Content uploaded by Gilbert Faccarello
Author content
All content in this area was uploaded by Gilbert Faccarello on Jun 01, 2021
Content may be subject to copyright.
‘I profess to have made no discovery’.
James Mill on comparative advantage
Gilbert Faccarello ∗
The study of Mr Ricardo’s work . . . has been much facilitated by the
labours of late writers . . . Mr Mill’s Elements of Political Economy is
a work of a higher order. (John Ramsay McCulloch, A Discourse on the
Rise, Progress, Peculiar Objects and Importance of Political Economy,
1824)
Without losing sight of Ricardo’s main book, we will focus on Mr Mill’s
Elements, one of the most esteemed and estimable authors, who pro-
fessed the same principles. (Guillaume Prévost, ‘Réflexions du traduc-
teur sur le système de Ricardo’, 1825)
His Elements of Political Economy . . . presented the views of Ricardo
in a precise and clear style. (The New American Cyclopaedia, XI, 1861)
1Setting the stage
After a first, brief reappraisal in the 1930s (Sraffa 1930), David Ricardo’s ideas
on international trade, stated in the chapter ‘On Foreign Trade’1of his 1817-
21 Principles of Political Economy and Taxation, have been submitted to new
scrutiny (see for example Yukizawa 1974 and 1978, Ruffin 2002, Maneschi
2004, Aldrich 2004, Faccarello 2015a and 2015b, Gehrke 2015 and 2017, de
∗Panthéon-Assas University, Paris. Email: gilbert.faccarello@u-paris2.fr. This paper is
forthcoming in The European Journal of the History of Economic Thought,29 (1), February
2022.
1Contrary to the chapter on value, Ricardo’s chapter on foreign trade remained un-
changed through the three editions of the Principles: only its numbering was modified —
from Chapter 6 in the first edition, it became Chapter 7 from the second edition onwards.
1
James Mill on comparative advantage 2
Boyer 2017, Tabuchi 2017, Hisamatsu 2019, Takenaga 2019). This innovative
research has led to a new understanding of the so-called ‘four magic numbers’
or ‘principle of comparative advantage’. It shows that the alleged ‘Ricardian’
model of foreign trade is not to be found in Ricardo. Based on a misunder-
standing of Ricardo’s text, it was actually proposed by some of his followers.
Ricardo’s statements
In a nutshell,2Ricardo supposes an equilibrium exchange between two coun-
tries, England and Portugal, where two commodities, cloth and wine, can be
produced and exchanged. The quantities exchanged are given, and so is the
international exchange ratio: England exports bunits of cloth to Portugal and
imports aunits of wine, while Portugal exports aunits of wine to England
and imports bunits of cloth. It is also stated that, if each country had to
produce these quantities of commodities, Portugal would employ 80 and 90
units of labour to produce aunits of wine and bunits of cloth respectively,
while England would need 120 and 100 units of labour respectively. These
numbers — the ‘four magic numbers’ — are also the respective labour values
of the quantities produced in each country.
In spite of the fact that Portugal has an absolute advantage — smaller
real costs — in the production of the quantities of both commodities, and
England none, this exchange is possible and is in fact supposed to happen.
This is because each country has a greater relative facility of production in
one commodity: wine is relatively less expensive to produce than cloth in
Portugal, and cloth relatively less expensive to produce than wine in England.
Moreover, such an exchange is profitable to both countries. For each of
them, the gains from trade are immediately given and consist in the difference
between the cost of production the country would have spent on the home
production of the quantity of the foreign commodity it imports and the cost
of the quantity of the home commodity it exports in exchange. Portugal’s
gains from trade are thus of 10 units of labour and England’s gains are 20.
Both countries can employ the units of labour they save in the production of
more wine or cloth or any other commodity. While the gains from trade are
2For a recent summary, see Faccarello 2015a.
James Mill on comparative advantage 3
not equal on both sides, both countries nevertheless enjoy a greater amount of
use values. Some other advantages can also arise, for example when a country
imports cheaper agricultural products, which could lower wages and thus raise
the profit rate.
Another meaning for the four numbers?
The Ricardian model supposes instead that the ‘four magic numbers’ are tech-
nical coefficients, that is, quantities of labour necessary to produce one unit
of cloth and wine respectively in the two countries. While the difference with
Ricardo’s approach seems trifling, it actually has far-reaching theoretical con-
sequences. In particular, the amounts of commodities exchanged between the
two countries and their international relative price are now unknown, as are
each country’s gains from trade. Consumers’ preferences and demand have to
be introduced into the analysis to determine them.
In striking contrast with Ricardo’s theory of value and prices, the ‘Ricar-
dian’ model came to be accepted and developed by almost all economists and
still lies at the heart of international trade theory.
Two questions, however, are pending: How could Ricardo’s celebrated chap-
ter on international trade have been misunderstood by generations of readers?
How and when was Ricardo’s ex-post approach, starting with an existing in-
ternational exchange, replaced by an ex-ante approach focusing on why it was
in the interest of countries to open to trade?
In this change of perspective, the role of John Stuart Mill is usually em-
phasised and two of his texts are referred to: the 1829–30 essay ‘Of the laws
of interchange between nations, and the distribution of the gains of commerce
among the countries of the commercial world’, published in 1844 in Essays
on Some Unsettled Questions of Political Economy, and his 1848 Principles of
Political Economy. The parts played by James Mill, James Pennington and
Robert Torrens are also usually mentioned.
The present paper focuses on the role of James Mill in the interpreta-
tion of the ‘four numbers’ and the creation of the Ricardian vulgate, even if
the idea of a Millian innovation in this respect is a kind of paradox because
Mill’s intention, in this field as in others, was just to state Ricardo’s principles
James Mill on comparative advantage 4
pedagogically in an introductory book. This paper endeavours to depict this
process as precisely as possible, focusing on Mill’s 1818 article ‘Colony’3in the
third volume, Part I, of the Supplement to the Fourth and Fifth Editions of the
Encyclopaedia Britannica,4and the three editions of his Elements of Political
Economy (1821, 1824, 1826). It shows how the evolution of Mill’s thought, in
part due to a didactic perspective and some difficulties raised by his numerical
examples, gradually distorted Ricardo’s approach. In Mill we find in the end
all the ingredients of what is called the ‘Ricardian’ model of foreign trade,
substituting an ex-ante perspective for Ricardo’s ex-post approach.
In the following, the chronological order of the texts is followed. I disregard
the question of a possible intervention by John Stuart Mill in the various
editions of the Elements (see for example Thweatt 1986, Aldrich 2004): I only
deal with the evidence of the texts. I also disregard the vexed question of
priority in the formulation of the ‘principle’ of comparative advantage (see for
example Thweatt 1976, Maneschi 1998, Ruffin 2005): as soon as the peculiarity
of Ricardo’s formulation is admitted, the question changes and becomes that
of how and when Ricardo was misinterpreted.
2‘Colony’, 1818
In his 1818 entry in the Supplement of the Britannica, James Mill‘s analysis
— confined to only two paragraphs towards the end of the text (1818, 26–7)
— unambiguously follows Ricardo’s approach to the question of international
3The title of James Mill’s 1818 entry is ‘Colony’, though sometimes referred to as
‘Colonies’. This is probably due to the fact that, in the table of contents of some col-
lected essays Mill had privately printed in the 1820s, the text is mentioned as ‘Colonies’ —
but ‘Colony’ is maintained as the title of the piece on the essay itself.
4The Supplement to the Fourth and Fifth Editions of the Encyclopaedia Britannica was
published progressively between 1815 and 1824 — the fifth edition (dated 1815) being almost
identical to the fourth (1801–9). A sixth edition was published in 1815–23, and the title
of the Supplement was changed accordingly for the (re)publication of the six Supplement
volumes in 1824. For the entry ‘Colony’, I could not find the original 1818 text: I therefore
refer to the official republication of it in Volume 3 of the Supplement, 1824 (pp. 257–73), and
more specifically to one of the reprints of the entry, privately made for Mill — the reprints
all have the same pagination (33 pp.) even when they are included in a collection of essays.
The sole difference between the text in the Supplement and the reprints is that the latter
are divided into sections while the original is not. The reprints are not dated.
James Mill on comparative advantage 5
trade, its direction and the benefits it generates (see also Thweatt 1986, 37-
38). Ricardo’s book was published in 1817: Mill was thus quick to adopt this
approach — but Mill’s active role in the publication of the Principles is to be
remembered (Sraffa 1951).
A nation exports to another country, not because it can make
cheaper than another country; for it may continue to export, though
it can make nothing cheaper. It exports, because it can, by that
means, get something cheaper from another country, than it can
make it at home. But how can it . . . get it cheaper than it can
make it at home? By exchanging for it something which costs it
less labour than making it at home would cost it . . . So long as
what it does give is produced by less labour, than the commodity
which it gets for it could be produced by at home, it is the interest
of the country to export. (1818, 26)
An example follows of an exchange between England and Poland: a given
quantity of Polish corn (say, aunits) is exchanged for a given quantity of En-
glish cotton (say, bunits), though Poland can produce both quantities cheaper
than England can do. The quantities of labour necessary to produce these
quantities — men per unit of (unspecified) period — are reported in Table 1.
Table 1: The ‘four magic numbers’ in James Mill’s entry ‘Colony’.
aunits of corn bunits of cotton
Poland 80 85
England 100 90
Both England and Poland benefit from trade. England purchases in Poland
‘the same quantity of corn which is produced in England by the labour of 100
men . . . with a quantity of cotton goods which she has produced with the
labour of 90 men’. And ‘Poland can get with 80 men’s labour, through the
medium of her corn, the same quantity of cotton goods which would cost her
the labour of 85 men, if she was to make them at home’ (1818, 26). Hence the
gains from trade: ‘Both nations . . . profit by this transaction; England, to
the extent of 10 men’s labour, Poland to the extent of 5 men’s labour’ (1818,
27).
James Mill on comparative advantage 6
Finally, it should be noted that James Mill seems to pay due attention to the
monetary context in which Ricardo developed the argument in his Principles.
This point is developed below.
3Elements of Political Economy, 1821
Three years after the publication of ‘Colony’, the themes of comparative advan-
tage and related benefits from trade were developed by Mill in his Elements
of Political Economy — Chapter 3: ‘Interchange’, Section IV: ‘Occasion on
which it is the interest of nations to exchange commodities with one another’.
His discourse changed, but progressively.5
The nature of this book is worth noting, because it played an important
role in the diffusion of Ricardo’s ideas and their various transformations. Mill
convinced Ricardo to write his Principles, but Ricardo’s treatise was certainly
not easy to read and a more accessible work was needed to spread the new doc-
trine. In his Elements, Mill’s intention was therefore purely didactic. ‘During
my confinement at home’, he wrote to Ricardo on 28 December 1820, ‘I have
been making good progress with my School Book of Political Economy. In fact
I have got over all the knotty points; and, as I think, clearly; so that any body
will understand them’. He added: ‘I wish it may appear to you calculated to
teach the science, easily and effectually. In that case I shall conclude that I
have done a good service; as diffusing of knowledge is now the work of greatest
importance’ (Mill to Ricardo, 28 December 1820, in Ricardo 1951–73, VIII,
327).
Mill’s aim is also clearly stated in the Preface to the book.6But while
the work was successful and praised by economists, who however sometimes
stressed that it was not as elementary as it professed to be — ‘and is, perhaps,
5The idea of a sudden break between Mill’s 1818 entry on ‘Colony’ and the first edition,
three years later, of the Elements, is to be discarded. Mill had his entry ‘Colony’ privately
reprinted a couple of times in the 1820s, without changing the content, and did not change it
either when his text was reprinted in the third volume of the 1824 Supplement to the fourth,
fifth and sixth editions of the Encyclopaedia Britannica.
6‘My object has been to compose a school-book of Political Economy; to detach the
essential principles of the science from all extraneous topics, to state the propositions clearly
and in their logical order, and to subjoin its demonstration to each’ (Mill 1821, iii).
James Mill on comparative advantage 7
better calculated for the use of those who are considerably advanced in the
science than of beginners’ (McCulloch, 1824, 71) — Ricardo himself was not
so satisfied, as he wrote to Mill on 10 December 1821. He read the book ‘with
attention’ and expressed some concerns: ‘you must be aware that there are few
things in it from which I can differ. There are, I perceive, a very few in which we
do not quite agree; I have taken notes of them’ (in Ricardo 1951–73, IX, 117).
On 18 December, he sent a long list of remarks to his correspondent (1951–
73, IX, 126–33), in which, however, nothing directly concerns comparative
advantage. Does this necessarily mean that Mill’s developments in this field
strictly followed Ricardo’s?
Trade between England and Poland
As already mentioned, Mill’s intention was to be didactic. This is probably
the reason why he tried to show exactly why and how an exchange between
two countries is likely to take place. He thus first specified a case in which such
an exchange cannot happen. Suppose two countries, he writes, England and
Poland, both able to produce a given quantity of cloth and a given quantity
of corn — say aunits of corn and bunits of cloth. And suppose first that the
number of days of labour to produce them are those given in Table 2: Poland
can produce these quantities with 100 days of labour each, and England with
150.
Table 2: The ‘four magic numbers’ in James Mill’s 1821 Elements of Political
Economy, Chapter 3, section IV, example 1.
aunits of corn bunits of cloth
Poland 100 100
England 150 150
Obviously, in this case, ‘no exchange would take place’. ‘With 150 days’
labour in cloth . . . England would only get as much corn in Poland as she
could raise with 150 days’ labour at home; and she would, on importing it,
have the cost of carriage besides’ (1821, 85–6). Poland is in a similar position:
with aunits of corn which cost 100 days of labour, she obtains bunits of cloth,
James Mill on comparative advantage 8
which could, however, have been produced at home at the same cost, and she
would have in addition to pay for the transport.
To explain the possibility of an exchange, then, something must be changed.
Mill supposes that the cost of production of the quantity of corn in England
is modified: it is now 200 days of labour instead of 150, as shown in Table 3.
Table 3: The ‘four magic numbers’ in James Mill’s 1821 Elements of Political
Economy, Chapter 3, section IV, example 2.
aunits of corn bunits of cloth
Poland 100 100
England 200 150
With these new numbers, England can now benefit from trade:
With a quantity of cloth which England produced with 150 days’
labour, she would be able to purchase as much corn in Poland as
was there produced with 100 days’ labour; but the quantity, which
was there produced with 100 days’ labour, would be as great as
the quantity produced in England with 200 days’ labour. Eng-
land, therefore, would obtain her corn with less labour, through
the medium of her cloth. (1821, 86)
Up to this point, it seems that Mill’s analysis conforms to Ricardo’s. How-
ever, there is already here an incipient change of perspective, which forms a first
departure from Ricardo’s views: while, in the Principles, an ‘actual’ exchange
is analysed, Mill is explaining the genesis of this exchange — an approach
already nascent in his Commerce Defended (Mill 1808, 35–39). In addition, a
problem arises with the chosen ‘four numbers’: the analysis cannot be applied
to Poland in the same way if Ricardo’s line of thought is to be maintained.
Poland has no interest in exchanging aunits of corn for bunits of English
cloth because, as before, she can produce these bunits of cloth at home with
the same quantity of labour that is necessary to produce the aunits of corn
given in exchange.
James Mill on comparative advantage 9
Yet Mill stresses that ‘Poland would profit in the same manner’ (1821, 86).
He is, however, uncomfortable and seems aware of the problem raised by the
new ‘magic numbers’. Three rather obscure sentences follow:
A quantity of corn which cost her 100 days’ labour, being equal
to the quantity produced in England by 200 days’ labour, would
purchase, in England, the produce of 200 days’ labour in any other
commodity; for example, in cloth. But the produce of 150 days’
labour in England in the article of cloth, is equal to the produce
of 100 days’ labour in Poland. If, with the produce of 100 days’
labour, she [Poland] can purchase, not the produce of 150, but the
produce of 200, she gains to the amount of 50 days’ labour; in other
words, a third. (1821, 86–7)
This passage has been interpreted in various ways. In my view, Mill is
sticking to the same data (see also Thweatt 1986, 35). However, he is changing
the way in which the quantities of commodities are ‘exchanged’. While England
is interested in trade on Ricardo’s basis, the fact that Poland cannot act on
the same basis induces him to suppose that Poland sends aunits of corn to
England and sells them there for 200. In the English market, with these 200,
Poland can buy bunits of cloth for 150 (the same quantity that she can produce
at home at the same cost as that of the corn she gives in exchange), plus ‘any
other commodity’ for 50 (her gain from trade), that is, a third of 150. If these
50, for example, are also spent on cloth, Poland’s gain would be a third of
bunits of cloth — ‘in other words, a third’ of the quantity of cloth she can
produce at home with 100 days of labour.
We thus see here a second notable departure from Ricardo’s approach:
Poland is trading in the English market at English domestic prices. But this
does not settle the problem because, with this transaction, England’s benefits
from trade vanish: in exchange for aunits of corn, she gives now bunits of
cloth plus 50 of ‘any other commodity’. Mill was to realise this five years later.
The above-quoted apparently obscure passage is left unchanged in the sec-
ond edition of the Elements but is significantly modified in the third. As will
be seen, the changes confirm the interpretation just suggested. For the time
being, it is interesting to note that the results of the analysis do not change
for England if we apply to this country the new exchange rule used for Poland.
Suppose England sends bunits of cloth to Poland and sells them there at the
James Mill on comparative advantage 10
local value of 100: in the Polish market, England will be able to buy, with
these 100, aunits of corn, and her gain from trade is still 50. Mill could thus
have had the impression of perfect coherence in his reasoning.
This case of an exchange between England and Poland is to be found again
later in the book, in sections XIII and XIV of the same Chapter 3, in the
context of a monetary analysis. Now (Table 4) Mill supposes that ‘four quarters
of corn and 20 yards of cloth required the same quantity of labour in England’
and that ‘in Poland 20 yards of cloth required twice as much labour as four
quarters of corn’. As a consequence, ‘four quarters of corn, which in England
would be of equal value with 20 yards of cloth, would in Poland be equal to
no more than 10 yards’ (1821, 135).
Table 4: The ‘four magic numbers’ in James Mill’s 1821 Elements of Political
Economy, Chapter 3, Section XIV
corn cloth
Poland, for a given quantity of labour 4 quarters 10 yards
England, for (another) given quantity of labour 4 quarters 20 yards
The meaning of the ‘four magic numbers’ has now changed: (i) the quanti-
ties of corn and cloth produced in each country with a given quantity of labour
(not necessarily the same in both countries) are known: 4 quarters of corn and
20 yards of cloth in England, and 4 quarters of corn and only 10 yards of cloth
in Poland; (ii) the quantities exchanged are therefore unknown: the four num-
bers are only technical data, and this forms a third departure from Ricardo’s
model.
In this context, the new way of describing the exchanges between the two
countries is generalised: both England’s and Poland’s exports are said to be
exchanged in foreign markets at the foreign domestic prices — with, for each
country, an erroneous calculation of the gains from trade.
If Poland . . . should send corn to England; the quantity of labour
which produced four quarters, or 10 yards of cloth, in Poland, would
purchase 20 yards in England. In like manner, if England should
send cloth to Poland; the same quantity of labour which produced
James Mill on comparative advantage 11
20 yards of cloth, would obtain, not four quarters, only, of corn, all
it could produce at home, but . . . eight quarters. (1821, 135–6)
With regard to comparative advantage and the gains from trade in the
England / Poland example, the first edition of the Elements does not stop
here: a short complement is added (1821, 87–8), in which Mill starts to deal
with units of commodities, referring to the production of ‘a quarter of corn’.
Mill’s presentation of Ricardo’s approach
At the end of Chapter 3, Section IV, after presenting the England / Poland
example, Mill sums up, in a few paragraphs, the ‘principle’ of comparative ad-
vantage — paragraphs left unchanged in the subsequent editions of the ‘School
Book’.
To produce exchange, therefore, there must be two countries, and
two commodities.
When both countries can produce both commodities, it is not
greater absolute, but greater relative, facility, that induces one of
them to confine itself to the production of one of the commodities,
and to import the other.7
When a country can either import a commodity, or produce it at
home, it compares the cost of producing at home with the cost of
procuring from abroad; if the latter cost is less than the first, it
imports.
The cost at which a country can import from abroad depends, not
upon the cost at which the foreign country produces the commod-
ity, but upon what the commodity costs which it sends in exchange,
compared with the cost which it must be at to produce the com-
modity in question, if it did not import it. (1821, 87; 1824, 117;
1826, 123)
7In this two-country / two-commodity framework, Mill supposes a complete specialisa-
tion of each country. The statement is also made in the context of the England / Poland
example (1821, 135). But, for Mill, this question does not seem to be important — his
framework is enlarged to three commodities when he introduces gold into the picture, and
even four when he supposes that a new commodity is produced in England (1821, 138),
the degree of specialisation being left aside. Ricardo himself thought that a two-country /
two-commodity framework was by far too restrictive and that incomplete specialisation was
a more realistic picture (Faccarello 2015b).
James Mill on comparative advantage 12
Ricardo’s approach is apparently followed here. But, because of the ap-
proach chosen — that is, explaining why it is advantageous for a country to
participate in international trade — the above-mentioned shifts in the analysis
are confirmed: Ricardo’s ex-post approach is replaced by an ex-ante perspective.
Mill’s argument is no longer in terms of ‘quantity’ of corn and ‘quantity’ of
cloth actually exchanged between two countries. The present formulation is
made at the individual level of commodities, and nothing can be said either
about the quantities exchanged and the international exchange ratio, or the
gains from trade.
The change is also perceptible in the fact that relative costs can be ex-
pressed through ‘ratios’. These ratios, which are not to be found in Ricardo,
are advanced by Mill (1821, 88) — who also sometimes writes of ‘the purchas-
ing power of one commodity with respect to another’ in each country (1824,
118; 1826, 124).
Thus, if labour in Poland produce corn and cloth, in the ratio of
eight yards to one quarter; but, in England, in the ratio of ten
yards to one quarter, exchange will take place.
Generally, then, when labour produces one of two commodities in
a greater ratio to the other in one than in another of two countries,
it is the interest of both to exchange. (1821, 88)
Finally, the role of the Elements in the evolution of the vocabulary used in
this approach can be further exemplified. The phrase ‘comparative advantage’
cannot be found in Ricardo’s chapter on foreign trade, but appears once inci-
dentally elsewhere in his Principles, and the same can be said of the phrases
‘comparative difficulty of production’ and ‘comparative facility [of] produc-
tion’. In Mill’s 1818 entry on ‘Colony’, these phrases are not used. However,
in Elements, Mill writes of a ‘comparative facility’ of production (1821: 84) and
clearly contrasts a ‘greater absolute’ with a ‘greater relative, facility’ of pro-
duction (1821, 87). He also speaks of ‘peculiar advantages’ with the meaning
of comparative advantage (1821, 84).8
8After Mill, things changed further. In the third and fourth editions of his Essay on
the External Corn Trade (1826, 1827), Torrens started to designate the doctrine with the
phrases ‘comparative advantage’ or ‘disadvantage’ (1826, vii) and ‘comparative costs’ (1827:
401). John Stuart Mill also used the phrase ‘comparative cost(s)’ (J. S. Mill 1829–30, 233,
James Mill on comparative advantage 13
Trade between England and Germany
A final and important point must be addressed as regards the first edition of
the Elements. In Chapter 4 (‘Consumption’), the final Section XV (‘Effects of
the taxation of commodities upon the value of money, and the employment of
capital’) presents another example of an international exchange (this example
is also to be found verbatim in the second edition of the book9but is abridged
in the third). It refers to an exchange between England and Germany, both
countries being able to produce broad cloth and linen (1821, 231). The new
data are shown in Table 5: the numbers are, in each country, the unit prices
of linen and broad cloth respectively.
Table 5: The ‘four magic numbers’ in James Mill’s 1821 Elements of Political
Economy, Chapter 4, Section XV
linen/yard broad cloth/yard
Germany 2 sh. 24 sh.
England 3 sh. 20 sh.
In this case, both countries can benefit from trade: ‘it would be the interest
of England to employ her labour in making broad cloth for Germany, instead
of linen for herself; and that of Germany, in making linen for England, instead
of broad cloth for herself’ (1821, 231). But what about the respective gains
from trade? Mill endeavours to specify England’s gains in the following way:
In England the same quantity of labour that produced one yard
of broad cloth, would produce only seven yards of linen. But in
Germany, one yard of this cloth would purchase twelve yards of
linen. England, therefore, by exporting broad cloth in exchange
for linen, would gain the whole of the difference between seven
yards of linen and twelve. (1821, 231–2)
235–6, 254): speaking of comparative advantage, he mentions a commodity in which the
‘advantage is least’ or the ‘advantage is greatest’ (1829–30, 233).
9Mill (1824, 295-6). The section was now numbered XVII.
James Mill on comparative advantage 14
The case of Germany can be stated in the same way although Mill does
not do so. In Germany, the same quantity of labour that produces one yard of
linen can produce only one twelfth of a yard of cloth, but in England one yard
of linen would purchase one seventh of a yard of cloth. By exporting linen in
exchange for cloth, Germany would gain ‘the whole of the difference’ between
one twelfth of a yard of cloth and one seventh. But Ricardo’s approach is
again left aside here. Led by his purpose — the analysis of the consequence
on trade of a tax levied on a commodity — Mill reasons in terms of units
of commodities, and the ‘four magic numbers’ are defined, in each country,
as the per-unit domestic prices of these commodities and no longer as the
quantities of labour respectively necessary to produce, in each country, the
amounts exchanged internationally. Mill also implicitly supposes that prices
reflect the labour values of the commodities: this is the reason why he can
state that, in England, ‘the same quantity of labour that produced one yard
of broad cloth, would produce only seven yards of linen’.10
What was already perceptible in Mill’s other examples and in his summary
of Ricardo’s ‘principle’ is confirmed. To state that both countries can benefit
from trade in the pre-tax world, Mill is obliged to deal with units of com-
modities and to adopt the device he previously used for Poland: each country
trades at the foreign domestic prices . . . and therefore appropriates the whole
gains from trade. The amounts exchanged between countries are unknown.
4Elements of Political Economy, 1824
In the second edition of the Elements, and as regards international exchange,
everything is left unchanged except that, at the end of Chapter 3, Section 4,
Mill replaces the last paragraph of the first edition by four new ones, which
develop the England / German example introduced at the end of the first
edition. The analysis carried out at the individual level of commodities is
confirmed: ‘Whenever the purchasing power of any commodity with respect
to another is less, in one of two countries, than it is in the other, it is the
10 Logically, however, the price of one yard of cloth in England should have been 21 instead
of 20.
James Mill on comparative advantage 15
interest of those countries to exchange these commodities with one another’
(1824, 118).
But, instead of defining the ‘four magic numbers’ as before (Table 5), as the
per-unit-prices of the two commodities in each country (an example maintained
at the end of the book, as already mentioned), Mill now refers directly to the
physical productivity of labour in these countries. As in the third presentation
of the England / Poland example in the first edition (Table 4), he now defines
the numbers as the quantities of the two commodities produced with a given
quantity of labour (not necessarily the same in the two countries). The relative
domestic values are different: 10 units of cloth = 20 units of linen in Germany,
10 cloth = 15 linen in England. These new numbers are shown in Table 6.
Table 6: The ‘four magic numbers’ in James Mill’s 1824 Elements of Political
Economy, Chapter 3, Section IV
linen cloth
Germany, for a given quantity of labour 20 yards 10 yards
England, for (another) given quantity of labour 15 yards 10 yards
The quantities exchanged between the two countries and the international
exchange ratio are unknown. Mill has no other choice here than to repeat
the device he used in the first edition, reasoning as if England, exporting her
cloth, exchanges it in Germany at the German domestic price, and Germany,
exporting her linen, exchanges it in England at the English domestic price.
. . . it is very evidently the interest of England to send broad cloth
to purchase linen in Germany because with 10 yards of broad cloth,
that is, as much cost of production as would produce 15 yards of
linen, she can obtain 20 yards.
It is equally the interest of Germany to send linen to purchase broad
cloth in England, because with 15 yards of linen she can purchase
10 yards of broad cloth in England, which, if made at home, would
cost her as much as 20 yards of linen. (Mill 1824, 118–9)
James Mill on comparative advantage 16
It is this example which is taken up by John Stuart Mill at the beginning
of his 1829–30 essay11 in order to stress the need to correct the miscalculation
of the gains and to develop the analysis further:
. . . the whole gain of both countries together, consisting in the
saving of labour; and the saving of labour being exactly equal to
the difference between the costs, in the two countries, of the one
commodity as compared with the other; the two countries taken
together gain no more than this difference: and if either country
gains the whole of it, the other country derives no advantage from
the trade. (J.S. Mill, 1829–30, 235)
Ironically, as we know, J.S. Mill attributes the mistake to Ricardo himself,
who ‘unguardedly expressed himself as if each of the two countries making the
exchange separately gained the whole of the difference between the comparative
costs of the two commodities in one country and in the other’ (J.S. Mill, 1829–
30, 235). In an attempt to play down his criticism, however, he adds that
the problem is not really due to an error but rather to ‘a mere oversight of
Mr. Ricardo, arising from his having left the question of the division of the
advantage entirely unnoticed’ (1829–30, 236). The problem, J.S. Mill points
out, was first corrected by his father in the third edition of his Elements —
even if, in his eyes, the correction was not satisfactory: ‘It can hardly, however,
be said, that Mr. Mill has prosecuted the inquiry any further [than Ricardo];
which, indeed, would have been quite as inconsistent with the nature of his
plan as of Mr. Ricardo’s’ (1829–30, 236). It is therefore time to turn to the
new developments offered by the 1826 edition.
11 ‘If England sends 10 yards of broad cloth to Germany, and is able to exchange them
for linen according to the German cost of production, she will get 20 yards of linen, with a
quantity of labour with which she could not have produced more than 15; and will gain,
therefore, 5 yards on every 15 . . . [I]f Germany sends 15 yards of linen to England, and
finding the relative value of the two articles in that country determined by the English costs
of production, is enabled to purchase with 15 yards of linen 10 yards of cloth; Germany now
gains 5 yards, just as England did before, — for with 15 yards of linen she purchases 10
yards of cloth, when to produce these 10 yards she must have employed as much labour as
would have enabled her to produce 20 yards of linen’ (J.S. Mill 1829–30, 235–6, emphasis
added).
James Mill on comparative advantage 17
5Elements of Political Economy, 1826
Some changes are indeed to be found two years later in the third edition of
Mill’s Elements. Two of them are of material importance for our subject: both
result from Mill’s attempts to dispel his former ambiguities about the countries’
gains from trade — ‘I have corrected an error of the former editions’, he states
in the new Preface (Mill 1826, iv): this was the result of criticisms raised
by some members of the discussion group, of which the young John Stuart
Mill was a member,12 which met twice a week at George Grote’s house in
Threadneedle Street.13
Development of the England / Poland example
First of all, the England / Poland example (Table 3) is still there in Chapter
3, Section IV, but the case of Poland is better explained. Mill admits that
if the exchange is made, for England, as stated before — that is, if England
exchanges bunits of cloth, produced with 150 days of labour, for aunits of
Polish corn that England can only produce with 200 days of labour — England
alone would benefit from trade and Poland would gain nothing (1826, 121).
But, he adds, ‘the power of Poland would be reciprocal’. This means that,
as already noted, Poland exports aunits of corn and exchanges them in the
English market at English domestic prices, thus getting bunits of cloth for
150 and ‘any other commodity’, including more cloth as is explicitly the case
here, for 50 (Poland’s gain from trade). But Mill also admits that, in this case,
England does not gain anything: the country imports aunits of corn, but in
12 ‘Those among us with whom new speculations chiefly originated, were Ellis, Graham,
and I: though others gave valuable aid to the discussions . . . The theories of International
Values and of Profits were excogitated and worked out in about equal proportions by myself
and Graham . . . I may mention that among the alterations which my father made in
revising his Elements for the third edition, several were grounded on criticisms elicited by
these Conversations’ (J.S. Mill 1873, 123-5). Thweatt (1986) presents a different view of
this story and concludes that the initial error concerning the division of the gains from trade
was made by John and the correction was made by James.
13 Viner remarked that William Ellis, in ‘Exportation of Machinery’, a paper published in
1825 in the Westminster Review, ‘had presented an arithmetical illustration similar to those
used by James Mill, and had concluded therefrom that the gain would be equally divided
between the two countries. It seems, therefore, that the error was detected about the same
time by several members of the group associated with James Mill’ (Viner 1937, 446).
James Mill on comparative advantage 18
the end gives in exchange bunits of cloth, plus an additional quantity of cloth
worth 50: that is, a quantity of commodities produced by 200 days of labour
in England, exactly the same cost of production of aunits of English corn.
With a quantity of corn which cost her 100 days’ labour, equal to
the quantity produced in England by 200 days’ labour, she [Poland]
could . . . purchase, in England, the produce of 200 days’ labour
in cloth. The produce of 150 days’ labour in England in the ar-
ticle of cloth would be equal to the produce of 100 days’ labour
in Poland. If, with the produce of 100 days’ labour, she [Poland]
could purchase, not the produce of 150, but the produce of 200, she
also would obtain the whole of the advantage, and England would
purchase corn, which she could produce by 200 days’ labour, with
the product of as many days’ labour in other commodities. (1826,
121–2)
Mill’s conclusion is puzzling: ‘The result of competition would be to divide
the advantage equally’ between the two countries (1826, 122).14 He is prob-
ably referring to the new developments brought to the England / Germany
example which immediately follows and in which the advantage is shared, but
not ‘equally’.
The statement is repeated in Section XV of the same chapter, where the
England / Poland example is again mentioned with different numbers (Table
4). In the first and second editions, Mill simply writes that a country’s export
is exchanged abroad at the foreign domestic prices, and, implicitly, that each
country benefits from trade (1821, 135-6; 1824, 173-4). In the third edition,
14 Ellis also supposed that the global gain from trade is equally divided between the two
countries. He depicts a situation where 100 men in England can produce 1,000 yards of silk
or 2,000 yards of cotton, while 100 men in France can produce 2,000 yards of silk and 1,000
yards of cotton. If each country specialises in one production (silk in France and cotton
in England, with 200 men in each sector) and exchanges, they both benefit from trade
because, together, they produce more commodities than before: but the individual gains
from trade are not specified. Ellis then supposes that the productivity of labour doubles
in England, and that 100 workers there can produce 2,000 units of silk and 4,000 units of
cotton. His general conclusion as regards the benefits from trade is unchanged: France and
England, specialising respectively in silk and cotton and each employing 200 units of labour,
produce together 4,000 units of silk and 8,000 units of cotton. Now suppose that free trade
is impossible: the global production of cotton falls to 5,000 yards and Ellis simply stresses
that the global loss will be shared equally, ‘the loss to be divided between the two countries
being 3,000 yards of cotton or 1,500 each’ (Ellis 1825, 388). A similar analysis is made in
the case of a more explicit technical progress (1825, 388–9).
James Mill on comparative advantage 19
this passage is deleted and replaced by another one. Mill now simply states
that the benefits from trade are divided between England and Poland, each
country’s gain being worth 5 yards of cloth — ‘In a traffic of these commodities,
between England and Poland, there would be a value of 5 yards of cloth to
be gained by each upon every repetition of the transaction’ (1826, 178) — a
statement which is clearly hardly more comprehensible.
Development of the England / Germany example
The England / Germany model was first developed at the end of the first edi-
tion (Chapter 4, Section XV) (Table 5). It remained unchanged in the second
edition, but in addition it was also inserted and more extensively presented
in the chapter on international trade, with new ‘magic numbers’ (Chapter 3,
Section IV) (Table 6). In the third edition, this latter presentation is further
developed, and that of the final section (now Section XVII of Chapter 4) on
taxation is abridged. In this last section and after mentioning the per-unit-
prices of the commodities in each country (as in Table 5 above), Mill simply
adds:
It has been already seen, how, in these circumstances, it would be
the interest of England to employ her labour in making broad cloth
for Germany, instead of linen for herself; and that of Germany, in
making linen for England, instead of broad cloth for herself. (1826,
300)
In Chapter 3, however, Mill recognises the mistake made in the former edi-
tions and admits that, when England, Poland or Germany trade at the domes-
tic prices of one of them, the country whose domestic prices are used does not
benefit from trade — and can even lose because of the ‘cost of carriage’. The
international relative prices must therefore lie between the domestic relative
prices. However, Mill does not say anything about the quantities exchanged
and the final global gains from trade.
The ‘magic numbers’ are those shown in Table 6: ‘10 yards of broad cloth
purchase 15 yards of linen in England; and 20 yards in Germany’. But now
Mill states that the domestic relative prices of the two countries must change
because of the opening to international trade and competition. At the begin-
ning, England, exporting 10 yards of broad cloth to Germany, can buy there
James Mill on comparative advantage 20
20 yards of linen and makes a benefit of 5 yards of linen. Germany, we can
suppose, exporting 20 yards of linen to England, can buy more than 10 yards
of cloth in the English market. But, Mill adds, these exchange ratios cannot be
maintained. They will change, and, as a consequence, the gains of each coun-
try will be modified: ‘it is evident that the advantage will be shared’ (1826,
122). The result however cannot be known in advance and it is no longer a
matter of dividing the advantage ‘equally’ between the two countries (1826,
122).
In England linen will fall, in relation to cloth, from the knowledge
that 10 yards of cloth will purchase more than 15 yards of linen in
Germany; and in Germany linen will rise as compared with cloth,
from a knowledge that 20 yards of linen, if sent to England, will
purchase more than 10 yards of cloth. It is the inevitable effect of
such an interchange to bring the relative value of the two commodi-
ties to a level in the two countries; that is, to make the purchasing
power of linen in respect to cloth, and of cloth in respect to linen,
the same in both; bating the difference in the cost of carriage, each
country paying the cost of the carriage of the commodity which
it imports, and the value of that article being so much higher in
the country which imports than in that which exports it. (1826,
122–3)
Mill does not further specify the working of competition. Are the domestic
prices of the exchanged commodities now disconnected from their labour val-
ues, and do market prices definitely prevail over natural prices, and supply and
demand over labour values? Or, in each country, are the domestic prices of
imported commodities equal to their natural prices in the exporting countries,
plus the ‘cost of carriage’? James Mill is elusive on that point, but it seems
that he adopts Ricardo’s solution, that is, the second one just mentioned (Fac-
carello 2015b, 774–5). Three years later, John Stuart Mill was to adopt the
first.
As a matter of fact, for Ricardo, if the transportation costs are neglected,
‘no more will be paid for commodities by foreign purchasers than by home
purchasers; the price which they will both pay will not differ greatly from
their natural price in the country where they are produced’ (1951–73, I, 340–
41). Suppose, Ricardo argues, that one unit of corn costs £3 in France and £6
in England because of a prohibition on the importation of corn in this country.
James Mill on comparative advantage 21
Suppose now that the prohibition is removed. Importation of corn from France
changes the corn price in England. But by how much? All other things being
equal, until the English consumer pays the French natural price:
. . . corn would fall in the English market, not to a price between
6l. and 3l., but ultimately and permanently to the natural price
of France, the price at which it could be furnished to the English
market, and afford the usual and ordinary profits of stock in France.
(1951–73, I, 374–5)
In the context of his monetary analysis (below), and as a consequence of an
exchange of corn and cloth between England and Poland, James Mill asserts
that, in the final state of equilibrium, ‘in England corn is dearer than in Poland,
by the expense of carrying corn; cloth is dearer in Poland than in England,
by the expense of carrying cloth, from the one country to the other’ (1821,
137). The idea is repeated some pages later (1821, 154), and nothing more can
be found in the text. Nevertheless, as he wrote his book in order to explain
and support Ricardo’s ideas, it is reasonable to suppose that he adopted the
same solution.15 John Stuart Mill however, introducing reciprocal demands
and elasticities of demand, put the analysis on a different track.16
6Money and international trade
The transformations made by James Mill thus lead us away from Ricardo. It
seems, however, that Mill remained faithful to an important aspect developed
15 Ricardo commented on Mill (1821, 153–4), but his criticism only concerns the costs
to be included in the price of an imported commodity. ‘If I consent to send money from
England, to import corn from Poland, when it arrives in England, it must be of a value,
not only equal to its cost in Poland, but also equal to the charges of sending the money and
conveying the corn. One hundred qrs of corn in Poland are worth we will suppose £200[,] the
charge of sending the money £5. If I send the money therefore the corn will cost me £205,
but the charge of bringing the corn to England we will suppose to be £10, when therefore it
arrives in England it will stand me in £215 and unless I sell it for more than that sum I get
no profit’ (Ricardo 1951-1973, IX, 130). The allusion to profit illustrates the frequent shift,
in the Principles, from labour values to production prices.
16 In his Autobiography, he observed with some regret that, while his father ‘modified
his opinions’ after the criticism made by Graham and himself regarding the theories of
international values and of profits, James Mill did not do so ‘to the extent of our new
speculations’ (J.S. Mill 1873, 125).
James Mill on comparative advantage 22
in the Principles of Political Economy and Taxation: the monetary context of
the chapter on foreign trade, which is given pride of place by Ricardo. The
‘four numbers’ in the examples developed in the Elements, Mill states, only
show what the countries’ interest would be to participate in international trade,
‘in the way of barter, and without the intervention of money’ (1821, 136).17
Money must therefore be introduced into the analysis. Mill’s developments are
to be found in Chapter 3, Sections XII to XV.18
Mill first correctly stresses that, strictly speaking, there is no barter, and
that the decisions to export or import commodities are made upon considera-
tion of their money (gold) prices at home and abroad.
In ordinary language . . . those commodities alone can be ex-
ported, which are cheaper in the country from which they go, than
in the country to which they are sent; and . . . those commodities
alone can be imported, which are dearer in the country to which
they come, than in the country from which they are sent. (1821,
128)
He also insists that gold is a commodity like any other and follows the same
rules of exchange.
Gold is cheap, when a greater quantity of it is required to pur-
chase commodities; and commodities are dear, for the same rea-
son; namely, when a greater quantity of gold is required to pur-
chase them. When the value19 of gold, therefore, in England, is
low, gold will be exported from England, on the principle that all
commodities which are free to seek a market, go from the place
where they are cheap to the place where they are dear. But as, in
the fact that gold is cheap, is implied the correlative and insepa-
rable fact that other commodities, at the same time, are dear, it
follows, that, when gold is exported, less of other commodities can
be exported; that no commodities can be exported, if the value of
gold is so low as to raise the price of all of them above the price in
other countries. (1821, 129)
17 This sentence is deleted in the third edition.
18 Sections XIII to XVI from the second edition onwards.
19 ‘When we speak of the value of the precious metal, we mean the quantity of other
things for which it will exchange’ (1821, 131).
James Mill on comparative advantage 23
Hence a general statement: ‘a country will export . . . commodities, other
than the precious metals, only when the value of the precious metals is high’
and, accordingly, ‘she will import, only when the value of the precious metals
is low’ (1821, 129). But the value of the precious metals, for a country without
gold or silver mines, is dependent on foreign trade, as the price-specie flow
mechanism states: an inflow of these metals lowers their value and an outflow
increases it. Money prices vary in the opposite direction.
Mill’s aim, however, is only to show that the introduction of money does
not change the analysis in real terms, that is, in terms of ‘barter’ — a per-
spective which supposes a dichotomy between real and monetary analyses in
Ricardo’s views on foreign trade.20 In his 1818 entry ‘Colony’, he alludes to
this. Concluding his example of a trade between England and Poland, he
writes:
In what manner this class of transactions is affected by the inter-
vention of the precious metals; in what manner the precious metals
distribute themselves, so as to leave the motives to this barter ex-
actly the same as they would be, if no precious metal intervened,
it would require too many words here to explain. (1818, 27)
Mill then refers the reader to Ricardo’s book.21 In the Elements, instead,
the analysis is developed in Chapter 3, Section XIV22 , where the England
/ Poland example is again examined with the changes already noted above
(Table 4). In each country, the quantities of corn and cloth produced there are
of equal value. In England, for instance, four quarters of corn and twenty yards
of cloth have the same value, being produced with the same amount of labour;
and similarly in Poland, where four quarters of corn has the same value as ten
20 In the second section of his essay ‘Of the laws of interchange between nations’, John
Stuart Mill develops the same approach. ‘We shall now examine’, he writes, ‘whether the
same law of interchange, which we have shown to apply upon the supposition of barter, holds
good after the introduction of money. Mr. Ricardo found that his more general proposition
stood this test; and as the proposition which we have just demonstrated is only a further
development of his principle, we shall probably find that it suffers as little, by a mere change
in the mode (for it is no more) in which one commodity is exchanged against another’ (J.S.
Mill 1829–30, 241).
21 ‘The reader who recurs for that explanation to Mr. Ricardo, the first author of it, will
not lose his time or his pains’ (1818, 27). Ricardo’s analysis, however, is more complex.
22 Section XV in the second and third editions.
James Mill on comparative advantage 24
yards of cloth. With the introduction of money, gold prices reflect values: in
Poland, ‘the price of four quarters of corn and 10 yards of cloth would be the
same’ and, in England, ‘the price of four quarters of corn and that of 20 yards
of cloth would be the same’ (1821, 136). Mill then compares the price of corn
in both countries and supposes that it is the same: as a consequence, the price
of one unit of cloth is twice as high in Poland as it is in England. Cloth will
be exported to Poland and exchanged there for gold.
In these circumstances, what will happen is obvious: the cloth,
which is cheap in England, will go to Poland, where it is dear;
and there it will be sold for gold, because there can be no counter
importation of corn, which, by supposition, is already as cheap in
England as in Poland. (1821, 136)
Some gold leaves Poland, where prices diminish, and flows into England,
where prices rise: as a consequence, ‘the price of corn in the two countries
gradually recedes from equality, the price of cloth gradually approaches it’
(1821, 137). Whenever the price of corn in England exceeds the price in Poland
by more than the transportation costs, England will import it. The process
stops when an equilibrium is found, that is to say, when the prices of corn
and cloth are the same in both countries, plus the respective transportation
costs which have to be added to the price of the imported commodity in the
importing country.
. . . prices regulate themselves in such a manner, that in England
corn is dearer than in Poland, by the expense of carrying corn; cloth
is dearer in Poland than in England, by the expense of carrying
cloth, from the one country to the other. At this point, the value
of the cloth imported into the one country, and that of the corn
imported into the other, balance one another. The exchange is
then at par, and gold ceases to pass. (1821, 137–8)
We can now understand the narrow scope of Mill’s analysis. To show ‘in
what manner the precious metals distribute themselves, so as to leave the mo-
tives to this barter exactly the same as they would be, if no precious metal
intervened’, as stated in 1818, just amounts to showing that, for each coun-
try, the direction of trade is the same, whether it is determined in a context
James Mill on comparative advantage 25
of barter or of monetary exchanges. But nothing more can be said, espe-
cially about the quantities exchanged and the gains from trade, which are still
unknown. Mill does not properly understand the complexity of Ricardo’s mon-
etary analysis or what Ricardo meant when he used the phrase ‘equilibrium of
barter’. It is true that he does not feel very comfortable with these matters.
‘Every thing, too, has come within a narrow compass, except money’, he wrote
to Ricardo on 28 December 1820, speaking of the writing of his ‘School Book’.
‘So many different circumstances had to be noticed, on that subject, that it
has been tedious to me in the writing’ (in Ricardo 1951–73, VIII, 327).
7Concluding remarks
James Mill’s Elements of Political Economy was intended to be a ‘School
Book’ explaining the basic developments of the new science of political econ-
omy. This is the reason why no reference to any author is made in the text: ‘I
have thought it advisable not to quote any authorities, because I am anxious
that the learner should fix his mind upon the doctrine and its evidence, without
extraneous consideration’ (1821, iv). It was obvious to readers, however, that
the four chapters were written along Ricardo’s line of thought and were based
on the Principles. For decades, the book was even preferred to the Principles
of Political Economy and Taxation, too difficult to read and understand. Ele-
ments was a kind of ‘Ricardo without tears’, despite McCulloch’s opinion that
the book was not really for beginners. During the mid-1820s, when political
economy was studied by the members of John Stuart Mill’s discussion group,
the Elements were discussed first, before the Principles.23
James Mill insisted: ‘I profess to have made no discovery’ (1821, iv). But
as regards the presentation of the ‘magic numbers’, at least, this assertion is
a half-truth, which left Mill’s new approach unnoticed — and most probably
unnoticed by Mill himself! Ricardo is, perhaps, somewhat responsible for this
state of things: after reading the Elements, he did not make any comments on
the way in which his views on foreign trade were presented, except on a few
23 ‘We chose some systematic treatise as our text-book; my father’s Elements being our
first choice . . . When we had finished in this way . . . we went in the same manner through
Ricardo’s Principles of Political Economy, and Bailey’s Dissertation on Value’ (J.S. Mill
1873, 123).
James Mill on comparative advantage 26
details. For a modern reader, this attitude is of course puzzling. One can only
presume that Ricardo was more directly concerned with topics which were
under fire at that time and on which Mill’s statements seemed particularly
inadequate: he criticised Mill on value and money.
What is more, the perspective of a perfect continuity with Ricardo was
asserted by John Stuart Mill. In 1829–30, after having stated that the pur-
pose of his essay ‘Of the laws of interchange between nations’ was to study
the distribution of the gains from trade among countries, he also insisted that
the analysis was simply the consequence of the first principles established by
Ricardo, which Ricardo himself, too busy with the establishment of the foun-
dations of the science, had no time to develop.
This question was not entered into by Mr. Ricardo, whose at-
tention was engrossed by far more important questions, and who,
having a science to create, had not time, or room, to occupy himself
with much more than the leading principles. When he had done
enough to enable any one who came after him, and who took the
necessary pain, to do all the rest, he was satisfied. He very rarely
followed out the principles of the science into the ramifications of
their consequences. (1829–30, 235)
All these facts, taken together, explain why James Mill’s new approach
was easily accepted as a faithful interpretation of the doctrine, to the point
that subsequent readings of Ricardo’s Chapter 7 were systematically biased
and made through Mill’s lens.
In James Mill, the process of change was subtle and gradual. (i) Mill’s
didactic purpose led him to start with an example involving two countries,
England and Poland, between which no exchange was possible because, in
each of them, aunits of corn and bunits of cloth can be produced with the
same quantity of labour (100 in Poland, 150 in England), a hypothesis which
also means that the relative domestic prices are the same in both countries.
The choice of such particular numbers was made in order to show when a
profitable exchange could take place between the two countries. (ii) In order
to make such an exchange possible, Mill changed only one of the ‘four numbers’
— England now needs 200 days of labour to produce aunits of corn — leaving
Poland’s costs of production unchanged and still equal in the two sectors of
production. He thus ended with a situation where, in order to show that
James Mill on comparative advantage 27
Poland can also benefit from trade, he was led to suppose that this country
exchanges aunits of corn in the English market at the English domestic relative
prices — therefore getting more than bunits of cloth. (iii) Later in the book
the ‘four numbers’ of the model appeared as simple technical data, that is to
say, quantities of commodities produced with a given amount of labour. (iv)
Finally, summing up the main principles of this approach, and still in order to
show when it is in the interest of two countries to exchange, he reasoned at the
individual (per-unit) level of commodities, an approach confirmed by a second
example involving England and Germany, in which the ‘magic numbers’ are
the per-unit prices of the commodities.
All this results from Mill’s change of perspective. Instead of the ex-post
analysis of the Principles of Political Economy and Taxation — that is, in-
stead of starting with an actual international exchange in which the quantities
of commodities exchanged, the exchange ratio and the global gains from trade
are given — Mill adopts an ex-ante approach, already outlined in his 1808
Commerce Defended. Starting from a no-trade situation, the direction of the
international flows of commodities is explained but the quantities exchanged,
the international relative price and the benefits from trade now have to be de-
termined. With Mill, the ‘four magic numbers’ acquire a prescriptive flavour.
And money is only introduced to show that it does not disturb the develop-
ments made on a ‘barter’ basis. For Ricardo however (Faccarello 2015b), what
has been called ‘comparative advantage’ is no ‘principle’ at all, it is not a rule
to be followed by countries in order to obtain ‘gains from trade’: it is just
the outcome, at the macro level, of the actions of free individuals in markets,
looking for profits and basing their decisions on money prices, the global ‘gains
from trade’ being simply the unintended consequences of these actions.
All this was achieved in 1821: the new developments found in the subse-
quent editions of the Elements only tried to solve some problems generated
by this approach. The essential ingredients of the ‘Ricardian’ model of for-
eign trade are already there and the interpretation of the ‘four numbers’ as
technical coefficients is merely a consequence of the new perspective.
James Mill on comparative advantage 28
Acknowledgements
This paper originated in discussions with Taichi Tabuchi during my visiting period
at Doshisha University in 2016. Comments by Victor Bianchini, Christian Gehrke,
Taro Hisamatsu, Shin Kubo, Heinz D. Kurz, Taichi Tabuchi, Susumu Takenaga and
two anonymous referees are gratefully acknowledged.
Keywords
David Ricardo, James Mill, John Stuart Mill, Comparative advantage, Four magic
numbers.
Abstract
During recent decades, David Ricardo’s ideas on international trade have been sub-
mitted to new scrutiny. This research has led to a new understanding of the so-called
‘principle of comparative advantage’ and shown that the alleged ‘Ricardian’ model
of foreign trade is based on a misunderstanding of Ricardo’s text. In this story, the
parts played by James and John Stuart Mill, James Pennington and Robert Torrens
in the creation of the Ricardian vulgate are usually mentioned. The present paper
focuses on the role of James Mill, examining in detail his article ‘Colony’ (1818) and
the three editions of his Elements of Political Economy (1821, 1824, 1826). It shows
how the evolution of Mill’s thought, in part due to a didactic perspective and some
difficulties raised by his numerical examples, was gradual and distorted Ricardo’s
approach. In Mill we find in the end all the ingredients of what is called the ‘Ri-
cardian’ model of foreign trade, substituting an ex-ante perspective for Ricardo’s
ex-post approach.
References
Aldrich, John (2004). ‘The discovery of comparative advantage’. Journal of the
History of Economic Thought,26 (3), 379–99.
Boyer, Jérôme de (2017). ‘Money and Ricardo’s “magic numbers”’. In
Shigeyoshi Senga, Masatomi Fujimoto and Taichi Tabuchi (eds), Ricardo and
International Trade, New York: Routledge, pp. 120–138.
Ellis, William (1825). ‘Exportation of Machinery’. Westminster Review, April,
386–394.
Faccarello, Gilbert (2015a). ‘Comparative advantage’. In The Elgar Companion to
David Ricardo, Heinz D. Kurz and Neri Salvadori (eds), Cheltenham: Edward
Elgar, pp. 69–77.
James Mill on comparative advantage 29
Faccarello, Gilbert (2015b). ‘A calm investigation into Mr Ricardo’s principles
of international trade’. The European Journal of the History of Economic
Thought,22 (5), 754–90.
Gehrke, Christian (2015). ‘Ricardo’s discovery of comparative advantage revisited.
A critique of Ruffin’s account’. The European Journal of the History of Eco-
nomic Thought,22 (5), 791–817.
Gehrke, Christian (2017). ‘Foreign trade, international values, and gains from trade:
Ricardo, Pennington, Whewell, and John Stuart Mill’. In Shigeyoshi Senga,
Masatomi Fujimoto and Taichi Tabuchi, Ricardo and International Trade, New
York: Routledge, pp. 139–63.
Hisamatsu, Taro (2019). ‘On the Ricardian model of international trade’. Paper
presented at the 32th HETSA Conference, University of Sidney, Australia.
Maneschi, Andrea (1998). Comparative Advantage in International Trade. A His-
torical Perspective. Cheltenham: Edward Elgar.
Maneschi, Andrea (2004). ‘The true meaning of Ricardo’s four magic numbers’.
Journal of International Economics,62 (2), 433–43.
McCulloch, John Ramsay (1824). A discourse on the rise, progress, peculiar ob-
jects, and importance of political economy, containing an outline of a course of
lectures on the principles and doctrines of that science. Edinburgh: Archibald
Constable.
Mill, James (1808). Commerce Defended. An Answer to the Arguments by which
Mr. Spence, Mr. Cobbett, and Others, have attempted to prove that Commerce
is not e Source of National Wealth. London: C. and R. Baldwin.
Mill, James (1818). ‘Colony’. Supplement to the Fourth and Fifth Editions of the
Encyclopaedia Britannica, vol. III. Reprinted in Supplement to the Fourth,
Fifth and Sixth Editions of the Encyclopaedia Britannica, vol. III, Edinburgh:
Archibald Constable, pp. 257–73. Quoted after a reprint of the entry, ‘The ar-
ticle Colony, reprinted from the Supplement of the Encyclopaedia Britannica’,
33 pp.
Mill, James (1821). Elements of Political Economy. First edition. London: Bald-
win, Cradock and Joy.
Mill, James (1824). Elements of Political Economy. Second edition. London:
Baldwin, Cradock and Joy.
Mill, James (1826). Elements of Political Economy. Third edition. London: Bald-
win, Cradock and Joy.
Mill, John Stuart (1829–30). ‘Of the laws of interchange between nations, and the
distribution of the gains of commerce among the countries of the commercial
world’. In Essays on Some Unsettled Questions of Political Economy, London:
John W. Parker, 1844, pp. 1–46. As in Collected Works of John Stuart Mill,
James Mill on comparative advantage 30
vol. IV: Essays on Economics and Society, 1824–1845, John M. Robson (ed.),
Toronto: The University of Toronto Press, 1967, pp. 232–61.
Mill, John Stuart (1873). Autobiography. London: Longmans, Green, Reader and
Dyer. As in Col lected Works of John Stuart Mill, vol. I, Autobiography and
Literary Essays, John M. Robson and Jack Stillinger (eds), Toronto: The
University of Toronto Press, and London: Routledge and Kegan Paul, 1981,
pp. 1–290.
Prévost, Guillaume (1825). ‘Réflexions du traducteur sur le système de
Ricardo, contenant l’exposé et la discussion sommaires des principes et des
conséquences de ce système’. In J.R. McCulloch, Discours sur l’origine, les
progrès, les objets particuliers, et l’importance de l’économie politique. Traduit
de l’anglais par G.me [Guillaume] Prévost, Genève: Jean-Jacques Paschoud.,
pp. 155–204.
Ricardo, David (1951–73). The Works and Correspondence of David Ricardo, edited
by Piero Sraffa with the collaboration of M. H. Dobb. Cambridge: Cambridge
University Press.
Ruffin, Roy (2002). ‘David Ricardo’s discovery of comparative advantage’. History
of Political Economy,34 (4), 727–48.
Ruffin, Roy (2005). ‘Debunking a myth: Torrens on comparative advantage’. His-
tory of Political Economy,37 (4), 711–722.
Sraffa, Piero (1930). ‘An alleged correction of Ricardo’. The Quarterly Journal of
Economics,44 (3), 539–545.
Sraffa, Piero (1951). ‘Introduction’. In The Works and Correspondence of David
Ricardo, edited by Piero Sraffa with the collaboration of M. H. Dobb. Cam-
bridge: Cambridge University Press, I, pp. xiii–lxv.
Tabuchi, Taichi (2017). ‘Yukizawa’s interpretation of Ricardo’s theory of compara-
tive costs’. In Shigeyoshi Senga, Masatomi Fujimoto and Taichi Tabuchi (eds),
Ricardo and International Trade, New York: Routledge, pp. 48–59.
Takenaga, Susumu (2019). ‘Ricardo’s theory of foreign trade: its relation to his
theory of money and theory of profit’. Paper presented at the 23rd Annual
Conference of the European Society for the History of Economic Thought
(ESHET), University of Lille, France.
Thweatt, William O. (1976). ‘James Mill and the Early Development of Compara-
tive Advantage’. History of Political Economy,8(2), 207–34.
Thweatt, William O. (1986). ‘James and John Mill on Comparative Advantage:
Sraffa’s Account Corrected’. In Trade in Transit, edited by H. Visser and E.
Schoorl, Dordrecht: Martinus Nijhoff, pp. 33-43.
Torrens, Robert (1826). Essay on the External Corn Trade. 3rd ed. London:
Longman, Rees, Orme, Brown and Green.
James Mill on comparative advantage 31
Torrens, Robert (1827). Essay on the External Corn Trade. 4th ed. London:
Longman, Rees, Orme, Brown and Green.
Viner, Jacob (1937). Studies in the Theory of International Trade. New York:
Harper & Brothers.
Yukizawa, Kenzo (1974). ‘The original meaning and the deformed interpretation
of Ricardo’s theory of comparative costs’. Shogaku Ronsan,15 (6), 25–51 (in
Japanese).
Yukizawa, Kenzo (1978) ‘Formation of classical theory of international trade: Ri-
cardo and James & John Stuart Mill’. In On Method and History of Social
Science, edited by K. Yukizawa, M. Tanaka, T. Hirai and K. Yamaguchi, Ky-
oto: Minerva Shobo, pp. 203–24 (in Japanese).
