IJAE MARCH 2015 3-29-2015 ManuscriptHeckscherOhlinAssignment PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 29

51

Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Heckscher-Ohlin Model with Assignment Problem with


Skilled and Unskilled Labor
Tsuyoshi Sasaki*

Osaka City University

Abstract: In this work, I construct an extended model of the Heckscher-Ohlin model, in


which there is an assignment problem of how to assign workers to two types of
endowments, skilled and unskilled labor. In the comparative statics for a small open
economy, when product prices change, the outputs change more than in the Heckscher-
Ohlin model. In the comparative statics for a whole economy, when consumer tastes or
production technologies change, the product prices change less than in the Heckscher-
Ohlin model. Furthermore, a greater elasticity of substitution between endowments
causes a greater extent of change in the outputs for a small open economy and a lesser
extent of change in the product prices for a whole economy. These results explain why we
see different kinds of relationships between wage changes and other economic changes.

Keyword: Heckscher-Ohlin model, assignment model, heterogeneous agents, income


differentials

JEL Classification: D33, F11, F61, I31, J31

1. Introduction

In the early 20th century, Heckscher and Ohlin developed the Heckscher-Ohlin model [see
Heckscer-Ohlin (1991)]. The Heckscher-Ohlin model assumes that differences in factor
endowments among countries cause international trade, and the pattern of international
trade depends on the differences in the factor endowments. This insight is different from
the Richardian model [see Richardo (1817)]. In the Richardian model, the causes of
international trade are technological differences across countries, and comparative
advantages determine the patterns of international trade.

The Heckscher-Ohlin model has been extended, modified, and examined by many
economists. Stolper and Samuelson (1941) formulated the Stolper-Samuelson Theorem,
which states that an increase in a product price increases the factor price used intensively
for that good and reduces the other factor price. Rybczynski (1955) found that an increase
in a factor endowment increases the output of the product using the factor intensively and
reduces the other output. Vanek (1968) presented and examined the multigood and
multifactor Heckscher-Ohlin model. The empirical research on the Heckscher-Ohlin
model includes Leontief (1953), Leamer (1980), Bowen, Leamer, and Sveikauskas (1987),
Trefler (1993), and others.1
52
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

In this paper, I extend the Heckscher-Ohlin model using concepts of assignment models.
Assignment models are models in which heterogeneous resources are assigned to tasks
based on characteristics of the resources and the tasks, and the outputs are then produced.
Roy (1951) considered models in which workers choose their jobs from among different
kinds of jobs based on which job is best suited for them. Tinbergen (1951, 1956, and 1970)
considered models in which the preference of each worker determines which job each
worker chooses. Sattinger (1979) introduced a one-to-one matching model in which
workers have different levels of ability and machines have different sizes. This model is
extended in many ways, including by Teulings (1995 and 2005), Costrell and Loury
(2004), Tervio (2008) and Sasaki (2012).

Martin (1976), Chen (1995), and others have discussed the effects of endogenous factor
supply on the Heckscher-Ohlin model. In these papers, the supply of factors depends on
the factor prices. For example, when workers determine how long they work, an increase
in wage increases their working hours and reduces their leisure time. Thus, the supply of
labor depends on the wage for labor.

Here, I consider an assignment problem between two factors, say skilled labor and
unskilled labor. Workers choose their jobs based on the factor price ratio between skilled
labor and unskilled labor. As the ratio changes, workers alter their jobs. As a result, the
supply of both skilled labor and unskilled labor depends on the wages for both skilled
labor and unskilled labor. Thus, this paper proposes another aspect of endogenous factor
supply on the Heckscher-Ohlin model.

In international trade literature, concepts of assignment models are used by Leamer


(1999), Grossman and Maggi (2000), Ohnsorge and Trefler (2007), and others. Leamer
(1999), assuming that effort levels affect productivity, analyzes income distribution in
developing countries and developed countries. Grossman and Maggi (2000) consider a
model in which countries have similar aggregate factor endowments, but different
diversity and ability relationships. These assumptions are what create types of trade
patterns. Ohnsorge and Trefler (2007) assume that there are heterogeneous workers, and
they analyze what types of trade pattern and income distributions are created in the
economy by sorting mechanism.

In this paper, my main focus is the relationship between assignment and prices of factor
and product. Heterogeneous workers choose to work as skilled workers or unskilled
workers. When product prices or factor endowments change, the changes cause factor
prices and outputs to change. However, the factor prices and output changes also alter
how factor endowments are assigned to sectors optimally because a change in factor
prices might alter which job each worker is most suited for. In the previous studies, these
relationships are not examined extensively.

Today, as globalization evolves, developed countries are exposed to extensive


competition with developing countries, which have cheap labor. Some have said that
globalization and competition with developing countries cause wages, especially for
unskilled work, to drop, widening income differentials. One explanation is the Factor Price
53
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Equalization Theorem of the Heckscher-Ohlin model. On one hand, the competition with
developing countries in unskilled labor might cause the wages of unskilled labor to
decrease and cause loss of jobs. On the other hand, competition within developed
countries also has consequences and might reduce the wages of others. Thus, when
considering the effect of competition with developing countries, one must take into
consideration not only competition with developing countries but also competition
within developed countries. In this paper, I attempt to extend the Heckscher-Ohlin model
to include both effects and explain the reality.

2. The Model

2.1 Assumption

In a Heckscher-Ohlin model, countries have fixed amounts of factor endowments. For


example, assuming that there are two countries and two resources, labor and capital, the
first country has 500 units of labor and 300 units of capital and the other has 400 units of
capital and 250 units of capital. More generally, we suppose that a country has A of labor
and B of capital. Using those endowments, countries produce products and trade a portion
of those products.

I think that one of the natural extensions of the Heckscher-Ohlin model is cases in which
endowments are given in a function form, not fixed amounts, and are, to an extent,
interchangeable. Considering labor, if one wants to produce a machine, more efficient
skilled workers are preferable. However, less efficient skilled workers might be also
acceptable to a extent. If the wage difference between the less and more efficient skilled
workers is large enough, less efficient skilled workers are more adequate from cost-
efficiency standpoint. Thus, with wages changing, there is a possibility that the amounts
of effective labor that each industry can use are changing.

This type of relationship is also true between labor and capital. If the capital price is too
high, producing capital from labor might be more efficient. Thus, the amount of capital
with which a country is endowed might not be considered a fixed amount.

Figure 1 is the Heckscher-Ohlin model's two-factor case, where x, on the horizontal axis,
is an amount of one of the two factors and y, on the vertical axis, is an amount of the other
factor. The figure depicts the endowment possibility frontier. The amounts of each factor
are fixed.

Conversely, Figure 2 depicts cases in which endowments are given in a function form. The
figure depicts the endowment possibility frontier. If one wants to increase the amount of
one of the factors, one can do that in exchange for a decrease in the amount of the other
factor. As the relative price changes, the amounts of factors change. The fact that
endowments are not fixed gives additional flexibility to the model.

In the Heckscher-Ohlin model case, Figure 1, whatever the factor prices are, the amounts
of the endowments are fixed. Conversely, in Figure 2, with factor prices changing, the
54
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

amounts of the endowments change too. This means that, as the factor prices change, a
portion of one of the factors transforms into the other factor.

For simplicity, we assume that we are able to add up different levels of worker ability and
that the sum of the abilities is the amount of a factor. For example, assume that there are
three workers, Mike, John, and Alice. Mike has three of skilled work abilities, John has two,
and Alice has four. They have the same unskilled work ability, say one. If all three are
utilized as skilled workers, there are nine skilled work factors. If all three are utilized as
unskilled workers, there are three unskilled work factors. In more complex models,
combinations of workers' abilities and jobs' difficulties must affect outputs. However, this
simplified assumption largely reduces the difficulty of analysis.

For example, let's consider skilled workers and unskilled workers. For skilled work, more
competent people are probably preferable to get a job done. However, if less competent
workers need much lower wages than more competent workers, less competent workers
might be preferable to companies. Considering the economy as a whole, as the wage ratio
of skilled workers to unskilled workers grows, more workers are assigned to skilled work,
which is done more efficiently. However, as more and more workers are assigned to
skilled work, the marginal increase of efficiency unit of knowledge worker per newly
assigned worker decreases. Since, as more and more workers are assigned to knowledge
work, the competence of newly reassigned workers decreases. As a result, the shape of
the function of factor endowments has to be like that shown in Figure 2.

Putting workers' wages on the vertical axis and workers' ability on the horizontal axis,
Figure 3 shows workers' wages. Assume that all workers have the same ability in
unskilled work and receive the same wage, 𝑤𝑢 , and that workers have different ability
levels, x, in skilled work and are compensated on a pay-for-ability basis, 𝑥𝑤𝑠 . Each worker
chooses the optimal work, skilled work or unskilled work, based on how much he or she
gets. As Figure 3 shows, workers with lower ability levels work as unskilled workers and
earn the wages of unskilled work, 𝑤𝑢 . Workers who have higher ability levels work as
skilled workers and earn the wages of skilled work, 𝑥𝑤𝑠 . The wage line of unskilled
workers and the wage line of skilled workers are connected because the point where the
wage lines meet is where workers can earn the same wage on skilled and unskilled work,
and it divides the workers into skilled workers and unskilled workers.

Next, we turn our attention to a comparison of the endowment functions between two
countries. Assume that there are two factors, X and Y. Let 𝑤𝑖 , 𝑖 = 𝑥, 𝑦 be the factor prices
𝑤𝑦
and w be the relative factor price, 𝑤 . Let 𝑥𝑖 (𝑤), 𝑖 = 1, 2 be the endowment of factor X that
𝑥
i country has at the relative factor price w and 𝑦𝑖 , 𝑖 = 1, 2 be the endowment of factor Y
that i country has at the relative factor price, w.
Definition 2. 1.

Assume that there are two factors, x and y, and two countries. When the relative factor price
𝑦 (𝑤)
is w, if the relative factor endowment of Y factor to X factor in country 1, 𝑥1 (𝑤), is larger than
1
55
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

𝑦 (𝑤)
the relative factor endowment of Y factor to X factor in country 2, 𝑥2(𝑤), then country 1 is Y
2
abundant at the relative price w.

This definition is straightforward. For example, given a relative factor price, if country 1
possesses 100 units of factor Y and 150 units of factor X and country 2 possesses 150 units
2
of factor Y and 200 units of factor X, the relative factor amount of country 1 is 3 and that
3
of country 2 is 4. Then, country 2 is Y abundant at the given relative price and country 1 is
X abundant at the given relative price.

Definition 2. 1. is true when a relative price is given. Next, we generalize Definition 2. 1.


to a property that holds true at any relative price.

Definition 2. 2.

Assume that there are two factors, x and y, and two countries. Whatever the relative price is,
𝑦 (𝑤)
if the relative factor endowment of Y factor to X factor in country 1, 𝑥1 (𝑤), is larger than the
1
𝑦2 (𝑤)
relative factor endowment of Y factor to X factor in country 2, 𝑥 , then country 1 is
2 (𝑤)
completely Y abundant.

Definition 2. 2. is a very useful property because, as explained later, being completely i


abundant guarantees that a country is abundant in a factor whatever the relative price is,
and it simplifies the analysis.

For example, assume that there are two types of jobs, skilled labor and unskilled labor.
For unskilled labor, each worker produces the same amount. Conversely, when engaging
in skilled work, each worker produces different amounts. Assume that there are two
countries and that each country has a continuum of labor, 𝑖 ∈ [0, 1]. The continuum of
labor is indexed by relative ability in skilled work. The 0-th worker is the least competent
and the 1-th worker is the most competent. Let 𝑔𝑗 (𝑖) be the ability level that the i-th
worker in country j has. The ability level is addable in production. We get the next
proposition.

Proposition 2. 1.

Assume that there are two countries and two types of jobs, unskilled and skilled, and that
every worker produces the same amount in the first type of job, an unskilled job. If 𝑔1 (𝑖) is
higher than 𝑔2 (𝑖) at any i, then country 1 is completely skilled labor abundant.

Proof 2. 1.

At a given relative factor price, country 2 has more workers assigned to unskilled work.
Because 𝑔1 (𝑖) is higher than 𝑔2 (𝑖) at any i, the number of workers who have less than a
56
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

certain ability level is larger in country 2 than in country 1. The ability of a worker at any i
is higher in country 1, and more workers are assigned to skilled work in country 1 than in
country 2. Thus, the sum of the ability level assigned to skilled work in country 1 is higher
than in country 2. Therefore, the relative factor amount of skilled labor to unskilled labor in
country 1 is larger than in country 2.

2. 2 Existence of Equilibrium

Assume that there are two countries. The countries have the same production
technologies. The production functions are assumed to be increasing, concave, and
homogeneous of degree one in inputs, skilled labor, 𝐿𝑠 , and unskilled labor, 𝐿𝑢 . There are
two products. Producing the first product, say 𝑋1, requires more skilled labor compared
to the other, 𝑋2. FIRs (factor intensity reversals) do not occur. The market is completely
competitive. We assume that skilled labor and unskilled labor are fully mobile within
countries but not mobile between countries. Both countries have the identical homothetic
taste. Endowments, skilled labor and unskilled labor, are given in a functional form. The
endowment functions are decreasing and concave, and one of the countries is completely
skilled labor abundant. Let 𝑝𝑖 , 𝑖 = 1, 2 be the prices of products and 𝑤𝑗 , 𝑗 = 𝑠, 𝑢 be the
prices of factors, skilled labor and unskilled labor.
𝑝
First, we consider whether equilibrium exists. Let 𝑝 = 𝑝1 be the relative price and 𝑋 𝑠 and
2
𝑋
𝑋 𝑑 , 𝑋1 , be the relative supply and demand of products. With the relative price on the
2
vertical axis and the relative supply and demand on the horizontal axis, the world demand
curve is a downward slopping curve. The relative supply curve is an upward slopping
curve because the endowment functions and the production functions are concave.
Therefore, the intersection is the equilibrium.

Proposition 2. 2.

If both countries produce both goods, the factor prices are equalized across countries.
As in a basic Heckscher-Ohlin model, the Factor Price Equalization Theorem holds true.
Given prices, if a country produces both goods, the factor prices are uniquely determined
to be where the unit-cost lines intersect because FIRs do not occur, the factor prices are
stable because the endowment functions are concave.

Proposition 2. 3.

Suppose that two countries produce two goods and engage in free trade and that FIRs do
not occur. If a country is completely X abundant, then the country exports the product using
X intensively.
When both countries produce both goods, according to Proposition 2. 2, the factor prices
are equalized between countries. This implies that the country that is completely skilled
labor abundant produces more skilled labor intensive goods relative to the other country.
Taking into consideration that the countries have the identical homothetic taste, the
skilled labor abundant country exports skilled labor intensive goods. When specialization
57
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

occurs, the only possibility is that the skilled labor abundant country will specialize in
producing the skilled labor intensive goods and the unskilled labor abundant country will
specialize in producing the unskilled labor intensive goods. Thus, each country exports
goods that use the abundant factor intensively.

Proposition 2. 4.

Assume that there are two countries, one is completely X abundant relative to the other, and
two goods, a good using X intensively and a good using Y intensively, that FIRs do not occur,
and that the countries change their trade policy from a closed economy to a open economy.
Then, in the completely X abundant country, the prices of X and a good using X intensively
increase, and the prices of Y and a good using Y intensively decrease. Furthermore, in the
completely Y abundant country, the prices of Y and a good using Y intensively increase, and
the prices of X and a good using X intensively decrease.

As Proposition 2. 2 and 2. 3 state, the product prices and the factor prices are equalized
across countries, and each country exports a good using each country's abundant factor.
This means that, comparing the situation before and after the trade policy change, each
country increases the amount of a good using each country's abundant factor and exports
them. Thus, if there is no trade between the countries, there will be a surplus of the good
using each country's abundant factor at the prices in the open economy. As a result, in
each country, the price of the good using each country's abundant factor in the closed
economy is lower than in the open economy. As explained later in equation (6) of the next
subsection, when the price of a good using a factor intensively increases, the price of the
factor increases.

2. 3 Model Structure

2. 3. 1 Functions

As in the previous subsections, let 𝑤𝑗 , 𝑗 = 𝑠, 𝑢, be the prices of labor (s represents skilled


and u represents unskilled) and 𝑝𝑖 , 𝑖 = 1, 2, be the prices of products. Let 𝐿𝑗 , 𝑗 = 𝑠, 𝑢, be
the amount of labor employed as skilled workers, s, and unskilled workers, u, and 𝐿𝑗,𝑖 , 𝑖 =
1, 2 and 𝑗 = 𝑠, 𝑢, be the amount of labor employed in industry i as skilled and unskilled
labor. Thus, the following holds.

𝐿𝑠 = 𝐿𝑠,1 + 𝐿𝑠,2 , 𝐿𝑢 = 𝐿𝑢,1 + 𝐿𝑢,2 .

Assume that the production functions for a good i, i=1,2, are 𝐹𝑖 (𝐿𝑠,𝑖 , 𝐿𝑢,𝑖 ) and have constant
return to scale, 𝐹𝑖 (𝑎𝐿𝑠,𝑖 , 𝑎𝐿𝑢,𝑖 ) = 𝑎𝐹𝑖 (𝐿𝑠,𝑖 , 𝐿𝑢,𝑖 ). Let 𝑦𝑖 , 𝑖 = 1, 2, be the output of good i,
𝐹1 (𝐿𝑠,1 , 𝐿𝑢,1 ) = 𝑦1 , 𝐹2 (𝐿𝑠,2 , 𝐿𝑢,2 ) = 𝑦2.

Let 𝑙𝑗,𝑖 , 𝑗 = 𝑠, 𝑢 and i=1, 2 be the amount of skilled and unskilled labor to produce one unit
𝐿𝑠,𝑖 𝐿𝑢,𝑖
of good i. Thus, each industry needs 𝑙𝑠,𝑖 = , 𝑖 = 1, 2, of skilled labor and 𝑙𝑢,𝑖 = ,𝑖 =
𝑦𝑖 𝑦𝑖
1, 2, of unskilled labor.
58
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

The market is completely competitive. Thus, profits of firms are zero. As a result, the
prices of products times the derivatives of the production functions are equal to the prices
of labor:

𝜕𝐹𝑖 (𝐿𝑠,𝑖 , 𝐿𝑢,𝑖 )


𝑝𝑖 = 𝑤𝑗 , 𝑖 = 1, 2, 𝑗 = 𝑠, 𝑢.
𝜕𝐿𝑗,𝑖

And the prices of products must be equal to the costs of production:

𝑝𝑖 = 𝑤𝑠 𝑙𝑠,𝑖 + 𝑤𝑢 𝑙𝑢,𝑖 , 𝑖 = 1, 2.

Assume that all consumers have the same utility function of Constant Elasticity of
Substitution (CES) form:
𝜇
𝜇−1 𝜇−1 𝜇−1
𝑢(𝑥1 , 𝑥2 ) = {(𝜏1 𝑥1 ) 𝜇 + (𝜏2 𝑥2 ) 𝜇 } ,

where 𝜏𝑖 , 𝑖 = 1, 2, is the parameter discribing how a consumer likes good I and 𝜇𝑢 is the
elasticity of substitution with respect to consumer demand. For notational purposes, we
omit subscript u here. Each consumer maximizes their utility, subject to their budget
constraints:

𝑚𝑎𝑥 𝑢(𝑥1 , 𝑥2 ), 𝑠. 𝑡. 𝑝1 𝑥1 + 𝑝2 𝑥2 = 𝐵𝑖 , 𝑖 ∈ 𝐶,
𝑥1 ,𝑥2

where 𝐵𝑖 is a budget of a consumer and C is set of all consumers. Since the utility functions
are the CES form, the following equation holds for all consumers:

𝜏1 𝑥1 𝜇 𝑝1
( ) = .
𝜏2 𝑥2 𝑝2
𝑥 𝑦
Taking into consideration 𝑥1 = 𝑦1, we get
2 2

(𝜏 𝑦)𝜇 = 𝑝, (1)
𝜏 𝑦
where 𝜏 = 𝜏1 and 𝑦 = 𝑦1.
2 2

Assume that the implicit endowment function is written as


𝑀 = 𝑚(𝐿𝑠 , 𝐿𝑢 ). (2)

The implicit endowment function means that 𝐿𝑠 and 𝐿𝑢 can take on values which satisfy
the implicit endowment equation. Arranging (2), we get

𝐿𝑠 = 𝑔(𝐿𝑢 ), (3)
59
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

where g(.) is the explicit endowment function. The explicit endowment function means
that, when the economy has 𝐿𝑢 of unskilled labor, the economy can have 𝐿𝑠 of skilled labor
in maximum. Since labor is divided into skilled and unskilled labor based on a relative
wage, the explicit endowment function satisfies the following equations:

𝑑𝑔 𝑑2 𝑔
< 0, 𝑑𝐿2 < 0, 𝑠. 𝑡. 𝐿𝑠 ≥ 0, 𝐿𝑢 ≥ 0.
𝑑𝐿𝑢 𝑢

The derivative of (3) must be equal to the ratio of the wage of unskilled labor to the wage
of skilled labor; otherwise, the equilibrium is not attained:

𝑑𝑔(𝐿𝑢 ) 𝑤𝑢 1
= ⟹ 𝑔′ = ,
𝑑𝐿𝑢 𝑤𝑠 𝑤

𝑑𝑔(𝐿𝑢 )
where 𝑔′ = and w is the ratio of the factor price of skilled labor to the factor price
𝑑𝐿𝑢
𝑤
of unskilled labor, 𝑤 𝑠 . When the ratio of factor prices, w, and the amount of labor, 𝐿𝑠 and
𝑢
𝐿𝑢 , change by 𝑑𝑤, 𝑑𝐿𝑠 , 𝑑𝐿𝑢 , the following equation must hold true:

𝑔′ 𝑑𝑤
𝑑𝐿𝑢 = − 𝑔′′ 𝑤 , (4)

𝑑2 𝑔(𝐿𝑢 )
where 𝑔′′ = .
𝑑𝐿2𝑢

2. 3. 2 Relationship between Product Prices and Factor Prices

We turn our attention to relationships among the product prices, the factor prices, the
outputs, and others.2 Let

𝑐𝑖 (𝑤𝑠 , 𝑤𝑢 ) = 𝑚𝑖𝑛{𝑤𝑠 𝑙𝑠,𝑖 + 𝑤𝑢 𝑙𝑢,𝑖 |𝐹𝑖 (𝑙𝑠,𝑖 , 𝑙𝑢,𝑖 ) = 1}, 𝑖 = 1, 2

be the unit-cost functions which is the minimum costs produce a unit of good i. The zero
profit conditions are

𝑝𝑖 = 𝑐𝑖 (𝑤𝑠 , 𝑤𝑢 ), 𝑖 = 1, 2.

Totally differentiating the above, we get

𝑑𝑝𝑖 = 𝑎𝑖,𝑠 𝑑𝑤𝑠 + 𝑎𝑖,𝑢 𝑑𝑤𝑢


𝑑𝑝 𝑤 𝑎 𝑑𝑤 𝑤 𝑎 𝑑𝑤
⟹ 𝑝 𝑖 = 𝑠𝑐 𝑖,𝑠 𝑤 𝑠 + 𝑢𝑐 𝑖,𝑢 𝑤 𝑢 , 𝑖 = 1, 2, (5)
𝑖 𝑖 𝑠 𝑖 𝑢
where 𝑎𝐼,𝑗 , 𝑖 = 1, 2, 𝑗 = 𝑠, 𝑢, are the optimal amount of skilled labor or unskilled labor
given ws and wu .3 Rearranging (5), we get

𝑝̂𝑖 = 𝜃𝑖,𝑠 𝑤
̂𝑠 + 𝜃𝑖,𝑢 𝑤
̂𝑢 , 𝑖 = 1, 2,
60
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

𝑑𝑝𝑖 𝑑𝑤𝑗
where 𝑝̂𝑖 = and 𝑤
̂𝑗 = , 𝑗 = 𝑠, 𝑢, and 𝜃𝑖,𝑗 are the cost share of j labor in the i industry.
𝑝𝑖 𝑤𝑗
Thus, 𝜃𝑖,𝑗 satisfies 𝜃𝑖,𝑠 + 𝜃𝑖,𝑢 = 1.

These equations are rewritten in matrix form and solved as

𝑝̂ 𝜃1,𝑠 𝜃1,𝑢 𝑤 ̂
( 1) = ( ) ( 𝑠)
𝑝̂2 𝜃2,𝑠 𝜃2,𝑢 𝑤 ̂𝑢
𝑤
̂𝑠 1 𝜃 2,𝑢 −𝜃1,𝑢 𝑝̂1
⟹ ( ) = |𝜃| ( ) ( ), (6)
𝑤
̂𝑢 −𝜃2,𝑠 𝜃1,𝑠 𝑝̂ 2

where |𝜃| denotes the determinant of the matrix on the upper equations and can be
expressed as

|𝜃| = 𝜃1,𝑠 𝜃2,𝑢 − 𝜃1,𝑢 𝜃2,𝑠 = 𝜃1,𝑠 − 𝜃2,𝑠 = 𝜃2,𝑢 − 𝜃1,𝑢 .

(6) means that changes in the product prices have a magnified effect on the factor prices.
Assuming that 𝑝̂1 − 𝑝̂ 2, we get

𝑤
̂𝑠 > 𝑝̂1 > 𝑝̂ 2 > 𝑤
̂𝑢 .

Jones (1965) called this equation the magnification effect.

2. 3. 3 Relationship between Amounts of Factors and Outputs

The full employment conditions for the economy are written as

𝐿𝑠 = 𝑎1,𝑠 𝑦1 + 𝑎2,𝑠 𝑦2
𝐿𝑢 = 𝑎1,𝑢 𝑦1 + 𝑎2,𝑢 𝑦2 ,

where 𝑦𝑖 , 𝑖 = 1, 2 are the outputs in i good. Totally differentiating the above, we get

𝑑𝐿𝑗 = 𝑎1,𝑗 𝑑𝑦1 + 𝑎2,𝑗 𝑑𝑦2


𝑑𝐿𝑗 𝑦1 𝑎1,𝑗 𝑑𝑦1 𝑦2 𝑎2,𝑗 𝑑𝑦2
⟹ = +
𝐿𝑗 𝐿𝑗 𝑦1 𝐿𝑗 𝑦2
̂
⟹ 𝐿𝑗 = 𝜆1,𝑗 𝑦̂1 + 𝜆2,𝑗 𝑦̂2 , 𝑗 = 𝑠, 𝑢,

where 𝜆𝑖,𝑗 is the fraction of the skilled or unskilled labor employed in industry i.4 These
equations are rewritten in matrix form and solved as
𝐿̂ 𝜆1,𝑠 𝜆2,𝑠 𝑦̂1
( 𝑠) = ( )( )
𝐿̂𝑢 𝜆1,𝑢 𝜆2,𝑢 𝑦̂2
𝑦̂ 1 𝜆2,𝑢 −𝜆2,𝑠 𝐿̂𝑠
( 1 ) = |𝜆| ( ) ( ), (7)
𝑦̂2 −𝜆1,𝑢 𝜆1,𝑠 𝐿̂𝑢
61
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

where |λ| is the determinant of the matrics on the upper equations and can be expressed
as

|𝜆| = 𝜆1,𝑠 𝜆2,𝑢 − 𝜆2,𝑠 𝜆1,𝑢 = 𝜆1,𝑠 − 𝜆1,𝑢 = 𝜆2,𝑢 − 𝜆2,𝑠 .

2. 3. 4 Substitution between Factors

Let w be the ratio of the factor price of skilled labor to the factor price of unskilled labor,
𝑤𝑠 𝑑𝑤
, and 𝑤
̂ is . Thus, 𝑤
̂ means 𝑤̂𝑠 − 𝑤
̂𝑢 . From (4), the transformation equations for the
𝑤𝑢 𝑤
factors, skilled labor and unskilled labor, are written as

𝐿̂𝑠 = 𝑏𝜂(𝑤
̂𝑠 − 𝑤
̂𝑢 )
̂𝐿𝑢 = 𝜂(−𝑤̂𝑠 + 𝑤
̂𝑢 ), (8)

𝑤𝑢 𝐿𝑢 𝑔′
where b is and 𝜂 is 𝑔′′ 𝐿 and a parameter that measures how much a change in the
𝑤𝑠 𝐿𝑠 𝑢
factor prices affects the endowments of labor, both skilled and unskilled. In (8), when 𝑤 ̂𝑠
𝑤𝑢 𝐿𝑢
and 𝑤̂𝑢 change by the same percentages, 𝐿̂𝑠 and 𝐿̂𝑢 must not change. Why b is 𝑤 𝐿 is
𝑠 𝑠
because the values of increased labor force and decreased labor force have to be the same.
𝑤 𝐿
This implies that b is 𝑤𝑢 𝐿𝑢.
𝑠 𝑠

2. 3. 5 Relationship between Outputs and Factor Ratios


𝑎 𝑎
Let 𝑎1 be the ratio of 𝑎1,𝑠 to 𝑎1,𝑢 , 𝑎 1,𝑠 , and 𝑎2 be the ratio of 𝑎2,𝑠 to 𝑎2,𝑢 , 𝑎 2,𝑠 . Let L denote
1,𝑢 2,𝑢
𝐿𝑠
the ratio of the endowment of skilled labor to the endowment of unskilled labor, .
𝐿𝑢
Assuming 𝑎2 < 𝐿 < 𝑎1 , the fraction of the skilled and unskilled labor employed in
industry i, 𝜆𝑖,𝑗 , 𝑖 = 1, 2, 𝑗 = 𝑠, 𝑢, is rewritten as

𝑎1 (𝐿 − 𝑎2 ) 𝐿 − 𝑎2
𝜆1,𝑠 = , 𝜆1,𝑢 =
𝐿(𝑎1 − 𝑎2 ) 𝑎1 − 𝑎2
𝑎2 (𝑎1 − 𝐿) 𝑎1 − 𝐿
𝜆2,𝑠 = , 𝜆2,𝑢 = .
𝐿(𝑎1 − 𝑎2 ) 𝑎1 − 𝑎2

Therefore, the amount of the outputs, 𝑦1 and 𝑦2 , are rewritten as

𝐿 − 𝑎2
𝑦1 = {𝑎 𝑏 + 𝑏1,𝑢 }𝐿𝑢
𝑎1 − 𝑎2 1 1,𝑠
𝑎1 − 𝐿
𝑦2 = {𝑎 𝑏 + 𝑏2,𝑢 }𝐿𝑢
𝑎1 − 𝑎2 2 2,𝑠

where 𝑏𝑖,𝑗 is the marginal productivity of the skilled and unskilled labor measured by the
𝑤
product, which can be rewritten as 𝑝 𝑗. Differentiating and arranging the above, we get
𝑖
62
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

𝑦̂1 = 𝜁1,1 𝑎̂1 , 𝑦̂1 = 𝜁1,2 𝑎̂2 , 𝑦̂2 = 𝜁2,1 𝑎̂1 , 𝑦̂2 = 𝜁2,2 𝑎̂2 , (9)

Where

−𝑎1 𝑎1 𝑤 𝑎2 𝑎2
𝜁1,1 = { + } , 𝜁1,2 = { − },
𝑎1 − 𝑎2 𝑎1 𝑤 + 1 𝑎1 − 𝑎2 𝐿 − 𝑎2
−𝑎1 𝑎1 𝑎2 𝑎2 𝑤
𝜁2,1 ={ + } , 𝜁2,2 = { + },
𝑎1 − 𝑎2 𝑎1 − 𝐿 𝑎1 − 𝑎2 𝑎2 𝑤 + 1
𝑑𝑦𝑖 𝑑𝑎𝑖
and 𝑦̂𝑖 and 𝑎̂𝑖 are and , respectively.
𝑦𝑖 𝑎𝑖

−𝑎1 𝑎1 𝑤 𝑎2 𝑎 −𝑎1 𝑎1 −𝑎1 𝑎1


Since {𝑎 +𝑎 } < 0, {𝑎 − 𝐿−𝑎2 } < 0, {𝑎 +𝑎 } > 0, and {𝑎 +𝑎 }>0
1 −𝑎2 1 𝑤+1 1 −𝑎2 2 1 −𝑎2 1 −𝐿 1 −𝑎2 1 −𝐿

hold true, a change to more skilled labor intensive production decreases the output using
skilled labor intensively and increases the output using unskilled labor intensively.

2. 3. 6 Substitution in Production and Consumer Demand

The relationships between the factor prices, 𝑤1 and 𝑤2 , and the ratios between the
optimal amounts of skilled or unskilled labor, 𝑎𝑖,𝑗 , is written as

𝑎̂1 = −𝜇1 𝑤
̂, 𝑎̂2 = −𝜇2 𝑤
̂, (10)

where 𝜇𝑖 , 𝑖 = 1, 2 are the elasticity of substitution with respect to production. Totally


differentiating (1), the relationship between the product prices, 𝑝1 and 𝑝2 , and the
outputs, 𝑦1 and 𝑦2 , is written as

𝑦̂ = −𝜇𝑢 𝑝̂ (11)

in cases in which a change in consumer taste does not happen, and

𝑦̂ + 𝜏̂ = −𝜇𝑢 𝑝̂ (12)

𝑑𝑦 𝑑𝜏 𝑑𝑝
in cases in which a change in consuer taste happens, where 𝑦̂,𝜏̂ , and 𝑝̂ are , , and ,
𝑦 𝜏 𝑝
respectively, and 𝜇𝑢 is the elasticity of substitution with respect to consumer demand.
63
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

2. 4 Comparative Statics for a Small Open Economy

We turn our attention here to the comparative statics. In this subsection, we examine the
comparative statics for a small open economy. We analyze the comparative statics for a
whole economy in the next subsection.

In this subsection, we consider a small open economy that is too small to affect the
international prices of goods and thus behaves as a price taker. We do not assume that the
small open economy and the other countries have the same technologies. Thus, the small
open economy maximizes the total value of the products they produce under the
conditions under which the product prices are given.

2. 4. 1 Effect of Product Price Changes

Proposition 2. 5.

Assume that endowments are given in a functional form. For a small open economy, when
the relative price of a product increases, the output of the product increases more than when
endowments are given by fixed amounts.

When the product prices, 𝑝1 and 𝑝2 , change by 𝑝̂ 𝑖 , 𝑖 = 1, 2 percentage, from (6), (7), (8),
(9), and (10), the effects are written as

𝜃2,𝑢 −𝜃2,𝑠 𝑏𝜂 −𝜂
𝑤
̂𝑠 = 𝑝̂1 , 𝑤̂𝑢 = 𝑝̂1 , 𝐿̂𝑠 = 𝑝̂1 , 𝐿̂𝑢 = 𝑝̂ ,
|𝜃| |𝜃| |𝜃| |𝜃| 1
−𝜇1 −𝜇2
𝑎̂1 = 𝑝̂1 , 𝑎̂2 = 𝑝̂ ,
|𝜃| |𝜃| 1
𝜆2,𝑢 𝑏𝜂 + 𝜆2,𝑠 𝜂 𝑝̂1
𝑦̂1 = {−𝜁1,1 𝜇1 − 𝜁1,2 𝜇2 + } ,
|𝜆| |𝜃|
𝜆1,𝑢 𝑏𝜂 + 𝜆1,𝑠 𝜂 𝑝̂1
𝑦̂2 = {−𝜁2,1 𝜇1 − 𝜁2,2 𝜇2 − }
|𝜆| |𝜃|

in a 𝑝̂1 change case, and

−𝜃1,𝑢 𝜃1,𝑠 −𝑏𝜂 𝜂


𝑤
̂𝑠 = 𝑝̂ 2 , 𝑤
̂𝑢 = 𝑝̂2 , 𝐿̂𝑠 = 𝑝̂ 2 , 𝐿̂𝑢 = 𝑝̂ ,
|𝜃| |𝜃| |𝜃| |𝜃| 1
𝜇1 𝜇2
𝑎̂1 = 𝑝̂ 2 , 𝑎̂2 = 𝑝̂ ,
|𝜃| |𝜃| 2
𝜆2,𝑢 𝑏𝜂 + 𝜆2,𝑠 𝜂 𝑝̂2
𝑦̂1 = {𝜁1,1 𝜇1 + 𝜁1,2 𝜇2 − } ,
|𝜆| |𝜃|
𝜆1,𝑢 𝑏𝜂 + 𝜆1,𝑠 𝜂 𝑝̂ 2
𝑦̂2 = {𝜁2,1 𝜇1 + 𝜁2,2 𝜇2 + }
|𝜆| |𝜃|

in a 𝑝̂2 change case. This result is similar to the Stolper-Samuelson Theorem in Stolper
and Samuelson (1941). The differences are that, in this paper's model, the changes in the
factor prices, caused by product price changes, make the factor endowments change. This
64
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

is because, when the factor prices change, it might be beneficial for some skilled or
unskilled workers to change from one job to the other. As a result, the product price
changes have a greater effect on the outputs. Thus, Proposition 2. 5 occurs.

2. 4. 2 Effect of Endowment Function Changes

Assume that the factor endowment function changes, and, as a result, the amounts of the
factors change by 𝐿̂𝑗 , 𝑗 = 𝑠, 𝑢 percentage. From (7), the effects are written as

𝜆2,𝑢 −𝜆1,𝑢
𝑦̂1 = 𝐿̂𝑠 , 𝑦̂2 = 𝐿̂
|𝜆| |𝜆| 𝑠

in the 𝐿̂𝑠 change case, and

−𝜆2,𝑠 𝜆1,𝑠
𝑦̂1 = 𝐿̂𝑢 , 𝑦̂2 = 𝐿̂
|𝜆| |𝜆| 𝑢

in the 𝐿̂𝑢 change case. This result is analougous with the Rybczynski Theorem in
Rybczynski (1955). However, in this paper's model, when a change in endowments
happens, both skilled labor and unskilled labor might change. For example, we consider
cases like that of Proposition 2. 1. If the change is that the later endowment function is
completely skilled labor abundant relative to the initial one, and the endowment of skilled
labor increases, the endowment of unskilled labor decreases.

2. 4. 3 Effect of Technological Changes

We turn our attention here to technological changes.


Assume that the technologies of industry i change, and the factor prices and the optimal
amount of skilled or unskilled labor change by 𝑤̂𝑗 and 𝑎̂1,𝑗 , 𝑗 = 𝑠, 𝑢. Since the unit cost line
of industry i is a downward slopping curve, if 𝑤
̂𝑠 > 0 holds true, then 𝑤 ̂𝑢 < 0; conversely,
if 𝑤
̂𝑢 > 0 holds true, then 𝑤
̂𝑠 < 0. From (7), (8), (9), and (10), the subsequent effects that
the technological changes make are written as

𝑎̂𝑗 = −𝜇𝑗 𝑤 ̂, 𝐿̂𝑠 = 𝑏𝜂𝑤 ̂, 𝐿̂𝑢 = −𝜂𝑤 ̂,


𝜆2,𝑢 𝑏𝜂 + 𝜆2,𝑠 𝜂
𝑦̂1 = 𝜁1,𝑖 𝑎̂𝑖 − 𝜁1,𝑗 𝜇𝑗 𝑤
̂+ 𝑤
̂,
|𝜆|
𝜆1,𝑢 𝑏𝜂 + 𝜆1,𝑠 𝜂
𝑦̂2 = 𝜁2,𝑖 𝑎̂𝑖 − 𝜁2,𝑗 𝜇𝑗 𝑤
̂− ̂, (𝑖, 𝑗) = (1, 2), (2, 1).
𝑤
|𝜆|

The increase in w causes the output of product 1 to increase through increasing skilled
labor and changing the factor mix in another industry to use less skilled labor. The
decrease of 𝑎1 directly increases the output of product 1 and decreases the output of
product 2. Assuming that w increases and 𝑎1 decreases or w decreases and 𝑎1 increases,
the changes expands industry 1 in functional endowments cases more than in fixed factor
endowments cases.
65
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

2. 5 Comparative Statics for a Whole Economy

In what follows, we consider the comparative statics for a whole economy. In the previous
subsection, the product prices are supposed to be fixed or given externally. In this
subsection, the equilibrium determines the product prices. For simplicity, the economy is
supposed to have the same technologies across countries.

Proposition 2. 6.

Assume that endowments are given in a functional form. For a whole economy, when a
change in consumer's taste or production technologies happens, the change in the outputs
is less than in cases in which endowments are given by fixed amounts.

In this subsection, we verify Proposition 2. 6.

2. 5. 1 Effect of Demand Change

We consider a change of consumer demand or taste. When the consumers' taste changes,
the outputs or the product prices have to change to compensate for the change in
consumer taste. Writing again (12), we get

𝑦̂ + 𝜏̂ = −𝜇𝑢 𝑝̂ .

If the consumers' taste changes, to satisfy the equation, the relative output, y, and the
relative product price, p, change accordingly. Solving (12) by using (6), (7), (8), (9), (10),
and (11), the changes in the relative product price and the relative output are written as5
−1
𝑝̂ = 𝜙+𝜇 𝜏, 𝑦̂ = 𝜙𝑝̂ (13)
𝑢

1 𝑎1 𝑤 𝑎1 𝑎2 𝑎2 𝑤 𝑏𝜂 + 𝜂
𝜙= {(− + ) 𝜇1 + ( + ) 𝜇2 + }.
|𝜃| 𝑎1 𝑤 + 1 𝑎1 − 𝐿 𝐿 − 𝑎 2 𝑎2 𝑤 + 1 |𝜆|

The other effects are written as

1 𝑏𝜂 −𝜂
𝑤
̂= 𝑝̂ , 𝐿̂𝑠 = 𝑝̂ , 𝐿̂𝑢 = 𝑝̂ ,
|𝜃| |𝜃| |𝜃|
−𝜇1 −𝜇2
𝑎̂1 = 𝑝̂ , 𝑎̂2 = 𝑝̂ ,
|𝜃| |𝜃|
𝜆2,𝑢 𝑏𝜂 + 𝜆2,𝑠 𝜂 𝑝̂
𝑦̂1 = {−𝜁1,1 𝜇1 − 𝜁1,2 𝜇2 + } ,
|𝜆| |𝜃|
𝜆1,𝑢 𝑏𝜂 + 𝜆1,𝑠 𝜂 𝑝̂
𝑦̂2 = {−𝜁2,1 𝜇1 − 𝜁2,2 𝜇2 − } .
|𝜆| |𝜃|
𝑎2 𝑎2 𝑤 𝑎1 𝑤 𝑎1
In (13), (𝐿−𝑎 + 𝑎 𝑤+1) is clearly more than one, and (− 𝑎 𝑤+1 + 𝑎 −𝐿 ) is more than one,
2 2 1 1
too, because
66
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

1 1
𝑎1 𝑤 𝑎1 𝑎1 𝑤 𝑎1
− + =( − 1) + (1 + ) > 1.
𝑎1 𝑤 + 1 𝑎1 − 𝐿 1 1 1
𝑎1 𝑤 + 1 𝐿 − 𝑎1

Although changes in the product prices have a magnified effect on the factor prices in the
Hechscher-Ohline type models including this paper's model, the outputs also change
greatly when the product prices change, and the changes in the factor prices change the
ratio between the optimal amounts of skilled and unskilled labor. The changes in the
factor prices have a greater effect on the outputs in the Heckscher-Ohlin type models than
𝑎 𝑎2 𝑤 𝑎1 𝑤 𝑎1
in one product models because (𝐿−𝑎2 + 𝑎 𝑤+1 ) and (− 𝑎 𝑤+1 + 𝑎 −𝐿 ) are more than one,
2 2 1 1
too. As a result, smaller changes in the product prices are able to offset the change in
consumer taste.

This analysis also applies to cases in which the productivity of one industry increases
because a decrease in the product price, caused by the increase in productivity, has a
similar effect to that of a consumer taste change.

2. 5. 2 Effect of Endowment Function Changes

Next, we consider changes in the endowment function. When the endowment function
changes, assuming that the relative factor price is constant, the amounts of skilled and
unskilled labor at the equilibrium factor prices change. Let 𝐿̂𝑗𝐼 be the initial effect of a
change in the endowment function on the amounts of skilled and unskilled labor. Then,
the economy has to satisfy the equations,

𝐿̂𝐼 𝐿̂𝐼
𝑦̂ + 𝜇𝑢 𝑝̂ = |𝜆|𝑠 , 𝑦̂ + 𝜇𝑢 𝑝̂ = − |𝜆|𝑢 . (14)

From (6), (7), (8), (9), (10), (11), (13), and (14), the total effects on the outputs and the
product prices are written as
𝜇𝑢 𝐿̂𝐼𝑠 −1 𝐿̂𝐼𝑠
𝑦̂ = , 𝑝̂ =
𝜇𝑢 + 𝜙 |𝜆| 𝜇𝑢 + 𝜙 |𝜆|

in the 𝐿̂𝐼𝑠 change case, and

−𝜇𝑢 𝐿̂𝐼𝑢 1 𝐿̂𝐼𝑠


𝑦̂ = , 𝑝̂ =
𝜇𝑢 + 𝜙 |𝜆| 𝜇𝑢 + 𝜙 |𝜆|

in the 𝐿̂𝐼𝑢 change case. The amounts of skilled and unskilled labor finally change to

𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝜂
𝐿̂𝑠 = 𝐿̂𝐼𝑠 , (15)
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
𝜂
𝐿̂𝑢 = 𝐿̂𝐼𝑠 ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
67
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

(𝑎1 𝑤) 𝑎1 𝑎2 𝑎2 𝑤
𝐴1 = − + , 𝐴2 = +
𝑎1 𝑤 + 1 𝑎1 − 𝐿 𝐿 − 𝑎2 𝑎 2 𝑤 + 1

in the 𝐿̂𝐼𝑠 change case, and

𝑏𝜂
𝐿̂𝑠 = 𝐿̂𝐼 ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂 𝑢
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂
𝐿̂𝑢 = 𝐿̂𝐼 (16)
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂 𝑢

in the 𝐿̂𝐼𝑢 change case. The other effects are written as

1 −𝜇1 −𝜇2
𝑤
̂= 𝑝̂ , 𝑎̂1 = 𝑝̂1 , 𝑎̂2 = 𝑝̂ ,
|𝜃| |𝜃| |𝜃|
(𝜆2,𝑢 𝑏𝜂 + 𝜆2,𝑠 𝜂) 𝑝̂
𝑦̂1 = {−𝜁1,1 𝜇1 − 𝜁1,2 𝜇2 + } ,
|𝜆| |𝜃|
𝜆1,𝑢 𝑏𝜂 + 𝜆1,𝑠 𝜂 𝑝̂
𝑦̂2 = {−𝜁2,1 𝜇1 − 𝜁2,2 𝜇2 − } .
|𝜆| |𝜃|

In (15) and (16), the final effects are less than the initial effects because an increase in the
skilled labor increases the output using skilled labor intensively, which reduces the price
of the skilled labor intensive goods and the factor price of the skilled labor, and this leads
to a decrease in the skilled labor.

2. 5. 3 Effect of Technological Changes

We turn our attention here to technological changes. Assume that a technological change
̂ 𝐼 and 𝑎̂𝑖𝐼 . Then the economy has to
happens and that the initial effects of the change are 𝑤
satisfy the equations:

𝑦̂ + 𝜇𝑢 𝑝̂ = 𝜓, (17)
𝑏𝜂 + 𝜂 𝐼
𝜓 = −𝐴𝑖 𝑎̂𝑖𝐼 + 𝐴𝑗 𝜇𝑗 𝑤
̂𝐼 + ̂ , (𝑖, 𝑗) = (1, 2), (2, 1).
𝑤
|𝜆|

From (6), (7), (8), (9), (10), (11), (13), and (17), the total effects on the outputs and the
product prices are written as

𝜇𝑢 −1
𝑦̂ = 𝜓, 𝑝̂ = 𝜓.
𝜇𝑢 + 𝜙 𝜇𝑢 + 𝜙

The relative factor price, the ratios between the optimal amounts of skilled and unskilled
labor, and the amounts of skilled and unskilled labor are written as

̂ 𝐼 − 𝐴𝑖 |𝜆|𝑎̂𝑖𝐼
𝐴𝑖 |𝜆|𝜇𝑖 𝑤
𝑤
̂= ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
68
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

(𝐴𝑗 |𝜆|𝜇𝑗 + 𝑏𝜂 + 𝜂)𝑎̂𝑖𝐼 + (𝐴𝑗 |𝜆|𝜇𝑗 + 𝑏𝜂 + 𝜂)𝜇𝑖 𝑤


̂𝐼
𝑎̂𝑖 = ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
−𝐴𝑖 |𝜆|𝜇𝑗 𝑎̂𝑖𝐼 − 𝐴𝑖 |𝜆|𝜇𝑖 𝜇𝑗 𝑤̂𝐼
𝑎̂𝑗 = ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
𝑏𝜂𝐴𝑖 |𝜆|𝜇𝑖 𝑤̂ 𝐼 + 𝑏𝜂𝐴𝑖 𝑎̂𝑖𝐼
̂𝐿𝑠 = ,
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂
−𝜂𝐴𝑖 |𝜆|𝜇𝑖 𝑤̂ 𝐼 − 𝜂𝐴𝑖 𝑎̂𝑖𝐼
̂𝐿𝑢 = .
𝐴1 |𝜆|𝜇1 + 𝐴2 |𝜆|𝜇2 + 𝑏𝜂 + 𝜂

The final effects of the initial effect part on the relative factor price, the ratios between the
optimal amounts of skilled and unskilled labor, and the amounts of skilled and unskilled
labor are less than the initial effects, respectively.

2. 6 Specialization

In the following, we consider specialization in this paper's models. Assume that there are
two countries, two factors of production, skilled labor and unskilled labor, and two goods,
skilled labor intensive goods and unskilled labor intensive goods.6 In the Heckscher-Ohlin
model, when the difference in the relative endowment between two countries is
sufficiently large, each country specializes in what each country is suited for. The skilled
labor abundant country mainly specializes in goods using skilled labor intensively, while
the unskilled labor abundant country mainly specializes in goods using unskilled labor
intensively.

Suppose that endowments are given in a functional form, that one of two countries is
completely skilled labor abundant, and that the difference in the relative endowment
between the two countries is sufficiently large. As in the Heckscher-Ohlin model, one of
following three cases occurs. First, the completely skilled labor abundant country
specializes in goods using skilled labor intensively, and the other in goods using unskilled
labor intensively. Second, the completely skilled labor abundant country produces both
goods using skilled labor intensively and goods using unskilled labor intensively, and the
other specializes in goods using unskilled labor intensively. Third, the completely skilled
labor abundant country specializes in goods using skilled labor intensively, and the other
produces both goods using unskilled labor intensively and goods using skilled labor
intensively. Because, as in the Heckscher-Ohlin model, considering the Production
Possibility Frontier (PPF) of the countries, given the relative product price, the relative
output of skilled labor intensive goods is larger in the skilled labor abundant country than
in the unskilled labor abundant country. As explained before, a completely skilled
(unskilled) labor abundant country means that the country is skilled labor abundant at
any relative factor price.

When the specialization of one or both countries occurs, the relative factor prices differ
between countries. The relative price of skilled labor in the completely skilled labor
abundant country is lower than in the completely unskilled labor abundant country. The
relative price of unskilled labor in the completely skilled labor abundant country is higher
69
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

than in the completely unskilled labor abundant country. This happens intensively
because the technologies in each country change to technologies under which production
uses the respective abundant factor more to use endowments fully. This leads to changes
in the relative factor prices.

When the relative factor prices change, factor endowments also change. The lower
relative factor price of skilled labor in the completely skilled labor abundant country
relative to the completely unskilled abundant country means that, in the completely
skilled labor abundant country, more competent workers are used as a unskilled worker.
In other words, to assign more competent workers as unskilled labor, the relative price of
unskilled labor has to increase. This explains why the next proposition occurs.

Proposition 2. 7

When specialization occurs, the relative price of skilled (unskilled) labor in the completely
skilled (unskilled) labor abundant country is lower than in the other. In the completely
skilled labor abundant country, more competent workers are used as unskilled labor relative
to the completely unskilled labor abundant country.

Thus, the income differentials are larger in the completely unskilled labor abundant
country than in the completely skilled labor abundant country.

In addition, when a completely unskilled abundant country is catching up with a


completely skilled labor abundant country in education -- in other words, the amount of
skilled labor increases relative to the amount of unskilled labor --the relative price of
skilled labor in the completely skilled labor abundant country increases, and the relative
price of skilled labor in the completely unskilled labor abundant country drops. Thus,
catching up in education in developing countries widens the income differentials of
developed countries.

2. 7 Summary and Effects of Functional Form Endowments

In this subsection, we summarize and analyze what we found. In the comparative statics
for a small economy, we found results similar to the Stolper-Samuelson Theorem and the
Rybcynski Theorem. The differences are that, when one of the factor prices increases, the
endowment whose factor price increases also increases. This means that an increase in
the product price using skilled labor intensively increases the endowment in skilled labor
by increasing the factor price of skilled labor and decreasing the factor price of unskilled
labor. As a result, the increase in the product price using skilled labor intensively
increases the output using skilled labor intensively more than in the Heckscher-Ohlin
models.

Generally, functional form endowments increase the extent of changes in the outputs
when the relative product price or the relative factor price move by a certain percentage.
Two substitutional effects, one of which is between the products and the other between
the factors, also increase the extent of changes in the outputs when the relative product
70
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

price or the relative factor price moves by a certain percentage. Of course, changes in
product prices have a magnified effect on the factor prices. As a result, a change in the
product prices has a large effect on the outputs, and this effect in this paper's model is
larger than in the basic Heckscher-Ohlin model.

The magnified effects of the functional form endowments, the two substitutional effect
and the effect of changes in the product prices on the factor prices means that, in the
comparative statics for a small open economy, a change in the product prices or the
production technologies has a large effect on the outputs. Therefore, a change in the
product prices or production technologies causes greater specialization.

Conversely, in the comparative statics for a whole economy, a change in product prices
has a large effect on the outputs, which means that a change in consumer taste, the
endowment function, or the production technologies has a smaller effect on the product
prices because, if the magnified effects have a larger effect on the outputs, even a small
change in the product prices facilitates the achievement of the equilibrium conditions.
Thus, in the comparative statics for a whole economy, the magnified effects do not
increase the extent of the changes in the product prices but decrease the extent of the
changes in the product prices.

As for functional form endowments, as explained before, this increases the extent of
change in the outputs for a small open economy and decreases the extent of change in the
product prices for a whole economy. For a small and whole economy, the functional form
endowments increase the extent of specialization in production in the economy. As 𝜂
increases, which means that a smaller change in the factor prices has a larger effect on the
endowments, the extent of change in the outputs for a small open economy increases
given a change in the product prices and the production technologies. For a whole
economy, as 𝜂 increases, the extent of change in the product prices decreases given a
change in consumer taste, the production technologies, and the endowments. As 𝜂
increases, more and more effects caused by a given change in consumer taste and so on
are absorbed into a change in the outputs. As a result, as 𝜂 approaches infinity, all effects
except the initial effects will disappear for a whole economy. This means that, in cases that
have large 𝜂, the features of the Heckscher-Ohlin model do not hold much. Thus, when
analyzing the economy, one must take into consideration how much the endowments
change given a change in the factor prices.

In an extreme case, when the elasticity of substitution between the factors of production
is infinite, the relative factor price is constant. Assuming that the infinite elasticity of
substitution between the factors and that each production needs only each one of the two
factors, such as 𝑎1,𝑠 > 0, 𝑎1,𝑢 = 0, 𝑎2,𝑠 = 0, and 𝑎2,𝑢 > 0, this is similar to the Richarian
model.7 Thus, by controlling the parameters, this paper's model can derive many different
solutions and analyses.
71
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

3. Implications

Invention, technological change, innovation, and other factors change the structure of
production, how to produce goods, which goods are produced, and what type of factors
are used. These changes affect how each factor is utilized in an industry, how much the
factors are compensated, how to divide the whole production into labor, capital, land, and
so on, and then the distribution between members of the economy. However, every
invention, technological change, innovation and so on is not the same. One technological
change might alter what workers do dramatically but not change the distribution of
earnings. Conversely, another change might has a little effect on the economy but create
workers who earns a lot of money.

One of the reasons is that, as this paper has explained, there are differences in the extent
to which one factor is substituted for another. When the elasticity of substitute in a factor
is large, a small change in the relative factor price compensates for a given technological
change. Conversely, when the elasticity of substitute in a factor is very small, even a small
increase in a factor supply leads to a very large increase in the factor price. For instance,
since the early nineteenth century, we have experienced a lot of changes in the structure
of the economy. New technologies and innovations have created new products, and the
shares of each good produced in the economy have been changing. The share of
agricultural products in the economy has been decreasing, the share of industrial
products had been increased until the mid-twentieth century, and the share of education
has been increasing steadily. Within agricultural and industrial products, what is
produced has been changing. Furthermore, what kind of job workers do has been
changing, too.

However, every change does not have an effect on relative wages in the same way. This is
because what jobs require from workers is similar between jobs in some cases but not in
others. When jobs are similar, workers are interchangeable between jobs, but when jobs
are not similar, new types of workers are needed. As a result, in some cases, the
emergence of an industry created high earners. For example, in the sports industry, Floyd
Mayweather earned 85 million dollars, Manny Pacquiao earned 62 million dollars, and
Tiger Woods earned 59.4 million dollars (Forbes 2012). Even though the size of the sports
industry is not very large, the sports industry creates millionaires.

Now, we turn our attention to a more specific topic. Income inequity within developed
countries has increased since the early 1980s. From the mid-1980s to the late-2000s,
inequality rose in 15 out of 19 countries (OECD 2013). Among the causes of this
phenomenon, skill biased technological change, outsourcing, and computer technologies
have been discussed much and relate to the Heckscher-Ohlin model. Feenstra and Hanson
(1999) found that computers explained about 35 percent of the increase in the relative
wage of nonproduction workers (compared to that of production workers), while
outsourcing explained 15 percent.8 Cheung and Fan (2002) and Fan and Cheung (2004)
analyzed the effects of trade between Hong Kong and China on the relative wage between
skilled and unskilled workers.
72
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

When analyzing income differences between workers, using the Heckscher-Ohlin model
poses some problems. The Heckscher-Ohlin model assumes that the factors of production
are completely distinctive and given by fixed amounts. Even though nonproduction
workers might have more productive ability than production workers, workers assigned
to production work could be used in nonproduction work in a less effective way. Thus, if
the relative wage between nonproduction and production workers changes, the
reallocation of some production workers to nonproduction work might be optimal and
increase the total value of outputs. This reallocation has to offset the initial effects of skill
biased technological change, outsourcing, and computer technologies to an extent. Thus,
using the Heckscher-Ohlin model under the assumption that each factor is completely
different has a problem.

There are three elements in the relationship between factor endowments and factor
prices. First, technologies determine the structure of an economy. Thus, a technological
change affects how goods are produced and how outputs are distributed between factors.
Second, the elasticities of substitutes between factors determines how much each
technological change affects factor prices and the extent to which substitution between
factors occurs. Finally, changes in endowments, whether in a functional form or not, affect
factor prices, product prices, and others. Thus, skill biased technological change,
outsourcing, and computer technologies cannot entirely determine factor prices.

Particularly in Japan and Europe, there is an excess supply of educated workers.


McGuinness (2006) examined over-education in developed countries and concluded that
the evidence on wages is in line with assignment theory, not human capital theory nor job
competition model, given the lower returns on surplus education and the over-education
penalty. However, since there are overeducated workers, there should be some kind of
over-supply of highly educated workers. This means that, when international competition
reduces the wages of unskilled workers, indirect competition between skilled workers
and unskilled but educated workers reduces the wages of skilled workers, too. Thus,
competition from low wage countries does not necessarily widen inequality. As a result,
the more general assumption under which skilled workers and unskilled workers are not
distinctive probably produces more realistic results.

Developed countries have unemployment problems and difficulty in finding


employment.9This might mean that the market is imperfect, and the market imperfection
widens the wage differences because if market works efficiently, the existence of
unemployed or overqualified workers reduces the wages of workers who could be
replaced with others.

As Pritchett (2001) said, education might be like piracy. People with more education get
hired as skilled workers and have higher wages. At the same time, Pritchett found that the
cross-national data suggested negative externalities of education. This means that it is
rational for individuals to get an advanced degree in education, but it is not for nations.
This might suggest that market imperfection causes the wages of unskilled workers to
drop and the wages of skilled worker remain relatively stable, even though some of the
unskilled workers have the ability to work as skilled workers. If so, the larger wage
73
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

differentials between skilled workers and unskilled workers impair efficiency and reduce
output. This is only a hypothesis, but the above facts and the model in this paper suggest
that this might be the case. Thus, further research is needed.

4. Conclusion

In this paper, I attempt to extend the Heckscher-Ohlin model and introduce an assignment
problem between two factors in the model. In the original Heckscher-Ohlin model,
endowments are given by fixed amounts. In the model proposed in this paper,
endowments are given in a functional form and are, to an extent, interchangeable. This
extension enables us to examine trade patterns, factor prices, and product prices more
thoroughly. In section 2, I construct the model and analyze it using comparative statics.
According to the model, if both countries produce both goods, the factor prices are
equalized across countries, as in the Heckscher-Ohlin model. In comparative statics for a
small open economy, if endowments are given in a functional form, then when the relative
price of a good increases, the output of the good increases more than when endowments
are given by fixed amounts. In comparative statics for a whole economy, if endowments
are given in a functional form, then when a change in consumer tastes or production
technologies occurs, the change in outputs is less than when endowments are given by
fixed amounts. In specialization, the completely skilled labor abundant country uses more
competent workers as unskilled labor, whose wages are higher than in the completely
unskilled labor abundant country. These results partially explain why observers have
seen many patterns in which technological changes affect the structure of the economy.

Endnotes

* The auther is a Ph.D. graduate student in the Osaka City University. I would like to
express my gratitude to Prof. Nakajima and an anonymous referee. Address: Osaka City
University, Osaka, Sumiyoshi, Sugimoto. E-mail: [email protected].
1. Feenstra (2004) is a good textbook for graduate students. The textbook thoroughly
explains and examines empirical and theoretical works of the Heckscher-Ohlin model.
2. We use the dual approach here. See Jones (1965).
∂c
3. From the first part to the second part, the envelope theorem is used, ∂wi = 𝑎𝑖,𝑠 +
s
𝜕𝑎𝑖,𝑠 𝜕𝑎𝑖,𝑢
(𝑤𝑠 + 𝑤𝑢 ) = 𝑎𝑖,𝑠 .
𝜕𝑤𝑠 𝜕𝑤𝑠
4. Here, we use the envelop theorem, too.
5. In the comparative statics for a whole economy, the product prices and the factor prices
are not determined uniquely because, as long as the relative prices are the equilibrium
prices, the product prices can move freely.
6. Here, we restrict our discussion to cases of two goods, but the result can be applied to
cases of more than two goods.
7. However, comparative statics cannot be done in this paper's method, because the
Richadian model has corner solutions.
8. Here, Feenstra and Hanson use nonproduction workers as a proxy for more skilled
workers, and production workers as a proxy for less skilled workers.
9. For example, Pedersen et al. (2011) in Europe and Ohta et al. (2008) in Japan.
74
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

References

Bowen, H. P., E. E. Leamer, and L. Sveikauskas. 1987. “Multicountry Multifactor Tests


of the Factor Abundance Theory.” American Economic Review, 77, 791-809.

Chen, Z. 1995. “The Heckscher-Ohlin Theorem with Endogenous Labour Supply.” Bulletin
of Economic Research, 47(4), 275-283.

Cheung, K. Y. and C. S. Fan. 2002. “Does trade lead to wage inequality? A cross-industry
analysis.” Journal of the Asia Pacific Economy, 7(2), 147-159.

Costrell, R. M. and G. C. Loury. 2004. “Distribution of Ability and Earnings in a


Hierarchical Job Assignment Model.” Journal of Political Economy, 112(6), 1322-1363.

Fan, C. S. and K. Y. Cheung. 2004. “Trade and wage inequality: The Hong Kong case.”
Pacific Economic Review, 9(2), 131-142.

Feenstra, R. C. 2004. ``Advanced International Trade.'' Princeton, Princeton University


Press.

Feenstra, R. C. and G. H. Hanson. 1999. “The Impact of Outsourcing and High-technology


Capital on Wages: Estimates for the United States, 1979-1990.” Quarterly Journal of
Economics, 114, 907-40.

Forbes 2012. Available at: http://www.forbes.com/athletes/list/ (Accessed: 14 April


2013).

Grossman, G. M. and G. Maggi. 2000. “Diversity and Trade.” American Economic Review,
90(5):1255-75.

Heckscher, E. and B. Ohlin. 1991. “Heckscher-Ohlin Trade Theory.” Edited by Ed. Harry
Flam and M. June Flanders, Cambridge, MIT Press.

Jones, R. W. 1965. “The Structure of Simple General Equilibrium Models.” Journal of


Political Economy, 73, 557-72.

Leamer, E. E. 1980. “Leontief Paradox, Reconsidered.” Journal of Political Economy, 88,


495-503.

Leamer, E. E. 1999. “Effort, Wages, and the International Division of Labor.” Journal of
Political Economy, 107(6), 1127-62.

Leontief, W. W. 1953. “Domestic Production and Foreign Trade: The American Capital
Position Re-examined.” Proceedings of the American Philosophical Society, 97, 332-49.
75
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

McGuinness, S. 2006. “Overeducation in the Labour Market.” Journal of Economic Surveys,


20(3), 387-418.

Martin, J. P. 1976. “Variable Factor Supplies and the Heckscher-Ohlin-Samuelson Model.”


The Economic Journal, 86(344), 820-831.

OECD. 2013. “Income inequality in OECD Factbook 2013: Economic, Environmental and
Social Statistics.” OECD Publishing.

Ohnsorge, F. and D. Trefler. 2007. “Sorting It Out: International Trade with


Heterogeneous Workers.” Journal of Political Economy, 115(5), 868-92.

Ohta, S., Y. Genda, and A. Kondo. 2008. “The Endless Ice Age.” Japanese Economy, 35(3),
55-86.

Pedersen, P. J. and T. D. Schmidt. 2011. “What is Unemployment in Europe? Concepts


and Measurement in the European Community Household Panel.” Empirical Economics,
40, 705-28.

Pritchett, L. 2001. “Where Has All the Education Gone.” The World Bank Economic Review,
15(3), 367-91.

Richardo, D. 1817. “On the Principles of Political Economy and Taxation.” London, John
Murray.

Roy, A. D. 1951. “Some Thoughts on the Distribution of Earnings.” Oxford Economic Papers,
3, 135-46.

Rybczynski, T. N. 1955. “Factor Endowments and Relative Commodity Prices.”


Economica, 22, 336-41.

Sasaki, T. 2012. “Flexible Capital-labor Assignment Model.” Economics Bulletin, 32(3),


2320-25.

Sattinger, M. 1979. “Differential Rents and the Distribution of Earnings.” Oxford Economic
Papers, 31(1), 60-71.

Stolper, W. F. and P. A. Samuelson. 1941. “Protection and Real Wages.” Review of


Economic Studies, 9, 58-73.

Tervio, M. 2008. “The Difference That CEOs Make: An Assignment Model Approach.”
American Economic Review, 98(3), 642-68.

Teulings, C. N. 1995. “The Wage Distribution in a Model of the Assignment of Skills to


Jobs.” Journal of Political Economy, 105(2), 280-315.
76
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Teulings, C. N. 2005. “Comparative Advantage, Relative Wages, and Accumulation of


Human Capital.” Journal of Political Economy, 113(2), 425-461.

Tinbergen, J. 1951. “Some Remarks on the Distribution of Labour Incomes.” International


Economic Papers, 1, 195-207.

Tinbergen, J. 1956. “On the Theory of Income Distribution.” Weltwirtschaftliches Archiv,


77(2), 155-173.

Tinbergen, J. 1970. “A Positive and a Normative Theory of Income Distribution.” Review


of Income and Wealth, 16(3), 221-34.

Trefler, D. 1993. “International Factor Price Differences: Leontief Was Right!” Journal of
Political Economy, 101, 961-87.

Vanek, J. 1968. “The Factor Proportions Theory: The N-Factor Case.” Kyklos, 21, 749-54.
77
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Figure 1. Heckscher-Ohlin Model


78
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Figure 2. Generalized Case


79
Sasaki, International Journal of Applied Economics, 12(1), March 2015, 51-79

Figure 3. Worker’s Wages

You might also like