Some Aspects of Protection Further Considered
Author(s): Frank D. Graham
Source: The Quarterly Journal of Economics, Vol. 37, No. 2 (Feb., 1923), pp. 199-227
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/1883929
Accessed: 21-06-2017 09:41 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The
Quarterly Journal of Economics
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
THE
QUARTERLY JOURNAL
OF
ECONOMICS
FEBRUARY, 1923
SOME ASPECTS OF PROTECTION
FURTHER CONSIDERED
SUMMARY
Persistence of protectionist sentiment and its causes. - The principle
of comparative advantage not an infallible guide in the economics
of commercial policy, 203. - Effect of increasing and decreasing
unit costs, 205. - Mathematical demonstration, 207. - Suggested
tests for the modification of the principle of comparative advantage,
210. - Analysis of the "unit of productive power," 212. - Effect of a
change in the composition of this unit, 213. - Protection as an agent
in bringing about such a change, 216. - Extension of the infant-industry
argument to cover infant finance, 218. - Past and future relation of
protection and prosperity, 225.
THE lack of correspondence of the theory of international
trade with the commercial policy of most
nations has provoked frequent comment. In noting the
resurgence of protection since 1860, M. Gide suggests
that, in spite of all that has been written on the subject,
the matter has not yet been completely explored.
"There must have been some general causes at the root
of this sudden irresistible and spreading epidemic of
protection," he says, "but it is not very easy to discover
them." And again, "the reaction in favor of Protection
is not so marked in theory as it is in trade policy . . .
the greater number of economists have remained faithful
to the Free Trade doctrines." I One may doubt
1. Gide, Political Economy, 3d ed., p. 349.
199
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
200 QUARTERLY JOURNAL OF ECONOMICS
whether, in explanation of this phenomenon, it is necessary
to look for any more recondite "general causes"
than national prejudices and jealousies, private interest
coupled with public indifference, and the mental inertia
which, by precluding the consideration of other than
obvious effects, offers an enormous tactical advantage
to the advocates of protection. But accepting M. Gide's
view that these are not fully adequate, the purpose of
the present article is to suggest some such "general
causes " as Gide seeks, causes which have perhaps played
their part in the new protection and which hitherto have
been somewhat obscure. The sequel will endeavor to
show: (1) that the principle of comparative advantage
is no infallible criterion of the best commercial policy,
even from a purely economic point of view; and (2) that
protection to manufactures may advantageously be
continued much longer than would seem adequate to
cover the infant stage, whether or not the industry
could maintain itself without such aid.
Economists have perhaps been too ready to attribute
protection to sheer illusion, to a Neo-Mercantilism
which has less justification than its original, much as
that original is divided. The cult of protection is not to
be explained altogether on the basis of illusion. If all its
roots lay in error, sooner or later its vogue might confidently
be expected to languish. But year by year it
seems to grow stronger, winning fresh adherents and
strengthening its hold upon its former devotees. This
penchant for protection suggests more solid origins than
the common specious pleas disclose, and in fact, when
all these flimsy supports have been swept away, some
economic props remain.
By far the most important cause for the spread of
protective measures has been the desire to encourage
manufactures. The plain citizen of almost every land
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 201
idealizes, nay almost idolizes, the industrial state. A
smiling and peaceful Arcadian land as a national ideal
has no longer any charms - the fascination of the machine
has changed all that - and men want their national
skies black with the smoke of industry and
clanging with its noises. This preference may hark back
to the instinct for contrivance, it may be a search for
security, it may have roots which I shall presently indicate,
in great part no doubt it has no profounder basis
than a crude and shallow Mercantilism. At all events it
is a preference and a force and, in view of this preference,
manufacturing might almost be considered an
economic good in itself. With scarcely an exception
non-industrial states display a strong desire to develop
manufacturing within their borders, while there is
nothing like the same ardor with regard to the development
of agriculture. Agriculture has indeed at times
been given protection in an industrialized state such as
Germany, but only as a sop, and against the most
strenuous opposition. No state has shown any desire to
move from an industrial toward an agricultural regime,
while the decline of agriculture in favor of manufacturing
industry is accepted with equanimity. This widespread
preference for manufactures is the secret of the
"irresistible and spreading epidemic of protection," for
it is for the development of manufactures that protection
has usually been invoked. And so where protection
is unnecessary or ineffective to promote manufacturing,
as in England, it is not popular, while its popularity
remains unshaken wherever manufacturing is likely to
be advanced as a result of its use.
But if the preference for manufactures is the secret of
protection, what is the secret of the preference for manufactures?
The typical economist scoffs at any preference
for manufactures per se. Relying on deductive logic he
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
202 QUARTERLY JOURNAL OF ECONOMICS
comes to conclusions which run foul of empirically obtained
lay opinions to the effect that manufactures, as
such, make for prosperity. These latter opinions are on
this point perhaps no further from the truth than his
own, which are apt to issue out of narrow and questionable
premises.
For it is the premises of the free trader, not his conclusions,
that have always been and always must be the
object of any effective attack. It was through this hole
in the free-trade armor that Hamilton and List made
their most telling thrusts, but the blows of neither were
as shrewd as they might have been. List's ideas of an
"educative" tariff and of the development of a nation
through various stages to what he regarded as its apogee,
the agricultural-manufacturing-commercial phase, are
fundamentally sound, and it can plausibly be maintained
that the stimulus that a tariff may give in this
development may well be worth a present loss, provided
it sets free latent productive powers. Manufactures
probably promote skill quite as much as skill promotes
manufactures. But List's error (a strange one for him to
make) lay in conceding too much to his opponents. For
he thought of "nurturing" protection as a temporary
policy merely, a temporary loss which could be compensated
only by the development of the protected industry
to a point where it could stand without support.
If the industry could not eventually do this, List would
have regarded the protection as unwisely granted, and
as well that it should be withdrawn. He was in fact a free
trader in principle, admitting protection as a temporary
expedient only. It was on this point that he failed to
make the most of his position. For, as I shall now endeavor
to show, it may be to a country's economic advantage
to protect an industry which could not grow up
or survive without protection and which never will be
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 203
able to survive without it, an industry w
comparative advantage when the protective duty is
first levied nor ever attains one under it.
To make this clear let us assume two countries, A and
B, under three distinct sets of conditions. In the first
set A and B are just opening up a trade which has
hitherto not existed. In the second set the trade has
developed greatly, free trade between the two has obtained,
and each country has specialized along lines of
comparative advantage. In the third set of conditions,
which are contemporaneous with the second, protective
tariffs are supposed to have been levied from the beginning
by B against A; A's products have been excluded,
and each country has produced at home its home consumption.
The trade which was being opened up under
the first set of conditions has been stifled. Let us assume
further that, instead of production being at constant
cost, as the classical theory of international trade supposed,
country A has under the first set of conditions,
and retains throughout, a comparative advantage in the
production of a commodity subject in the prevailing
conjuncture to decreasing unit cost, - watches, let us
say, - while country B's comparative advantage lies in
the production of a commodity subject in the prevailing
conjuncture to increasing unit cost, say wheat. This
means, of course, that the effort cost petr unit of A's
product will decline solely by reason of an extension of
output; while the effort cost per unit of B's product will
increase for the same reason.2 Assume further that the
2 These are aspects of the law of proportionality The assumed situation is such that
an increased output of watches makes possible more efficient combination of agents,
while an increased output of wheat renders inevitable a less efficient combination In
the case of the commodity produced under conditions of decreasing unit cost, watches,
an extension of total output may be achieved by simply increasing the number of individual
producing organizations without any increase in the size of the individual
organization If this is the case, the only economies obtained will be external economies,
which come slowly and appear on a considerable scale only if there be a very great
increase in output If, on the other hand, some of the existing plants share in the exThis
content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
204 QUARTERLY JOURNAL OF ECONOMICS
total sales of the commodity in which A has a comparative
advantage, viz., watches, are likely to grow, while
those of the commodity in which the comparative advantage
lies with B, viz., wheat, will tend to remain
constant. Then the three sets of conditions referred to
above may be outlined as follows:
CASE I
In Country A
10 days' labor produces 40 units wheat.
10 days' labor produces 40 watches.
In Country B
10 days' labor produces 40 units wheat.
10 days' labor produces 30 watches.
Country A has a comparative advantage in watches,
country B in wheat. Trade is possible on any terms between
the limits 40 wheat = 30 @ 40 watches. Suppose
it is opened up and equilibrium established on the basis
of 40 wheat for 35 watches. Then country A, for 20
days' labor devoted to the production of watches ( = 80
watches) can, by the exchange with B of 35 watches for
40 wheat, obtain 40 wheat plus 45 (80 - 35) 3 watches,
while country B for 20 days' labor devoted to the production
of wheat (= 80 wheat) can, by the same exchange,
obtain 40 (80 - 40) wheat plus 35 watches. Both
tension of output, and the tendency toward a larger-scale producing organization is
increased thereby, internal economies will be obtained, and the unit cost may be diminished
very considerably. The reasoning in the text simply assumes that a decreasing
unit cost is obtained by an expansion of the production of watches; whether the cause of
it be external or internal economies is immaterial to the theory, tho it would, of course,
affect the degree of its applicability. Conversely, an expansion of the production of
wheat is assumed to result in an increasing unit cost, the causes being immaterial to the
present theory. As a matter of probability, the more effective combination of agents in
the production of watches will be due to the increased technical efficiency which the
expanded production makes possible, while in the production of wheat the less effective
combination of agents will be due to the relative scarcity of good agricultural land.
3. This and all similar expressions which follow are not to be construed as factored
expressions. The figures in brackets are intended to show merely how the expression
immediately preceding the bracket is obtained. Thus the 45 watches are the remainder
of A's total product of 80 watches after 35 watches have been exchanged with B for
40 units of wheat.
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 205
countries get 5 more watches than they could get without
the international exchange of products, for without
such exchange A could get for 20 days' labor equally
divided between wheat and watches 40 wheat plus 40
watches instead of 40 wheat plus 45 watches, and B in a
similar way would get 40 wheat plus 30 watches instead
of 40 wheat plus 35 watches.
It should be noted that in the case here assumed the
international exchange of products is just at its inception
and so will be on a small scale. For the most part,
both countries are producing their own consumption of
both commodities. Let us go now to the conditions of
Case II. Trade has been kept free and has developed so
that both countries specialize in the lines in which they
have a comparative advantage. By the extension of its
output of watches (assumed to be produced at decreasing
cost) A gets a lower cost per unit, while its withdrawal
from wheat production (assumed to be produced
at increasing cost) would give it a lower cost per unit of
wheat also, if it should still produce any as an alternative
to watches. B, on the other hand, by the extension
of its output of wheat raises its cost per unit of this product,
while its withdrawal from watch production gives
it a higher cost per unit for watches also if it should still
produce any as an alternative to wheat. Let the conditions
be represented thus:
CASE II (a)
In Country A
10 days' labor produces 45 units wheat
10 days' labor produces 45 watches.
In Country B
10 days' labor produces 35 units wheat.
10 days' labor produces 20 watches.
B still has a comparative advantage in the production
of wheat and may be assumed to have had it throughout
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
206 QUARTERLY JOURNAL OF ECONOMICS
the period of transition from Case I. Specialization in
wheat will be advantageous for B as things are at the
moment, and will have been advantageous, it may be
assumed, at any given moment throughout the period of
transition. Trade is now possible on any terms between
35 wheat = 20 @ 35 watches, or, what is the same thing,
40 wheat = 224 @ 40 watches. Let us assume for the
moment no change from Case I in the terms of exchange
that is that 40 units of wheat exchange for 35 watches
(tho in view of our premises and the changed conditions
this is a ratio unduly favorable to B). Then country A
for 20 days' labor devoted to the production of watches
(= 90 watches) can, by exchanging 35 watches for 40
wheat, obtain 40 wheat plus 55 (90 - 35) watches; while
country B, for 20 days' labor devoted to the production
of wheat ( = 70 wheat) can by this same exchange obtain
30 (70 - 40) wheat plus 35 watches. Under these conditions
both countries will again benefit by the trade,
since without it A for 20 days' labor equally divided
between wheat and watches would get but 45 wheat and
45 watches as compared with 40 wheat and 55 watches
obtained through trade; while without trade B would
get 35 wheat plus 20 watches as compared with 30 wheat
plus 35 watches obtained with trade. The result of the
trading is to give A, ten, and B, fifteen more watches
than they would have without trade, while both countries
lose 5 units of wheat. But 5 units of wheat are
worth not more than 5 watches in either country, so the
result for both is a net gain.
Compare, however, the situation of the two countries
under Case I, on the assumption that no trade had ever
been opened up, with that under Case II with trade
free. It is as follows:
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 207
IN COUNTRY A
Case I
20 days' labor without trade yields 40 wheat plus 40 watches.
Case II
20 days' labor with trade yields 40 wheat plus 55 watches
Here under Case II with trade A makes a clear gain of
15 watches as compared with its position under Case I
without trade.
IN COUNTRY B
Case I
20 days' labor without trade yields 40 wheat plus 30 watches.
Case II
20 days' labor with trade yields 30 wheat plus 35 watches.
Here under Case II with trade free B gets 10 less
wheat and 5 more watches than under Case I with no
trade. This is a net loss, since 10 units of wheat are
worth more than 5 watches in either country under
either set of conditions. The specialization along lines of
comparative advantage has been disadvantageous to
B. At any given moment of the transition from Case I
to Case II it will pay B to specialize in wheat, but the
final result of the specialization is to bring about a situation
in which the citizens of B get less reward for their
efforts than if they had never carried on international
trade at all.
To make this conclusion invulnerable, it is necessary
to consider the nature of the figures given in the illustrations.
If these figures be figures representing marginal
costs, the conclusion that country B will lose by
specialization is not inevitable. For the extension of the
production of wheat will mean not that the cost of every
unit of the product rises but merely of that part of the
whole product which was not grown previous to the increase
in output. All rents will be raised by the lowering
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
208 QUARTERLY JOURNAL OF ECONOMICS
of the margin of cultivation, and while this means a
lower net return to all growers who are not landlords,
the change is merely one in distribution and will not
affect the national income arising from the lands which
had been cultivated from the first.
The average unit cost, however, will increase, since
the new output is produced at higher cost than any
part of the old; and whether country B will sustain a net
loss or not depends (1) upon the rapidity with which
costs rise upon an extension of output and (2) upon the
magnitude of that extension. If then, instead of regarding
the figures as figures of marginal unit cost, they be
understood to mean average unit cost or (perhaps) the
cost of representative producers, the conclusion that
country B under the conditions assumed must lose by
free trade is inevitable, the only effect of the differing
costs of growing wheat in B being to narrow the limits
within which trade could be advantageous. Thus, if the
following conditions are assumed for country B,
10 days' labor will produce 40 units of wheat on some farms.
10 " " a " 35 " " " as an average.
10 a " a " 32 " a " on the margin
10 " " " " 20 watches,
the limits of advantageous trade with country A under
Case II would be 32 wheat = 20 @ 32 watches or 40
wheat = 25 @ 40 watches, as compared with the 40
wheat = 226 @ 40 watches, which are the limits when
35 units of wheat are taken as the marginal product (see
page 206). But this narrowing of limits is of no consequence,
since the actual terms set by the equation of international
demand are assumed to be 40 wheat = 35
watches.
The question of varying costs does not arise with regard
to country A's product, watches, since an extension
of output will be reflected here in a lower unit cost
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 209
for the whole product and not for the new
merely, and all producers must have approximately
equal costs or be put out of business by their competitors.
For a given fluctuation in the cost of their respective
marginal units, the total cost of the production of
watches will fall to a greater degree than the total cost
of the production of wheat will rise, since in the former
case the decreased cost applies to all the units produced,
while in the latter the increased cost applies only to the
new increments and only to the full extent to the last of
them. There is, of course, no reason to expect that the
marginal costs in the two cases will fluctuate to the same
degree. Whether the total cost and average unit cost of
the commodity produced under conditions of increasing
cost will mount more or less rapidly than the total and
unit cost of the commodity produced under conditions
of decreasing cost falls, depends upon the respective
nature of those conditions. An extension of the output
of wheat may mean a rapidly or slowly rising marginal
and average unit cost according to the conditions in
country B, and an extension of the output of watches
may mean a rapidly or slowly declining unit cost according
to the changes that take place in A consequent
upon such increased output. In what might be considered
the more normal case, the unit cost of the commodity
produced at increasing cost will rise with less
rapidity than the unit cost of the commodity produced
at decreasing cost will fall.
But even if we adopt the case most favorable to B
and proceed according to the method of limits to suppose
that the increased output of wheat is produced at a
unit cost no greater than prevailed before the production
was expanded, it may well happen that B will lose
by free trade. To make this evident let us suppose the
following conditions:
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
210 QUARTERLY JOURNAL OF ECONOMICS
CASE II (b) (marginal units)
In Country A
10 days' labor produces 45 units wheat.
10 days' labor produces 45 watches.
In Country B
10 days' labor produces 40 units wheat.
10 days' labor produces 20 watches.
Trade may be carried on advantageously on any
terms between 40 wheat = 20 @ 40 watches. Assume
that the play of international demand brings equilibrium
on the basis of 40 wheat for 28 watches. Then
country B for 20 days' labor devoted to the production
of wheat will by the exchange of 40 wheat for 28 watches
obtain 40 (80 - 40) wheat plus 28 watches. But if trade
had never been opened up and B had never specialized
on the commodity in which it had a comparative advantage,
viz., wheat, it could have had 40 wheat plus
30 watches.
It might be contended that the declining cost of production
of watches and the advancing cost of production
of wheat would be reflected in a shifting of the terms of
exchange not against B, as the illustration above assumes,
but in B's favor, and that this shifting of the
terms of exchange would cause both countries to gain
instead of one gaining and the other losing. But there
is no reason for supposing that this would happen. It
would happen only if there were perfect mobility of
labor and capital between the two countries. But the
lack of any such mobility is a fundamental of the theory
of international trade and is, indeed, the one thing that
calls for a theory of international trade distinct from
the theory of domestic trade. The play of reciprocal demand
which decides the terms of exchange of the commodities
entering into international trade may operate
to keep prices up even when the effort cost of production
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 211
falls, or to depress them when the effort cost of production
rises. If, as was assumed at the outset, the sales
of watches tend to increase while those of wheat tend
to remain constant, the terms of exchange will move
not in favor of B but, as is here supposed, against that
country.
Turn now to Case III. The hypothesis is simply
that protection has been given to watches in country B
while the conditions of production of Case I prevailed.
If the protection is fully effective, B will do no worse
than hold the productiveness of Case I, which, as has
already been pointed out, leaves it in a better position
at the later period represented by Case II and Case III
than if trade had been left free. Protection will still be
necessary if B's watch manufacture is to survive, and
there is no reason to suppose that it will ever be unnecessary.
Nevertheless B is economically benefited by protection
and may do well to keep it indefinitely. This
possibility of advantageous permanent protection was
not seen by Hamilton and List, no doubt because it
challenges the assumption of constant cost, an assumption
which in their time was probably more consonant
with facts than at present. Later economists, such as
Professors Sidgwick, Edgeworth, and Carver, have attacked
this assumption in one way and another, but so
far as the writer is aware have not brought out the foregoing
implications of increasing and decreasing cost of
production.
A comparison of the figures in the cases assumed so
far will show that, while country B loses as a result of
specialization along lines of comparative advantage, the
total production of the two countries taken together
increases. With the present intensity of national feeling
this general advantage will make little difference to B
(if we assume that it is a country in esse), since B would
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
212 QUARTERLY JOURNAL OF ECONOMICS
be unlikely to adopt a policy of free trade hurtful to its
own interests merely because it could be shown to give
more to other countries than B itself loses. But even
if we could assume that B would be so altruistic as to do
this, it might not have the opportunity. For to specialize
along lines of comparative advantage may result
in a next loss all round. To illustrate this point take
again the conditions of Case I, with the figures representing
average rather than marginal product.
In Country A
10 days' labor produces 40 units wheat (on the average).
10 days' labor produces 40 watches.
In Country B
10 days' labor produces 40 units wheat (on the average).
10 days' labor produces 30 watches.
Assume no trade. Then the total product of 20 days'
labor in both countries will be: 80 (40 plus 40) wheat
plus 70 (40 plus 30) watches. Now suppose conditions
after trade has been opened up and specialization along
lines of comparative advantage has taken place to be
(solely as a result of increased production in the lines of
respective comparative advantage) as follows:
In Country A
10 days' labor produces 45 units wheat (on the average).
10 days' labor produces 42 watches.
In Country B
10 days' labor produces 30 units wheat (on the average).
10 days' labor produces 25 watches.
Then the total production of 20 days' labor in both
countries with trade will be: 60 (2 X 30) wheat plus
84 (2 X 42) watches. Comparing this with the production
of a similar amount of labor without trade under
Case I, 80 wheat plus 70 watches, it appears that under
Case II with the specialization which free trade would
tend to bring about there would be obtained 20 less
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 213
wheat and 14 more watches than under Case I without
trade. But this is a net loss, since 20 wheat are worth
more than 14 watches in either country under either set
of conditions.
These figures illustrate the loss that List may have
had in mind when he insisted upon the advantage of the
development of productive forces. But neither List nor
his followers have shown that this was anything more
than an emotion, tho Patten has brought applicable
instances of the sort in support of his ingenuities in
defense of protection.
The upshot of the considerations advanced above is
that comparative advantage is by no means an infallible
guide to the maximum return in the long run to the
productive effort of a nation. A rational, forward-looking
policy might deliberately reject development along
lines of comparative advantage. Nor is absolute advantage
at any given moment a reliable test, Professor
Patten to the contrary notwithstanding. Taking comparative
advantage as a basis, modifications of that
principle may be necessary according to the answers to
these two questions (1) will the development of production
along lines of comparative advantage bring about
through specialization an absolutely less productive
regime? (2) what is the trend of world demand and
supply for the actual or potential products of the given
nation? The first of these considerations affects the
volume of a nation's product, the second, the terms on
which it exchanges its products for the products of other
countries. It may well be disadvantageous for a nation
to concentrate in production of commodities of increasing
cost despite a comparative advantage in those lines;
it will the more probably be disadvantageous to do so if
the world demand for goods produced at decreasing
cost is growing in volume more rapidly than that for
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
214 QUARTERLY JOURNAL OF ECONOMICS
goods produced at increasing cost, while at the same
time competition in the supply of the former grows
relatively less intense as compared with competition in
the supply of the latter. For in this case the operation
of the law of reciprocal demand will throw the terms of
the exchange of commodities more and more to the disadvantage
of the country producing the goods of increasing
cost.
Since goods exchange in international trade on the
principle of reciprocal demand, the tendency actually is
for competition to restrict the gains obtained in the production
and exchange of increasing-cost goods, and to
enlarge the gains in the production and exchange of decreasing-cost
goods. For in the production of increasing-cost
goods competition develops apace, as every
extension of production through specialization brings
in (as a result of rising costs) hitherto impossible outside
competition; while in the production of decreasingcost
goods, competition becomes more and more ineffective
as costs in the specializing country fall. The
quickening of competition in the marketing of increasing-cost
goods and its decline in that of commodities
produced at decreasing cost operates to lessen the spread
between the effort cost of the former class of goods and
the commodity returns received for them, while it tends
to augment that spread in the case of the latter class.
Conditions of supply being irrelevant in the determination
of the terms of exchange, if we assume for the moment
that the relation between the demand for the
commodities of increasing and those of decreasing cost
remains constant, then the country producing commodities
at decreasing cost will secure a growing differential
between cost and return, while the country producing
commodities at increasing cost will find its gain declining.
Real wages will rise in the former country as a
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 215
result of the relatively effective application of its labor,
for it is a commonplace of the theory of international
trade that the standard for wages in general is set in the
exporting industries,4 and depends on their effectiveness;
they will fall in the latter country for the opposite
reason. This movement of wages will prevent any tendency
toward a readjustment of the terms of exchange of
the two types of commodities in favor of the increasingcost
goods; it is in fact but a reaffirmation of the principle
that cost of production is not effective in setting
those terms.
The principle just laid down may go far to explain
why regions of slender natural resources devoted to
manufactures often surpass in prosperity regions of
much greater natural resources where extractive industry
prevails, tho no great difference exists in the native
ability of their respective populations. One may instance
the almost constant complaint of the West in
the United States during the period when its effort was
almost solely devoted to extractive industry, over its
"exploitation" by the East, particularly New England.
This is a case strictly analogous to international exchange
under free-trade conditions, since mobility of the
factors of production was considerably restricted.
So far the situation has been observed from the supply
side of production only, the relation between the demand
for the two types of commodities being assumed
to be constant. If now demand for the commodities
produced at decreasing cost grows in volume relatively
to that of products produced at increasing cost, the
terms of exchange will move still more in favor of
the producers of decreasing-cost commodities, and it
should be borne in mind that the only check on this.
4 Strictly, it is the standard for money wages that is set in the exporting industries,
but increased money wages are here due to increased effectiveness in production and to
more favorable terms of exchange, and so will mean increased real wages as well
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
216 QUARTERLY JOURNAL OF ECONOMICS
movement is that set by the possibility of the home
production of the imported decreasing-cost commodities
by the country which is specializing in the commodities
produced at increasing cost - a possibility which such
a country grows ever less competent to realize, since its
abandonment of the production of the decreasing-cost
commodities results in higher unit cost for such an
amount of these goods as it may continue to produce.
The movement of the terms of exchange in favor of the
producers of the increasing-cost commodities is, on the
contrary, quickly checked by the increasing competence
of the country which is specializing in goods of decreasing
cost to meet the competition of the other in the
production of the increasing-cost goods, since its abandonment
of the production of increasing-cost commodities
results in lower unit costs for such an amount of
the~e goods as it may continue to produce.
Departing now from deduction let us look into actual
conditions. In the main the commodities produced at
lower unit cost on an increase of output are manufactured
goods, while those produced at a higher unit
cost are the commodities of extractive industry. England
was the first country to obtain great advantages of
decreasing cost. The extension of free trade tended to
enlarge those advantages for England and to render her
prosperous beyond any other country in which conditions
were at all comparable. For England's increasing
concentration upon manufactured goods was constantly
tending to lower her effort cost of production, while the
increasing specialization which free trade operates to
bring about was tending to raise the effort cost of production
of those commodities (largely extractive) which
England took in exchange for her manufactures. The
play of reciprocal demand, which alone sets the terms
of exchange of commodities in international trade, was
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 217
probably also operating to England's advantage, since
the secular trend of demand must have been increasing
for manufactured goods relatively to the commodities of
extractive industry, as is shown by the sheer growth of
manufacturing industry since the Industrial Revolu-
tion.
The spurt in England's prosperity could not be approached
in old countries devoted to extractive industry,
and could be matched only by new countries whose
wealth of land made extractive industry highly profitable.
The situation meant a quasi-monopoly position
for England in her production while the countries producing
extractive. commodities were being subjected
to increasingly severe competition. In this situation a
tariff to build up industries of decreasing unit cost in
these countries might well be advantageous to them,
tho their comparative advantages were permanently to
remain in industries of increasing unit cost. In these
premises a scientific commercial policy for such countries
would, on purely economic grounds, modify the
principle of comparative advantage in anticipation of
the trend of world demand and of the conditions of
supply, and this is approximately what the bulk of tariff
policies, however blindly adopted, has actually done.
They were adopted, in part at least perhaps, because
empirically obtained opinions pointed to their efficacy
5. This statement may perhaps be questioned. Professor Taussig in his Free Trade,
the Tariff and Reciprocity (page 87) seems to hold the contrary view, when he attributes
the relative prosperity of the United States to the fact that our exports have in the
past been agricultural products largely, for which the demand has been, as he says,
stable and steadily increasing in volume. The conflict is, I think, superficial only. The
volume of the total demand for manufactured goods must have increased at least as
rapidly. Elasticity of demand for manufactured goods is undoubtedly greater and so
there would be less stability in the demand; but this elasticity is a prime factor in promoting
a secular increase in the volume of such goods taken, and it is the secular trend
which is here important. Increased demand for manufactured goods of course involves
an increased demand for extractive products but it would seem that in the last century
and a half the percentage of worked-up goods must constantly have been an increasing
proportion of the whole. The very appearance of a great variety of new inventions is
evidence of this.
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
218 QUARTERLY JOURNAL OF ECONOMICS
to advance prosperity. The free-trade movement
itiated in England in 1846 made headway outside of
England for a few years only. After that it was repudiated
by most old countries,6 and even by the new as the
first exceptionally favorable conditions for the production
of extractive commodities began to depart.7 We
have seen that the opinion commonly held that manufactures
mean prosperity is in some cases not without
scientific foundation, and self-sufficiency may conceivably
have been a justifiable policy, even on economic
grounds, for the countries which adopted it.
So much for the first of the "general causes" of the
epidemic of protection. There is another, which I shall
now indicate. The accepted exposition of free-trade
principles speaks in terms of "units of productive
power," usually commuted, as in this article, to so many
days' labor. But it fails to analyze the composition of
those units. They are not identical in different countries
but may be any combination whatever of land,
labor, and capital. Suppose in country A one day's labor
is equivalent in value to the use of 10 units of capital or
3 units of land, while in country B one day's labor equals
in value the use of 5 units of capital or 8 units of land.
The unit of productive power in the two countries will
be very different things. The comparative advantage
of A will lie in the production of commodities which
take much capital and little land, while in B it will lie
in the production of commodities which take much land
and little capital. But this situation may be completely
changed in B by the growth of capital relatively to the
other factors. It may well be that, given the capital, the
comparative advantage of country B would be in the
6. France 1871, Germany 1879, Italy 1877, Austria 1878.
7. United States 1861, Argentina 1878 onward, Canada 1879 onward, Australia 1902,
Chile 1916.
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 219
production of the very things in which country A now
has that comparative advantage. B's genius and resources
may lie in that direction, and remain dormant
because the scarcity of capital renders it uneconomic to
develop them. (China is perhaps an instance in point,
tho of course many other factors enter.) In such a case
a country's genius and prosperity may be thwarted by a
lack of capital and the all-important question is how to
obtain capital most quickly.
There are, of course, the two alternatives of borrowing
and saving. As to borrowing. The lending of capital
and the price of its use are determined to a very considerable
degree by propinquity, physical, and spiritual.
If this were not so, the rate of interest would be the
same the world over for investments of similar character.
But it is not the same the world over (tho its
fluctuations probably grow less), and its variations run
in large measure according to the remoteness in space
or acquaintance or good-will from the source of supply.
Here again it is this factor of immobility that differentiates
international from domestic trade and has given
rise to a separate theory concerning it. Create a home
supply of capital and you change conditions funda-
mentally.
Now as to saving.8 The confirmed free trader will contend
that the direction of industry along lines of immediate
comparative advantage will yield the greatest
return and so promote most successfully the desired
accumulation of capital and the quickest shift to the
industries which will be the more productive in the long
run. But this argument assumes that, because under
free trade more capital could be saved, more will be
8 The argument developed in this and succeeding paragraphs has been in large
measure anticipated by A S Johnson in an article in the Political Science Quarterly,
vol xxin, No 2, p 220 The present article had been completed when this fact was
brought to the writer's attention
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
220 QUARTERLY JOURNAL OF ECONOMICS
saved. The assumption is a dubious one for two reasons,
both of which turn on the fact that not only the
amount but the distribution of wealth is important in
the accumulation of capital. Protection in the circumstances
here assumed will affect the distribution of
wealth in two significant ways: (a) it will tend to concentrate
it; (b) it will concentrate it in the hands of
people who are well-nigh inevitable savers. For let it
be remembered that the assumption is that the long-run
interests of the country require a shifting to the industries
which are relatively large users of capital, that is,
to the manufacturing industries in the main. Now, tho
no satisfactory statistics exist, so far as the writer is
aware, to show clearly the effect of manufacturing industry
on the distribution of wealth, we know that
England, the typical manufacturing country, shows the
greatest concentration of wealth 9 and agricultural
Australasia, perhaps, the least. In our own country the
growth of huge fortunes seems to have been tied up in
large measure with the growth of manufacturing, and
the concentration of wealth seems to be increasing with
our gradual transition toward an Industriestaat.1 I
shall presently adduce deductive reasons for this phenomenon,
but, taking it for the moment to be a fact,
then effective protection to manufactures would act as
a force striking the plain of wealth distribution, and
crumpling it up into hills and valleys. There is no
doubt that the great bulk of saving is done by wealthy
people 2 and that the unequal distribution of wealth is a
prime factor in the accumulation of capital. (It is interesting
to note in this connection that the export of
capital comes from countries in which manufactures are
9. King, Wealth and Income of the People of the United States, p 97
1. Ibid., p 74.
2. Taussig, Principles of Economics, vol. ii, p 42.
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 221
highly developed and its import to those in which it is
not.) It seems to be true that where income is evenly
distributed the standard of living rises to the point
where a very high proportion of the total national income
is consumed immediately, to the ultimate detriment
of the whole country. As Professor Friday says,
quoting Keynes, "the immense accumulations of capital,
which to the great benefit of mankind were built up
during the half century before the war (1914), could
never have come about in a society where wealth was
divided equitably" 3 (that is, more equally than at
present).
The chief reason for the unequal distribution of
wealth which attends a manufacturing regime, as well
as its concentration in the hands of men who are almost
forced to save, is again the fact that an extension of output
in manufacture is likely to be accompanied by a
falling unit cost. As Professor Friday points out in the
work already cited enormous additions to capital are
made in the reinvestment of earnings in plant extensions.4
Whether by reinvestment of profits or otherwise,
this saving through the force of competition is
practically compulsory in an industry subject to increasing
returns, for the alternative to expansion is
bankruptcy. On the other hand in the extractive industries,
where an extension of output is likely to be
accompanied by a rising unit cost, this stimulus to expansion
and increase of equipment is lacking. The form
that industry takes is vital in the accumulation of
capital. The processes of extractive industry are relatively
direct, there is a comparatively slight need for
great investments in fixed capital. The processes of
manufacture on the other hand are indirect, and the
3 Friday, Profits, Wages, and Prices (1920), p. 66.
4 Ibid, chap iv
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
222 QUARTERLY JOURNAL OF ECONOMICS
extension of manufactures, even at the cost of extractive
industry, is thus itself inevitably an accumulation
of capital. The potential savers, the rich men, in a state
given over to extractive industries, are landlords. But
landlords are not good savers; on the contrary, they are
in the main a "spendthrift class." The reason for this
is that there is with them but slight stimulus to saving
and slight opportunity to invest profitably any savings
that might be made. There is a significant passage in
The Education of Henry Adams: "The American wasted
money more recklessly than anyone ever did before,
and except for the railway system the enormous wealth
taken out of the ground since 1840 had disappeared (in
1892).... West of the Alleghenies the whole country
might have been swept clean and could have been replaced
in better form in one or two years." 6 From these
words it would seem that capital accumulation had been
slight in the country as a whole while it was predominantly
agricultural, and that the only exception was in
the manufacturing East.
A considerable part of a manufacturing community is
likely to be, engaged in the production of producers'
goods, and the self-interest of firms so engaged promotes
the acquisition of capital. Every sale that they make
means the formation of capital, every extension of credit
that they receive from the banks based upon their anticipated
output promotes capital accumulation - a
condition approaching automatic saving.6 There is
nothing like the same facility of saving in a non-manufacturing
state where effort is devoted to the production
of consumers' goods and saving can take place only
through personal abstinence. In this latter case there
5. Education of Henry Adams, p. 328.
6. See Wolfe, "Savers' Surplus and the Interest Rate," Quarterly Journal of Economics,
vol. xxxv, No. 1, p. 1 et seq.
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 223
is always competition between saving and
the formation of capital involves a decision to save on
the part of those who have the alternative of spending.
There is no such alternative when the production of
capital goods is a prevalent type of activity. The industrial
and business processes themselves declare for
saving, and the choice is made for capital accumulation
by the nature of the industrial technique. So far as the
industries indicated are concerned, it is either capital
accumulation or no production at all; there is no alternative
of spending open as a choice to anyone. Of course
the demand for the capital goods must be present, but
no one conversant with modern business methods imagines
that demand is not stimulated by producers, as well
as producers by demand. It is perhaps not too much
to say that much of the capital found in non-industrial
regions, such as farm machinery, has been created as
a result of the energy and enterprise of its producers,
rather than on the initiative of those who make use
of it.
The promotion of manufactures by protection almost
inevitably advances capital formation in yet another
way. The protection which must be given to encourage
men to essay untried and doubtful projects must
ordinarily promise more than normal returns, and these
are as a rule realized by the more effective producers.
In industries of decreasing cost these producers tend to
extend their operations, all the while making great gains,
which they are almost certain to put back into so profitable
a business. They become rich men and so save
easily, they reinvest because it is a condition of success.
The operation of a protective tariff would be to reduce
the gains both of employers and wage earners in the
naturally more advantageous industries in the conditions
of the moment. It is probable that, if these are
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
224 QUARTERLY JOURNAL OF ECONOMICS
extractive industries, this will mean diminished consumption
rather than diminished capital accumulation;
at least more capital is likely to be saved by the
manufacturers than fails to be saved by the others. This
will improve the situation all round in the long run.
The protection to infant industries argument then
may perhaps be extended to an argument for the
protection of infant capitalism or finance - of course
a much longer process. The intimate connection of
modern industry with finance may make finance and
the accessibility of capital on favorable terms the factor
determining comparative advantage. The growth
of finance and manufacturing industry are interdependent.
The latest stage of economic development is
not List's final agricultural-manufacturing-commercial
stage, but an agricultural-manufacturing-commercialfinancial
stage. Perhaps no state can with advantage
permanently depend on another for its supply of capital,
not only because it may be exploited thereby, but be7
cause owing to difficulties of mobility the supply is
almost sure to be inadequate to the development that
would be economic were capital more plentiful. There
may be a considerable range of industries which would
be competent industries if capital were more abundant
but are in the present subordinate to industries which
use relatively little capital. The immediate effect of
protection to manufactures may be to increase the
demand for capital and so raise its price, making these
industries still less able to compete without protection
than they had been before, and this incapacity may
continue for a long period as new manufacturing industries
are thus encouraged. The gradual gain which may
be expected in the supply of capital in its race with demand
may serve, however, to realize the real genius of a
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 225
people which had hitherto been unable to express itself
for lack of the necessary means. In this case the protective
measures would have justified themselves even
tho they are necessary for a much longer time than
would ordinarily seem sufficient to cover the infant
stage.
It is difficult to assess the weight of the considerations
advanced in this paper. They may be purely
academic. It is certain that they have not been in the
minds of legislators of protection. Have they been of
any importance in the wide diffusion of the idea that
protection makes or preserves prosperity? Has the
belief that actual experience has shown protection to be
good any roots in them? France, Germany, and the
United States became powerful and prosperous, accumulators
and lenders of wealth, along with industrialism
and persistent protection - protection against
England mainly, England which had the start, the
comparative advantage in industries which show a
strong tendency toward a lower unit cost, the rapidly
accumulating capital. There can be no question that
much of this development would have taken place in any
event; but on the other hand protection has undoubtedly
played its part, perhaps a truly economic part.
This paper has been kind to the protectionists. It
may provide support for the comforting belief that t
United States in its long-continued protective policy has
builded better than it knew. But even assuming that
all the economic advantages of protection which it indicates
as possibilities have actually been realized,
those possibilities are now a thing of the past. Without
going into the far-reaching considerations of the most
advantageous distribution of wealth from the economic
point of view, or of an indefinite postponement of immediate
consumption for the sake of greater productive
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
226 QUARTERLY JOURNAL OF ECONOMICS
powers and a larger consumption in a future
gin fades forever as we move, it is clear that now, when
our export of manufactured goods has reached predominant
proportions, a pervasive system of protection is no
longer effective in these industries except to injure them
by curtailing exports pars passu with the restriction of
imports and by increasing cost of production in the manner
made plain by the classical economists. Pervasive
protection cannot now advance the interest of decreasing-cost
industries' as against those of increasing-cost;
indeed the loss of capital in Europe consequent upon the
war and our own rapid accumulation of it in the last few
decades may have caused a very considerable reversal of
the lines of comparative advantage which prevailed in
the last century, so that our comparative advantage
may now tend toward manufactures. In so far as there
is any "need" for the protection now advocated for
agriculture in this country this must be the explanation.
Our present situation as a lending country shows that
the stage of infant capitalism has been passed, that so
far as the supply of capital is an important element in
the production of any given commodity we are in a
most favorable position. The interests of finance and
manufacturing industry are still tied together, but they
lie in the direction of freer trade. The extension of
manufacturing equipment to meet the war demand has
created a situation in which the productive capacity
of our manufactures exceeds the proportion of those
manufactures which our home population cares to take,
and we cannot shift to the lines in which demand
is proportionately great without the loss of much
capital already irrevocably fixed. A foreign outlet as
large as possible is highly desirable. The tendency of
all decreasing-cost industries is of course to exceed the
demands of the home market, owing to the advantages
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms
SOME ASPECTS OF PROTECTION 227
derived from increasing output, and the huge scale attained
by our manufactures has operated to throw the
comparative advantage into that field of production.
We are in fact in the position to the world taken as a
whole, in which England found herself about a century
ago, and just as free trade was adopted there to her own
undoubted advantage, we too would find it working in
our favor now. As a result of the operation of the forces
I have indicated in this paper the high protection which
has been an historical fact in this country since 1860
has perhaps not been uneconomic in the long run, but
the very factors which lend a color of validity to this
opinion are now working against protection. If they
were of sufficient importance in the past to counteract
the advantages arising from free trade by so much do
they in the present reinforce the arguments for a policy
of unrestricted commerce,.
FRANK D. GRAHAM.
PRINCETON UNIVERSITY,
This content downloaded from 147.251.185.127 on Wed, 21 Jun 2017 09:41:05 UTC
All use subject to http://about.jstor.org/terms