Trade Imbalance and the Leontief Paradox

Author(s): Bee Yan Aw

Source: Weltwirtschaftliches Archiv, Bd. 119, H. 4 (1983), pp. 734-738
Published by: Springer
Stable URL: https://www.jstor.org/stable/40439179
Accessed: 07-11-2023 16:54 +00:00

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
https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to

Weltwirtschaftliches Archiv

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms
 Bemerkungen
Trade Imbalance and the Leontief Paradox

By

BeeYanAw

Introduction

the Heckscher-Ohlin-Vanek (HOV) model of trade, Learner

[1980] demonstrates that the Leontief paradox arose by drawing
inferences about a country's relative factor endowment based only
on an examination of the ratio of the capital per man embodied in imports
to that embodied in exports. Learner shows that it is theoretically in-
correct to focus on this ratio if there are more than two commodities and
if the country is a net exporter or importer of both labor and capital
services. In this context, factor abundance can only be inferred through a
comparison of the factor inputs in net exports and in consumption. By
assuming the existence of balanced trade and a third factor, land, he
shows that it is possible for the exports of a capital-rich country like the
U.S. to be more labor-intensive than its imports.
The objective of this paper is to extend Learner's results and show that
it is not necessary to include both a third factor and require trade to be
balanced in order to explain the net export of labor-intensive goods by a
capital-rich country. A trade surplus may in some cases be a sufficient
explanation.

Factor Content of Trade and Factor Abundance

The HOV theorem [Vanek, 1968] states that given the standard as-
sumptions of international trade models1 there exists a set of positive
scalars oq, i = 1, . . . , I such that the vector of net exports of country i, Tit
the vector of factor endowments of country i, Eif and the η χ η matrix

Remark: Helpful comments were provided by J. David Richardson and an anonymous

referee.

1 These assumptions are: 1) identical and linearly homogeneous production functions

across countries; 2) identical and homothetic preferences among countries; 3) non-revers-
ability of factor intensities; 4) perfectly competitive markets, free trade, no transport costs,
and complete international immobility of productive factors.

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms
 Bee Yan Aw 735
Trade Imbalance

of total factor requirements A, bear the following relationship to each

other:

ATi=Ei-Ewai (1)

where Ew is the world's en

trade in factor services
factors. Using Learner's no
country A and the world
spectively. Let (KT, Lx) b
of country A, then by def
(Kx, KM) is the capital con
labor content of exports
the two-factor case to

KT = KA-aAKw Lt^La-oiaL* (2)

Equation (2) can be rewritten as

KW/KA = (KA - KT)/aA KA LW/LA = (LA -

Definition: Country A is said to be abundant i

labor if and only if KA/KW > LA/Lw .

Definition: A capital-abundant country is in "a

balance" if the outflow of capital embodied in net ex
by an inflow of labor embodied in net exports. T
This condition is also consistent with small surplu
same country: small surpluses (deficits) would mea
capital was higher (lower) than under approximate
inflow of labor was lower (higher). A capital-abun
run a "sufficiently large" trade surplus (or defici
exporter (importer) of both labor and capital ser
(or KT < 0, Lx < 0).
Several corollaries follow from the above model. C
closely related to Learner's treatment2, however, the
trade balance or imbalance and the implication this h
method of computing trade-revealed factor abundan
differ significantly from Learner's analysis in th
itself is shown to be a possible explanation of the
intensive (capital-intensive) goods by a capital-rich

1 The proof of this result is in Learner [1980].

2 In particular, to Corollaries 1 to 3 in his article.

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms
73^ Bemerkungen

Corollary ι: If trade is balanced so that KT

with the positive net exports is revea
abundant factor, KA/KW > LA/LW .

Proof: Equation (2) shows that if KT > 0, L

LA/Lw < «a, that is KA/KW > LA/LW.
This corollary states that if the net export
net export of labor services are opposite in sig
and small trade surplus/deficits) then the fac
is revealed to be the relatively abundant fac

Corollary 2: Given "approximately" bala

then KA/KW > LA/LW if Kx/Lx > KM/

Proof: Given KT > 0, Lj < 0, then by Corolla

0 < KT = Kx - KM implies KX/KM > 1 a
plies LX/LM < 1. Thus KX/KM > LX/LM or
Thus, when trade is balanced or when onl
occur (in the sense that net export capital a
posite in sign) then for the capital abundant c
embodied in exports exceeds the capital per

Corollary 3: When country A's trade surplu

that KT > 0 and Lr > 0 then KA/KW
> Ka/La.i
Proof: From equation (3) it follows that Kw
< (LA - Lt)/La or KA/KW > La/L* if -
> 0, KT > 0 the latter inequality becomes K
Thus, a country which is an exporter of bot
is revealed by trade to be relatively capita
capital intensive than its national endowme

Corollary 4: When country A's trade deficit is

KT < OandLr < 0 then KA/KW > .LJL

Proof: From equation (3) KA/KW > LA/LW

Lp < 0, KT < 0, the latter inequality becom

1 In this case, this inequality is equivalent to Κχ/L

Learner used to empirically infer that the U.S. is revea
Beginning with equation (3) and Kc = Ka - Κχ and L
if and only if Kc/KA < Lc/LA or Kc (Lc + Lr) < Lc

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms
 Bee Yan Aw 737
Trade Imbalance

Thus, if a country is an importer of both capital and labor services,

then it is revealed by trade to be relatively capital abundant if trade is
less capital intensive than its endowment1.
Corollaries 1 through 4 show that the appropriate indicator of a
country's relative factor abundance depends on whether or not its trade
is balanced. When trade is "approximately" balanced, comparing the
capital per man embodied in imports with that in exports (KM/LM)/
(Kx/Lx) will reveal the underlying relative factor abundance. However,
when there is a "sufficiently large" trade surplus or deficit, the theoretical-
ly correct comparison must be between capital per man embodied in net
exports and the nationally endowed capital per man.
Corollaries 5 and 6 will show that when there is a large trade surplus
or deficit, the measure (KM/LM)/(KX/LX) cannot be used to make inferences
about the country's relative factor abundance.

Corollary 5; When there is a "sufficiently large" trade surplus so that

KT > 0, Lx > 0, Kx/Lx > KM/LM does not imply KA/KW
>LA/LW.
Proof: 1. From equation (2) and Corollary 3, it is clear that KT > 0 and
Lj > 0 is compatible with either KA/KW > LA/LW (when KT/Lx > KA/LA)
or KA/KW < LA/LW (when KT/Lx < KA/LA).

2. Note further that Kx/Lx > KM/LM is consistent with KT > 0,

Lj > 0. This can be illustrated numerically. Let Kx = 100, KM = 20,
Lx = 10, LM = 4, KT = Kx - KM = 80 > 0, Lx = Lx - LM = 6 > 0
and Kx/Lx > KM/LM (i.e., 100/10 > 20/4). Thus, Kx/Lx > KM/LM may be
consistent with either KA/KW > ILJL^ or KA/KW < LA/LW .

In other words, when there is a large trade surplus one might find
a labor-rich country's exports to be more capital intensive than its imports.
Conversely, with a large trade surplus one might also find the exports of
a capital-rich country to be more labor intensive than its imports. Thus,
the Leontief factor-content ratio (defined as (KM/LM)/(KX/LX)) may be a
misleading indicator of a country's relative factor endowment in the
presence of "sufficiently large" trade surpluses.

1 An alternative test whenever Κχ and Lx are of the same sign (whether both are positive
or negative) is to compare Κχ/ΚΑ and Lx/LA . If KA/KW > Là/I^ , then Κχ/ΚΑ > Lx/LA .
This can be shown easily. From equation (2) we can write
1 - Κχ/ΚΑ = oca KW/KA and 1 - LT/LA = oca Lw/La
KW/KA < Lw/La if Kt/Ka > LT/LA . But KW/KA < Lw/La implies KA/KW > LA/Lw.

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms
 738 Bemerkungen
Bee Yan Aw, Trade Imbalance

Corollary 6: When there is a "sufficiently large" trade deficit so that

KT < 0, Lr < 0, Kx/Lx > K^/Lm does not imply KA/KW
>La/Lw·
Proof: From equation (2) and Corollary 4, it is possible to show:
1. KT < 0, LT < 0 is compatible with either KA/KW > LA/LW (when
Kt/Lt < KA/LA) or KA/KW < La/L* (when KT/LT > KA/LA).
2. Kx/Lx > KM/LM is consistent with KT < 0, LT < 0. Again a numerical
demonstration will suffice. Let Kx = 20, KM = 100, Lx = 2, LM = 20.
Thus, KT = Kx - KM = - 80 < 0 and LT = Lx - LM = 2 - 20 = - 18 < 0.
And Kx/Lx > KM/LM (i.e., 20/2 > 100/20). Therefore, Kx/Lx > KM/LM
may be consistent with either KA/KW > LA/LW or KA/KW < LA/LW .
Thus, when a country has a large deficit so that it imports both capital
and labor services, it is again not possible to use the Leontief factor-
content ratio to infer the country's relative factor abundance.
The implication of Corollaries 5 and 6 is that a trade imbalance by
itself may in some cases be sufficient to explain the findings that (a) exports
are more labor intensive than imports for a relatively capital-abundant
country and that (b) exports are more capital intensive than imports for
a relatively labor-abundant country.
As shown by Learner in Corollary 4 of his article, these results are
also possible by including a third factor (e.g., land) and assuming balanced
trade so that the net export of labor and capital through trade is balanced
against a net import of land-using goods. However, our analysis shows
that these assumptions may not be necessary. A "sufficiently large"
trade surplus by itself may in some cases be sufficient to explain why the
exports are more labor intensive than imports for a relatively capital-rich
country like the U.S.

References

Learner, Edward E., "The Leontief Paradox, Reconsidered". Journal of Political

Economy, Vol. 88, 1980, pp. 495 - 503.
Vanek, Jaroslav, "The Factor Proportions Theory: The η-Factor Case". Kyklos,
Vol. 21, 1968, pp. 749 - 756.

This content downloaded from 180.179.211.153 on Tue, 07 Nov 2023 16:54:10 +00:00
All use subject to https://about.jstor.org/terms