Effects of International Trade Between Two-Factor Economies

Effects of International Trade
Between Two-Factor Economies
Having outlined the production structure of a two-factor economy, we can now look at what
happens when two such economies, Home and Foreign, trade. As always, Home and Foreign
are similar along many dimensions. They have the same tastes and therefore have identical
QC 2 QC 1
QF 2 QF 1
Output of
food, QF
Output of
cloth, QC
slope = –PC
/PF
QF
2
QF
1
2
slope = –PC
/PF
1
TT1
TT 2
QC
Q 2 C
1
Figure 5-8
Resources and Production
Possibilities
An increase in the supply of
labor shifts the economy’s
production possibility frontier
outward from to , but
does so disproportionately in
the direction of cloth production.
The result is that at an unchanged
relative price of cloth (indicated
by the slope ), food
production actually declines
from to QF.
Q 2 F
1
-PC/PF
TT1 TT2
6The biased effect of resource changes on production was pointed out in a paper by the Polish economist T. M.
Rybczynski, “Factor Endowments and Relative Commodity Prices,” Economica 22 (November 1955), pp. 336–341.
It is therefore known as the Rybczynski effect.
90 PART ONE International Trade Theory
relative demands for food and cloth when faced with the same relative prices of the two
goods. They also have the same technology: A given amount of labor and capital yields the
same output of either cloth or food in the two countries. The only difference between the
countries is in their resources: Home has a higher ratio of labor to capital than Foreign does.
Relative Prices and the Pattern of Trade
Since Home has a higher ratio of labor to capital than Foreign, Home is labor-abundant
and Foreign is capital-abundant. Note that abundance is defined in terms of a ratio and not
in absolute quantities. For example, the total number of workers in the United States is
roughly three times higher than that in Mexico, but Mexico would still be considered
labor-abundant relative to the United States since the U.S. capital stock is more than three
times higher than the capital stock in Mexico. “Abundance” is always defined in relative
terms, by comparing the ratio of labor to capital in the two countries; thus no country is
abundant in everything.
Since cloth is the labor-intensive good, Home’s production possibility frontier relative
to Foreign’s is shifted out more in the direction of cloth than in the direction of food. Thus,
other things equal, Home tends to produce a higher ratio of cloth to food.
Because trade leads to a convergence of relative prices, one of the other things that will
be equal is the price of cloth relative to that of food. Because the countries differ in their
factor abundances, however, for any given ratio of the price of cloth to that of food, Home
will produce a higher ratio of cloth to food than Foreign will: Home will have a larger
relative supply of cloth. Home’s relative supply curve, then, lies to the right of Foreign’s.
The relative supply schedules of Home (RS) and Foreign ( *) are illustrated in Figure 5-9.
The relative demand curve, which we have assumed to be the same for both countries, is shown
as RD. If there were no international trade, the equilibrium for Home would be at point 1, while
the equilibrium for Foreign would be at point 3. That is, in the absence of trade the relative
price of cloth would be lower in Home than in Foreign.
When Home and Foreign trade with each other, their relative prices converge. The relative
price of cloth rises in Home and declines in Foreign, and a new world relative price of
RS
Relative price
of cloth, PC/PF
Relative quantity
of cloth, QC /QF
2
1
3
RS
RS*
RD
Figure 5-9
Trade Leads to a Convergence
of Relative Prices
In the absence of trade, Home’s
equilibrium would be at point 1,
where domestic relative supply
RS intersects the relative demand
curve RD. Similarly, Foreign’s
equilibrium would be at point 3.
Trade leads to a world relative
price that lies between the pretrade
prices, that is, at point 2.
CHAPTER 5 Resources and Trade: The Heckscher-Ohlin Model 91
cloth is established at a point somewhere between the pretrade relative prices, say at point 2.
In Chapter 4, we discussed how an economy responds to this trade opening based on the
direction of the change in the relative price of the goods: The economy exports the good
whose relative price increases. Thus, Home will export cloth (the relative price of cloth
rises in Home), while Foreign will export food. (The relative price of cloth declines in
Foreign, which means that the relative price of food rises there).
Home becomes an exporter of cloth because it is labor-abundant (relative to Foreign)
and because the production of cloth is labor-intensive (relative to food production).
Similarly, Foreign becomes an exporter of food because it is capital-abundant and because
the production of food is capital-intensive. These predictions for the pattern of trade (in
the two-good, two-factor, two-countries version that we have studied) can be generalized
as the following theorem, named after the original developers of this model of trade:
Hecksher-Ohlin Theorem: The country that is abundant in a factor exports the good
whose production is intensive in that factor.
In the more realistic case with multiple countries, factors of production, and numbers of
goods, we can generalize this result as a correlation between a country’s abundance in a
factor and its exports of goods that use that factor intensively: Countries tend to export
goods whose production is intensive in factors with which the countries are abundantly
endowed.7
Trade and the Distribution of Income
We have just discussed how trade induces a convergence of relative prices. Previously we
saw that changes in relative prices, in turn, have strong effects on the relative earnings of
labor and capital. A rise in the price of cloth raises the purchasing power of labor in terms
of both goods while lowering the purchasing power of capital in terms of both goods.
A rise in the price of food has the reverse effect. Thus international trade can have a powerful
effect on the distribution of income, even in the long run. In Home, where the relative
price of cloth rises, people who get their incomes from labor gain from trade, but
those who derive their incomes from capital are made worse off. In Foreign, where the relative
price of cloth falls, the opposite happens: Laborers are made worse off and capital
owners are made better off.
The resource of which a country has a relatively large supply (labor in Home, capital in
Foreign) is the abundant factor in that country, and the resource of which it has a relatively
small supply (capital in Home, labor in Foreign) is the scarce factor. The general conclusion
about the income distribution effects of international trade in the long run is: Owners of a
country’s abundant factors gain from trade, but owners of a country’s scarce factors lose.
This conclusion is similar to the one reached in our analysis of the case of specific factors.
There we found that factors of production that are “stuck” in an import-competing industry
lose from the opening of trade. Here we find that factors of production that are used intensively
by the import-competing industry are hurt by the opening of trade. The theoretical argument
regarding the aggregate gains from trade is identical to the specific factors case: Opening to
trade expands an economy’s consumption possibilities (see Figure 4-11), so there is a way to
make everybody better off. However, there is one crucial difference regarding the income
distribution effects in these two models. The specificity of factors to particular industries is
often only a temporary problem: Garment makers cannot become computer manufacturers
7See Alan Deardorff, “The General Validity of the Heckscher-Ohlin Theorem,” American Economic Review 72
(September 1982), pp. 683–694, for a formal derivation of this extension to multiple goods, factors, and countries.
92 PART ONE International Trade Theory
overnight, but given time the U.S. economy can shift its manufacturing employment from
declining sectors to expanding ones. Thus income distribution effects that arise because labor
and other factors of production are immobile represent a temporary, transitional problem
(which is not to say that such effects are not painful to those who lose). In contrast, effects of
trade on the distribution of income among land, labor, and capital are more or less permanent.
We will see shortly that the trade pattern of the United States suggests that compared
with the rest of the world, the United States is abundantly endowed with highly skilled
labor and that low-skilled labor is correspondingly scarce. This means that international
trade has the potential to make low-skilled workers in the United States worse off—not just
temporarily, but on a sustained basis. The negative effect of trade on low-skilled workers
poses a persistent political problem, one that cannot be remedied by policies that provide
temporary relief (such as unemployment insurance). Consequently, the potential effect of
increased trade on income inequality in advanced economies such as the United States has
been the subject of a large amount of empirical research. We review some of that evidence
in the box that follows, and conclude that trade has been, at most, a contributing factor to
the measured increases in income inequality in the United States.
Case Study
North-South Trade and Income Inequality
The distribution of wages in the United States has become considerably more unequal
since the late 1970s. In 1979, a male worker with a wage at the 90th percentile of the wage
distribution (earning more than the bottom 90 percent but less than the top 10 percent of
wage earners) earned 3.6 times the wage of a male worker at the bottom 10th percentile of
the distribution. By 2005, that worker at the 90th percentile earned more than 5.4 times the
wage of the worker at the bottom 10th percentile. Wage inequality for female workers has
increased at a similar rate over that same time-span. Much of this increase in wage
inequality was associated with a rise in the premium attached to education. In 1979, a
worker with a college degree earned 1.5 times as much as a worker with just a high school
education. By 2005, a worker with a college degree earned almost twice as much as a
worker with a high school education.
Why has wage inequality increased? Many observers attribute the change to the
growth of world trade and in particular to the growing exports of manufactured goods
from newly industrializing economies (NIEs) such as South Korea and China. Until the
1970s, trade between advanced industrial nations and less-developed economies—often
referred to as “North-South” trade because most advanced nations are still in the temperate
zone of the Northern Hemisphere—consisted overwhelmingly of an exchange of
Northern manufactures for Southern raw materials and agricultural goods, such as oil and
coffee. From 1970 onward, however, former raw material exporters increasingly began to
sell manufactured goods to high-wage countries like the United States. As we learned
in Chapter 2, developing countries have dramatically changed the kinds of goods they
export, moving away from their traditional reliance on agricultural and mineral products
to a focus on manufactured goods. While NIEs also provided a rapidly growing market
for exports from the high-wage nations, the exports of the newly industrializing
economies obviously differed greatly in factor intensity from their imports. Overwhelmingly,
NIE exports to advanced nations consisted of clothing, shoes, and other relatively
unsophisticated products (“low-tech goods”) whose production is intensive in unskilled
CHAPTER 5 Resources and Trade: The Heckscher-Ohlin Model 93
labor, while advanced-country exports to the NIEs consisted of capital- or skill-intensive
goods such as chemicals and aircraft (“high-tech goods”).
To many observers the conclusion seemed straightforward: What was happening was a
move toward factor-price equalization. Trade between advanced countries that are abundant
in capital and skill and NIEs with their abundant supply of unskilled labor was raising
the wages of highly skilled workers and lowering the wages of less-skilled workers in the
skill- and capital-abundant countries, just as the factor-proportions model predicts.
This is an argument with much more than purely academic significance. If one regards
the growing inequality of income in advanced nations as a serious problem, as many people
do, and if one also believes that growing world trade is the main cause of that problem,
it becomes difficult to maintain economists’ traditional support for free trade. (As we have
previously argued, in principle taxes and government payments can offset the effect of
trade on income distribution, but one may argue that this is unlikely to happen in practice.)
Some influential commentators have argued that advanced nations will have to restrict
their trade with low-wage countries if they want to remain basically middle-class societies.
While some economists believe that growing trade with low-wage countries has
been the main cause of rising income inequality in the United States, however, most
empirical researchers believed at the time of this writing that international trade has
been at most a contributing factor to that growth, and that the main causes lie elsewhere.
8 This skepticism rests on three main observations.
First, the factor-proportions model says that international trade affects income distribution
via a change in relative prices of goods. So if international trade was the main driving
force behind growing income inequality, there ought to be clear evidence of a rise in
the prices of skill-intensive products compared with those of unskilled-labor-intensive
goods. Studies of international price data, however, have failed to find clear evidence of
such a change in relative prices.
Second, the model predicts that relative factor prices should converge: If wages of
skilled workers are rising and those of unskilled workers are falling in the skill-abundant
country, the reverse should be happening in the labor-abundant country. Studies of
income distribution in developing countries that have opened themselves to trade have
shown that at least in some cases, the reverse is true. In Mexico, in particular, careful
studies have shown that the transformation of the country’s trade in the late 1980s—
when Mexico opened itself to imports and became a major exporter of manufactured
goods—was accompanied by rising wages for skilled workers and growing overall wage
inequality, closely paralleling developments in the United States.
Third, although trade between advanced countries and NIEs has grown rapidly, it
still constitutes only a small percentage of total spending in the advanced nations. As a
result, estimates of the “factor content” of this trade—the skilled labor exported, in
effect, by advanced countries embodied in skill-intensive exports, and the unskilled
labor, in effect, imported in labor-intensive imports—are still only a small fraction of
the total supplies of skilled and unskilled labor. This suggests that these trade flows
cannot have had a very large impact on income distribution.
8Among the important entries in the discussion of the impact of trade on income distribution have been Robert
Lawrence and Matthew Slaughter, “Trade and U.S. Wages: Giant Sucking Sound or Small Hiccup?” Brookings
Papers on Economic Activity: Microeconomic 2 (1993), pp. 161–226; Jeffrey D. Sachs and Howard Shatz, “Trade
and Jobs in U.S. Manufacturing,” Brookings Papers on Economic Activity 1 (1994), pp. 1–84; and Adrian Wood,
North-South Trade, Employment, and Income Inequality (Oxford: Oxford University Press, 1994). For a survey of
this debate and related issues, see Robert Lawrence, Single World, Divided Nations?: International Trade and
OECD Labor Markets (Paris: OECD Development Centre, 1996).
94 PART ONE International Trade Theory
What, then, is responsible for the growing gap between skilled and unskilled workers in
the United States? The view of the majority is that the villain is not trade but rather new
production technologies that put a greater emphasis on worker skills (such as the widespread
introduction of computers and other advanced technologies in the workplace).
How can one distinguish between the effects of trade and those of technological
change on the wage gap between skilled and unskilled workers? Consider the variant of
the model we have described where skilled and unskilled labor are used to produce
“high-tech” and “low-tech” goods. Figure 5-10 shows the relative factor demands for
producers in both sectors: the ratio of skilled-unskilled workers employed as a function
of the skilled-unskilled wage ratio (LL curve for low-tech and HH for high-tech).
We have assumed that production of high-tech goods is skilled-labor intensive so the
HH curve is shifted out relative to the LL curve. In the background, there is an SS curve
(see Figure 5-7) that determines the skilled-unskilled wage ratio as an increasing function
of the relative price of high-tech goods (with respect to low-tech goods).
In panel (a), we show the case where increased trade with developing countries generates
an increase in wage inequality (the skilled-unskilled wage ratio) in those countries (via an
(a) Effects of trade (b) Effects of skill-biased technological change
Skilled-unskillled
wage ratio, wS /wU
LL
HH
wS /wU
SL/UL SH/UH
Skilledunskilled
employment,
S /U
Skilledunskilled
employment,
S /U
Skilled-unskillled
wage ratio, wS /wU
LL HH
wS /wU
SL/UL SH/UH
Figure 5-10
Increased Wage Inequality: Trade or Skill-Biased Technological Change?
The LL and HH curves show the skilled-unskilled employment ratio, , as a function of the skilled-unskilled
wage ratio, , in the low-tech and high-tech sectors. The high-tech sector is more skill-intensive than the lowtech
sector, so the HH curve is shifted out relative to the LL curve. Panel (a) shows the case where increased trade
with developing countries leads to a higher skilled-unskilled wage ratio. Producers in both sectors respond by
decreasing their relative employment of skilled workers: and both decrease. Panel (b) shows the case
where skill-biased technological change leads to a higher skilled-unskilled wage ratio. The LL and HH curves shift
out (increased relative demand for skilled workers in both sectors). However, in this case producers in both sectors
respond by increasing their relative employment of skilled workers: SL /UL and SH/UH both increase.
SL /UL SH/UH
wS/wU
S/U
CHAPTER 5 Resources and Trade: The Heckscher-Ohlin Model 95
increase in the relative price of high-tech goods). The increase in the relative cost of skilled
workers induces producers in both sectors to reduce their employment of skilled workers
relative to unskilled workers.
In panel (b), we show the case where technological change in both sectors generates
an increase in wage inequality. Such technological change is classified as “skill-biased,”
as it shifts out the relative demand for skilled workers in both sectors (both the LL and
the HH curves shift out). Then, a given relative price of high-tech goods is associated
with a higher skilled-unskilled wage ratio (the SS curve shifts). In this case, the technological
change induces producers in both sectors to increase their employment of skilled
workers relative to unskilled workers.
We can therefore examine the relative merits of the trade versus skill-biased
technological change explanations for the increase in wage inequality by looking at
the changes in the skilled-unskilled employment ratio within sectors in the United
States. A widespread increase in these employment ratios for all different kinds of
sectors (both skilled-labor-intensive and unskilled-labor-intensive sectors) in the
U.S. economy points to the skill-biased technological explanation. This is exactly
what has been observed in the U.S. over the last half-century.
In Figure 5-11, sectors are separated into four groups based on their skill intensity. U.S.
firms do not report their employment in terms of skill but use a related categorization of
0.16
0.18
0.00
Non−Production−Production Employment
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
year
Least Skill−Intensive
0.20
0.22
0.24
0.26
0.28
0.24
0.26
0.00 Non−Production−Production Employment
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
year
2nd Least Skill−Intensive
0.28
0.30
0.32
0.34
0.30
0.35
0.00
Non−Production−Production Employment
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
year
2nd Most Skill−Intensive
0.40
0.45
0.60
0.70
0.00
Non−Production−Production Employment
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
year
Most Skill−Intensive
0.80
0.90
1.00
Figure 5-11
Evolution of U.S. Non-Production–Production Employment Ratios in Four Groups of Sectors
Sectors are grouped based on their skill intensity. The non-production–production employment ratio has
increased over time in all four sector groups.
96 PART ONE International Trade Theory
production and non-production workers. With a few exceptions, non-production positions
require higher levels of education—and so we measure the skilled-unskilled employment
ratio in a sector as the ratio of non-production employment to production employment.9
Sectors with the highest non-production to production employment ratios are classified as
most skill-intensive. Each quadrant of Figure 5-11 shows the evolution of this employment
ratio over time for each group of sectors (the average employment ratio across all sectors in
the group). Although there are big differences in average skill intensity across the groups,
we clearly see that the employment ratios are increasing over time for all four groups. This
widespread increase across most sectors of the U.S. economy is one of the main pieces of
evidence pointing to the technology explanation for the increases in U.S. wage inequality.
Yet, even though most economists agree that skill-biased technological change has
occurred, recent research has uncovered some new ways in which trade has been an
indirect contributor to the associated increases in wage inequality, by accelerating this
process of technological change. These explanations are based on the principle that
firms have a choice of production methods that is influenced by openness to trade and
foreign investment. For example, some studies show that firms that begin to export also
upgrade to more skill-intensive production technologies. Trade liberalization can then
generate widespread technological change by inducing a large proportion of firms to
make such technology-upgrade choices.
Another example is related to foreign outsourcing and the liberalization of trade and
foreign investment. In particular, the NAFTA treaty (see Chapter 2) between the United
States, Canada, and Mexico has made it substantially easier for firms to move different
parts of their production processes (research and development, component production,
assembly, marketing) across different locations in North America. Because production
worker wages are substantially lower in Mexico, U.S. firms have an incentive to move to
Mexico the processes that use production workers more intensively (such as component
production and assembly). The processes that rely more intensively on higher-skilled,
non-production workers (such as research and development and marketing) tend to stay
in the United States (or Canada). From the U.S. perspective, this break-up of the production
process increases the relative demand for skilled workers and is very similar to skillbiased
technological change. One study finds that this outsourcing process from the
United States to Mexico can explain 21 to 27 percent of the increase in the wage premium
between non-production and production workers.10
Thus, some of the observed skill-biased technological change, and its effect on
increased wage inequality, can be traced back to increased openness to trade and
foreign investment. And, as we have mentioned, increases in wage inequality in
advanced economies are a genuine concern. However, the use of trade restrictions
targeted at limiting technological innovations—because those innovations favor
relatively higher-skilled workers—is particularly problematic: Those innovations
also bring substantial aggregate gains (along with the standard gains from trade) that
would then be foregone. Consequently, economists favor longer-term policies that
ease the skill-acquisition process for all workers so that the gains from the technological
innovations can be spread as widely as possible.
9On average, the wage of a non-production worker is 60% higher than that of a production worker.
10See Robert Feenstra and Gordon Hanson, “The Impact of Outsourcing and High-Technology Capital on Wages:
Estimates for the United States, 1979–1990,” Quarterly Journal of Economics 144 (August 1999), pp. 907–940.
CHAPTER 5 Resources and Trade: The Heckscher-Ohlin Model 97
Factor-Price Equalization
In the absence of trade, labor would earn less in Home than in Foreign, and capital would
earn more. Without trade, labor-abundant Home would have a lower relative price of cloth
than capital-abundant Foreign, and the difference in relative prices of goods implies an
even larger difference in the relative prices of factors.
When Home and Foreign trade, the relative prices of goods converge. This convergence,
in turn, causes convergence of the relative prices of capital and labor. Thus there is
clearly a tendency toward equalization of factor prices. How far does this tendency go?
The surprising answer is that in the model, the tendency goes all the way. International
trade leads to complete equalization of factor prices. Although Home has a higher ratio of
labor to capital than Foreign does, once they trade with each other, the wage rate and the
capital rent rate are the same in both countries. To see this, refer back to Figure 5-6, which
shows that given the prices of cloth and food, we can determine the wage rate and the
rental rate without reference to the supplies of capital and labor. If Home and Foreign face
the same relative prices of cloth and food, they will also have the same factor prices.
To understand how this equalization occurs, we have to realize that when Home and
Foreign trade with each other, more is happening than a simple exchange of goods. In an indirect
way, the two countries are in effect trading factors of production. Home lets Foreign have
the use of some of its abundant labor, not by selling the labor directly but by trading goods
produced with a high ratio of labor to capital for goods produced with a low labor-capital
ratio. The goods that Home sells require more labor to produce than the goods it receives in
return; that is, more labor is embodied in Home’s exports than in its imports. Thus Home
exports its labor, embodied in its labor-intensive exports. Conversely, since Foreign’s exports
embody more capital than its imports, Foreign is indirectly exporting its capital. When viewed
this way, it is not surprising that trade leads to equalization of the two countries’ factor prices.
Although this view of trade is simple and appealing, there is a major problem with it: In
the real world, factor prices are not equalized. For example, there is an extremely wide
range of wage rates across countries (Table 5-1). While some of these differences may
reflect differences in the quality of labor, they are too wide to be explained away on this
basis alone.
To understand why the model doesn’t give us an accurate prediction, we need to look at
its assumptions. Three assumptions crucial to the prediction of factor-price equalization
are in reality certainly untrue. These are the assumptions that (1) both countries produce
TABLE 5-1 Comparative International Wage Rates (United States = 100)
Country
Hourly Compensation
of Production Workers, 2005
United States 100
Germany 140
Japan 92
Spain 75
South Korea 57
Portugal 31
Mexico 11
China* 3
*2004
Source: Bureau of Labor Statistics, Foreign Labor Statistics Home Page.
98 PART ONE International Trade Theory
both goods; (2) technologies are the same; and (3) trade actually equalizes the prices of
goods in the two countries.
1. To derive the wage and rental rates from the prices of cloth and food in Figure 5-6,
we assumed that the country produced both goods. This need not, however, be the case.
A country with a very high ratio of labor to capital might produce only cloth, while a
country with a very high ratio of capital to labor might produce only food. This implies
that factor-price equalization occurs only if the countries involved are sufficiently similar
in their relative factor endowments. (A more thorough discussion of this point is given in
the appendix to this chapter.) Thus, factor prices need not be equalized between countries
with radically different ratios of capital to labor or of skilled to unskilled labor.
2. The proposition that trade equalizes factor prices will not hold if countries have
different technologies of production. For example, a country with superior technology
might have both a higher wage rate and a higher rental rate than a country with an
inferior technology. As described later in this chapter, recent work suggests that it is
essential to allow for such differences in technology to reconcile the factor-proportions
model with actual data on world trade.
3. Finally, the proposition of complete factor-price equalization depends on complete
convergence of the prices of goods. In the real world, prices of goods are not
fully equalized by international trade. This lack of convergence is due to both natural
barriers (such as transportation costs) and barriers to trade such as tariffs, import
quotas, and other restrictions.