will allow aprice reduetion, henee an inerease in eonsumers' welfare
(the gains from trade).
If ineome distribution is different in the two autarkie eountries, the num-
ber of firms that ean eoexist in the world economy is greater; but for our
purposes it is suffieient to observe that the result will be in any ease the
ereation of intra-industry trade to satisfy eonsumers' demands in both eoun-
tries.
than before, are still greater than those that it would obtain giving up any
export to country 2 and only producing for its domestic market like in the
pre-trade situation.
BOX 17.2 Strategie trade poliey: Boeing vs Airbus
The aireraft seetor provides a textbook example of governmental startegie trade
poliey, namely an industry in which trade poliey eould affeet the strategie interac-
tion between a domestie and an international rival: by subsidizing produetion, the
government ean affect the outeome of the eompetitive game in such a way as to shift
rents in favour of the domestie firms as argued by Brander and Speneer (1985). In
the eommercial aerospace industry the produetion has been direetly and indirectly
supported by using the market failure argument. The aerospace industry is surely
subject to market failure, notably beeause of large seale eeonomies in produetion
and the importanee of research and development. Given this industry's market
structure it is diflicult for individual eountries to face international eompetition, so
aireraft industry has given rise to signifieant international cooperation. One of the
most famous ease of such cooperation is the European Airbus Consortium which
was formed in the late 19608 to challenge the dominance of the Boeing Corporation
in international world markets. The publie support to Airbus has mostly taken the
form of a reduetion in fixed development eost.
Although not reeent, the ease of Boeing VB. Airbus eontains useful background
information on the subsidy issue. Boeing has long been the leader in the world
aviation industry and when Airbus was created the eommercial aireraft was almost
eontrolled by US firms. Airbus slowly but steadily expanded its market share during
the first two decades of its existenee and with other eompetitors out of the pieture
(Lockheed and MeDonnel Douglas) the battle for market share in the 19908 and
beyond is being waged direetly at Boeing's expense.
Boeing and MeDonnell Douglas accused Airbus to be state-supported with virtu-
ally unlimited (henee unfair) financial resourees in the form of cheap loans, the
repayment of whieh was eontingent on Airbus's profits. On the other side the Eu-
ropeans argued that Ameriean aireraft manufactures received indireet government
subsidies-from the Department of Defense and NASA-of eomparable magnitude.
The battle over the appropriateness of subsidies raged for the first twenty-two years
of Airbus' presenee. The US government lodged a eomplaint against Airbus under
the GATT and in 1992, an "Airbus Agreement" was signed between the United
States and the European Community. This agreement eontained three main points:
- Direet government subsidies for aireraft were eapped at 33% of developments eosts.
Loans made to the eonsortium were to be repaid aceording to striet scheduling and
interest-rate requirements.
- Indirect subsidies were limited to 3% of the turnover of civil aircraft manufacturers;
- A bilateral panel would monitor compliance of the previous two points, and inerease
the ''transpareney'' of the eommercial aireraft industry.
Strategie trade poliey emphasizes its results in the presenee of oligopoly: any exter-
nal intervention alters the strategie interaction between players on the market. If a
domestie firm is apart of an international oligopoly and reeeives any kind of sup-
port from its government, it eompetes suecessfully. There seems to be little doubt
that the Airbus projeet would not be in a position of sueh prominenee without
government support.
In such a situation the government of country 2, being aware of the strategie
interaction, may even calculate (and impose) a tariff that takes away from
the foreign firm all profits in excess of profits this firm earns by selling only
to consumers in country 1. In such a case this firm is indifferent between
17.8. Suggested Further Reading 307
309
310 Chapter 18. Growth, 'Irade, Globalization
traded along the lines of the Heckscher-Ohlin model; since the economy under
consideration is assumed to be a small open economy, the terms of trade
or relative price PB/PA is exogenously given by the international market.
Production takes place using capital and labour under constant returns to
scale and neutral technical progress.
The third sector produces a nontradable good, say Z, and is the cru-
cial one. It is the R&D sector, which can be considered as the sector that
"produces" technical progress by using primary factors. More precisely, this
sector employs capital and labour to provide R&D services to the traded
goods sectors to increase their efficiency. Technical progress is purely 'local',
as it only accrues to domestic firms, with no international spillovers.
The production of R&D services takes place under constant returns to
scale, but the increase in efficiency that accrues to the tradables when there is
an increase in these services is subject to diminishing marginal productivity.
Since the same primary factors are used to produce the three goods, and
production functions are homogeneous of the first degree, it follows that at
some initial time to (in which we can take the index of technological efficiency
as equal to one) the relative price of the nontradable, pz, is also determined
by the international market for traded goods.
We are now facing a problem of optimal allocation of resources, whose
solution will yield the rate of endogenous technical progress. In fact, there is
a trade-off between current and future outputs of tradable goods (on which
social welfare ultimately depends). Since the amount of primary factors is
given and constant through time, if more factors are allocated to the R&D
sector , there will be less current output of tradables but more output of
them in the future due to the higher rate of technical progress. If less factors
are allocated to R&D, there will be more current output but less future
output of tradables. Let v be the value of the output of tradables and v its
instantaneous change, a function of z, the per-capita output of R&D services.
As in all optimization problems involving trade-offs, we can apply the
usual optimization rule that equates the marginal benefit to the marginal
cost of an incremental expenditure on R&D. The instantaneous marginal
benefit is dzi/dz, namely the increment in zi due to an increment in z. Since
this benefit accrues from now to infinity, we must calculate its present value.
In general, the present value of an infinite stream having a constant value in
each unit of time is obtained dividing such a constant value by the discount
rate. Thus the present value we are looking for is (dzi / dz) /6, where 6 is a
discount rate (the interest rate or the social discount rate). This represents
the marginal benfit to be compared with the marginal cost, which is simply
pz, the relative price of the non-traded good. Marginal benefit and cost are
equated when (dzi/dz)/6 = pz, namely dzi/dz = 6pz.
The result can be shown in Fig. 18.1, taken from Findlay (1995, p.
89). The curve OF shows zi(z), the increase in value of tradable output as
a function of z. This curve embodies the trade-off between the reduction
312 Chapter 18. Growth, Trade, Globalization
v H
The model is rather complex, and its results can be found only by a formal
analysis. They can be summarized as folIows.
In steady-state growth, two main types of equilibria may occur. In the
first type, followers are relatively efficient at innovation (though less so than
leaders), hence both leaders and followers engage in innovation. This equilib-
rium gives rise to a complex history of product cycles because at any moment
the market leadership can pass from one Northern firm to another (formerly
a folIower) or from North to South (when imitation in South is successful
and R&D in North fails to develop a higher-quality product). Product cycles
go back and forth.
The second type of equilibrium (the inefficient folIower case) occurs when
followers have a relatively large inferiority in the research lab with respect
to leaders, so that only these latter carry out R&D in the steady state. In
this case the outcome is more clear-cut, as there will be alternating phases
of production between North and South, with Northern firms developing
new products and being market leaders until Southern firms displace them
thanks to successful imitation, after which there will be another innovation
by a Northern firm and so on.
Grossman and Helpman (1991a, Chap. 12) also examine the consequences
on world growth and trade of subsidies to R&D by either the Northern or
Southern government. The results depend in an essential way on the type of
equilibrium that obtains.
In the inefficient follower case, technological progress and hence growth
is favourably affected by the introduction of a research subsidy by either
government. Not only a research subsidy by the Northern government to its
innovative firms fosters technological progress, but also a higher pace of imi-
tation (brought ab out by a subsidy by the Southern government to its firms)
has the same effect, causing Northern firms to increase their research efforts
to regain market leadership after losing it to Southern imitators. This result
is the same that can be obtained in a similar model of North-South trade
with imitation in which, however, product differentiation is of the horizontal
type, so that innovation consists of the development of new varieties of the
product (Grossman and Helpman, 1991a, Chap. 11).
Results of government intervention may however be strikingly different
in the efficient follower case: "In this case an expansion in the size of South
may slow down the rate of innovation in North, and policies that might be
used to promote domestic productivity gains spill over abroad with adverse
consequences for the foreign rates of technologieal progress" (Grossman and
Helpman, 1991a, p. 327). As in the ease of strategie policies in a statie
context, results of government intervention in a dynamic context are heavily
model dependent, which eomes as no surprise.
18.2. Globalization and the new Economic Geography 315
broader field of economic geography, a field that would also indude interna-
tional trade theory as a special case. It would then seem quite natural to
observe a dose integration between international trade theory and location
theory in the broader context of economic geography, but this has not been
the case, for several reasons examined for example by Krugman (1991, 1993).
The present section briefly examines the relations between location of
production, cost of transport, and international trade in the context of both
the orthodox and the new theories of international trade.
While both N and L are in fixed supply, the supply of capital is endogenous.
Commodity A is assumed to be capital-intensive with respect to B, and turns
out to be also resource-intensive. To show this, we observe that commodity
A-apart from the amount of N directly used to produce Z which is specific
to A-indirectly requires more N with respect to B. In fact, since K embodies
the part of A that has been invested (hence K indirectly embodies N), it
follows that A, being capital-intensive with respect to B, indirectly requires
more N.
Consider now two identical count ries except for the endowment of natural
resources, which is larger in country 1. Since Nd LI is greater than N 2 / L 2 , we
speak of 1 and 2 as the resource-rich and resource-poor country, respectively.
The first step is the introduction of international trade in final goods
only. The raw material Z is assumed to be non traded due to prohibitive
transport costs when it is in unprocessed formj these costs disappear when it
is embodied in the capital-intensive final good A. With these assumptions the
model behaves like the standard 2 x 2 Heckscher-Ohlin model, hence country
1 will export the resource-intensive commodity A and import commodity
B, while country 2 will export commodity B and import commodity A. In
country 1 the A sector will expand and the B sector will contract, while the
opposite will take place in country 2.
Thus, when there is free trade in final goods only, what happens is a
higher extraction of the raw material input in the resource-rich countrYj this
entails an increase in the capital stock to meet the needs of the higher output
of the capital-intensive exportable commodity A. In the other country the
opposite will happen: the resource sector shrinks because of the reduction
in the output of the import-competing commodity A, with a corresponding
decrease in the long-run capital stock. As Findlay (1995, p. 168) notes,
"free trade clearly enhances the initial difference in wealth between the two
count ries based on the difference in natural resource endowment."
The second step is to allow free trade in all commodities, because a trans-
port revolution takes place so that the resource input Z Can be traded at
zero transport cost like the two final goods. We now have a model with three
traded goods (one of which is a factor of production) and three factors (one
of which is traded). Given the assumption of internationally identical tech-
nologies with constant returns to scale, if we further assume that all three
commodities are produced in both countries, factor prices will be equalized.
In the long-run equilibrium, agents must have the same per capita utility
level in both countriesj this implies that per capita income and per capita
wealth (and hence total wealth, given the assumption of identical labour
force) must also be equal. Total wealth is made up of two components, the
capital stock and the capitalized value of the rents from the natural resources
N used to produce Z.
Commodity- and factor-price equalization implies that the price of Z in-
creases (with respect to the pre-trade situation) in the resource-abundant
318 Chapter 18. Growth, Trade, Globalization
country 1 (where before trade it was lower than in the resource-scarce coun-
try) and decreases in the resource-scarce country 2. The resource sector
shrinks in country 2 and expands in country 1, which implies that, in the
long run equilibrium, the natural-resource component of wealth is greater in
country 1 than in country 2. This in turn entails a greater long-run equilib-
rium capital stock in country 2 than in country 1, as total wealth must be
equal in both countries.
The final result is that in the long-run equilibrium country 2 may become
the exporter of the capital-intensive commodity A.
"In other words, the possibility of sharing on equal terms in a global
pool for access to the intermediate input enables the resource-poor country
to build up its capital stock per head to such an extent that it leads to
areversal of its former comparative advantage in the labor-intensive good.
[... ] It is now the less naturally well endowed countries that will have a
higher proportion of physical capital per capita in their port folios and will
thus export the capital-intensive industrial goods on the basis of imported
intermediate inputs as, for example, in the case of Japan" (Findlay, 1995, p.
170 and p. 172).
the world into rich and poor nations, further integration of world markets
will bring about a eonvergenee in ineomes and eeonomie strueture.
Further developments are eontained in Fujita, Krugman, and Venables
(1999).
Krugman, P., 1993, On the Relationship between Trade Theory and Location
Theory, Review oE International Economics 1. 110-22.
Krugman, P. and A.J. Venables, 1995, Globalization and the Inequality of
Nations, Quarterly Journal oE Economies 110, 857-80.
Midelfart-Knarvik, K.H., H.G. Overman, S.J. Redding and A. Venables,
2000, The Location oE European Industry, European Commission, Brus-
sels.
Ohlin, B., 1933, Interregional and International 'Irade, Harvard University
Press.
Ohlin, B., P. Hesselborn and P.M. Wijkman (eds.), 1977, The International
Allocation oE Economic Activity, London: Macmillan.
Stiglitz, J.E., 2002, Globalization and Its Discontents, New York and London:
W.W. Norton.
Index
323
324 Index
X
Xeno-currencies, 29
Xeno-markets, 30
Z
Zamalloa, L., 174
List of Figures
337
338 List of Figures
17.1 Orthodox theory and the new theories of international trade . 287
17.2 Example of SITC Classification . . . 293
339
List of Boxes
341
nter:na ·onal.~_
Economics
i
PlHse order from
Springer . Customer Service · Haberstr. 7 '69126 Heidelberg, Germany
Tel.: +49 (0) 6221 - 345 - O· Fax: +49(0) 6221 - 345 - 4229
e-mi!: ordel'[email protected] · or through JOUr booLts.ller
All Euro and GSP prices ar~ ne:t· prKes subj«' 10 locaJ VAT, e.g, in Ge:rmany "", VAT for booh.
Springer
Pneu and olh~r d~tails an 5ubjt:CI 10 chans~ withOut nOtlU_d&p ' BA 21133/1