South Korea’s Record Wage Rates: Labor in

Late Industrialization

ALICE H. AMSDEN”

Despite “unlimited labor” and political repression, real average wages in

South Korea have risen rapidly, possibly faster than in any previous
industrialization-including Japan’s-or in any contemporay one. This
study suggests that the process of industrializing late through learning,
which has been especially intense in Korea, coupled with that country’s
special institutions, underlie the spectacular wage increases. The analysis
also highlights the unique skill-set of late industrialization that high wages
reward.

COUNTRIES THAT HAVE INDUSTRIALIZED “LATE” (by borrowing

technology rather than by innovating) have tended to do so in the presence
of extraordinary hostility toward unions and “unlimited” labor reserves.
Korea is no exception, yet its average real wage rates have soared since the
late sixties. In part, such increases reflect the policies of a populistic military
dictatorship which has favored higher wages for production workers and a
narrowing of the wage gap between production workers and managers. In
part, such increases reflect the need of Korean companies to industrialize
by means of “learning” (borrowing technology from more economically
advanced countries), the emphasis of this paper.
By characterizing labor market behavior in Korea against the background
of earlier and contemporary industrializations, this article raises several issues
of general interest. First, why real wages rose in Korea amid what appears
to have been abundant labor reserves cannot be resolved single-handedly
with the economists’ traditional supply and demand tools. In part, Korean
wages have risen rapidly because learning in Korea has been especially
intense; higher wages are a reward to workers who exercise their intelligence
at the workplace to facilitate technology transfer. Although labor economists

* Department of Economics, New School for Social Research.

Vol. 29, No. 1 (Winter 1990). 0 1990 Regents of the University of California
INDUSTRIALRELATIONS,
00 19/86761901220177/$10.00

77
78 / ALICE H.AMSDEN

have developed the concept of “implicit contracts” that involve higher wages
in return for greater effort and “loyalty” to the firm, this is an ad hoc
addition to traditional theory. It also remains to be shown, as this paper
attempts to do, precisely what the nature is of the effort being rewarded.
Second, Korea’s wage phenomena bring to the fore the issue of labor market
segmentation. Segmentation arose in Korea in what by all accounts is an
almost “perfect” labor market. The explanation for segmentation highlights
the nature of the skill set associated with late industrialization in general.
Third, Japanese industrialization, which has aroused intense interest because
of the competition it poses in world markets, is easier to understand when
it is coupled with Korean industrialization, both comprehended in the
context of late industrialization through learning.
A final caveat remains. This article is concerned with drawing relationships
over long time periods and across countries, and the attempt to do both
simultaneously makes it difficult to employ econometric methods of analysis.
Although the presumed determinants of Korea’s rapid wage increases are
discussed, no attempt is made to assign weights to their respective
significance.

The General Paradigm of Late Industrialization

Industrialization in the eighteenth century in Britain may be said to have
occurred on the basis of invention, or the search, largely through trial and
error, for better products and processes in the context of the small-scale
firm. Industrialization in the nineteenth century in Germany and the United
States may be said to have occurred on the basis of innovation, or the mass
commercialization of invention through a process involving systematic
problem solving, often in the context of the R&D laboratory of the modern
industrial enterprise. While these two modes of industrialization differ, if
only in their degree of scientific content and firm organization, they share
one key characteristic: the generation of new technology by leading
companies. Developing new technology colored the way in which leading
firms from the U.K . , U. S . , and Germany realized productivity improvements
and competed for market share.
By contrast, late industrializers have generated no new technology. The
mode of late industrialization has been one of learning, or borrowing
technology from more economically advanced societies. Industrializing on
the basis of learning rather than inventing or innovating involves a different
form of competition, a different model of productivity improvement, and,

See Amsden (1989) for a fuller treatment of issues raised in this paper.
 South Korea’s Record Wage Rates I 79
in general, different characteristics of firm behavior and business-government
relations from those of the past. These differences arise from the fact that
leading firms do not have the competitive advantage of novel technology.
The attributes of late industrialization appear in countries as disparate as
South Korea, Taiwan, Brazil, Mexico, Turkey, India, and Japan (in many
respects).
Perhaps the most striking difference among latecomers is their economic
performance. All are bound by having industrialized through learning, but
their growth rates have differed sharply. Japan, South Korea, and Taiwan
have tended to grow faster than the others for a more sustained time period.
The typical explanation for these three countries’ success is that each has
conformed more to the free market model than other late learnem2 Such
an explanation is not persuasive, however. Governments in all late
industrializing countries have used subsidies to get relative prices “wrong”
in order to stimulate investment. What has differed are the principles
governing these allocations. Whereas the subsidy was typically dispensed as
a giveaway in Turkey and Mexico, for example, it was dispensed according
to the principle of reciprocity in Korea and Japan, in exchange for transparent
performance standards. This lent a measure of discipline to the process of
accumulation. As for the labor market, although abundant labor reserves
and political repression have characterized almost all late industrialization,
Korea and Japan have distinguished themselves in two respects. First, in
Korea in particular, average real wages have risen faster, possibly, than in
any previous or contemporary industrial revolution. Second, while wage
segmentation by gender and manufacturing industry is common to all late
industrializations, segmentation has taken an extreme form in Korea and
Japan.
Rapidly rising real wages from a very low wage base have allowed firms
in Korea and Japan to reap high returns while driving the more prudent
among them to invest in greater technological capability in the future. And,
in both countries, a segmented work force has had the virtue (whatever its
vices) of providing labor-intensive businesses with a cheap source of labor
and capital-intensive businesses with a labor aristocracy that is motivated.

The Role of Labor Supply

Countries that have industrialized since World War I1 have done so on
the basis of large labor reserves, or what Lewis (1954) calls un unlimited
supply. Greater labor reserves may be inferred from differences in population

* For the case of Korea, see Kim and Roemer (1979).

80 I’ ALICE H. AMSDEN

growth rates, international migratory flows, and obstacles to labor organization

(one cannot infer surplus labor from unemployment rates because these are
statistically unreliable).

Population. According to Kuznets’ (1966) estimates for countries that are

currently industrialized, the growth rate of population from the mid-
seventeen hundreds and onwards was rarely more than 1.5 per By
contrast, in developing regions, except for Japan, the population growth
rate was significantly higher, and it grew steadily.‘ By the sixties, population
growth rates averaged 2.4 per cent in “low income” countries, 2.5 per cent
in “middle income” countries (which include Korea and other late-
industrializing states), and 1.O per cent in industrialized ones (World Bank,
1979).
Given Korea’s high population density and inequitable income distribution,
a scarcity of labor never existed, even in the mines and plantations of the
Japanese colonialists. During the colonial period, the labor market behaved
as one would expect; between 1910 and 1940, real wages fell (Grajdanzev,
1944). Further evidence of excess reserves came during World War I1 when
the Colonial Office drafted 2,000,000 Koreans for wage employment in
Japan’s least remunerative industries. Although land reform in the late
forties was labor-absorbing, by 1960 it was estimated that perhaps one-fifth
of the industrial and agricultural work force in Korea, two million people,
were unemployed (Reeve, 1963).

International migration. By comparison with countries that industrialized

relatively ear1.y in world history, Japan’s industrialization marked a turning
point after which industrializing countries experienced faster population
growth rates coupled with far fewer opportunities to send their surplus
populations abroad. Since the end of World War 11, capital has become
more mobile internationally, while labor has become far less so. An estimate
of European migration during the period 1851-1920, when the population
of all of Europe was far smaller than that of the developing world today,
puts the number at more than 40 million (Woodruff, 1973). Fewer than
half a million Koreans emigrated permanently in the two decades following
1962.’
The United States was an exception; its population grew much faster because of inward
migration.
From 1869 to 1940, the Japanese population is estimated to have grown at only about 1
per cent per annum, and after the war, it grew by only about 1.1 per cent (Patrick and
Rosovsky, 1976).
Temporary migration to the Middle East was heavy in the seventies, however. See S. K.
Kim (1982).
 South Korea’s Record Wage Rates I 81
Obstacles to labor organization. Countries that have industrialized since
World War I1 generally have tended to do so in the absence of a coherent
union movement or of a parliamentary labor party. Whereas skilled labor
in earlier industrializing countries was at the vanguard of worker protest
movements and trade union organization, a property of late industrialization
is a relatively small role for the same type of manual skilled labor in any
capacity-be it labor leader or scarce resource.
The part played by the artisan in late industrialization has varied by
country. In Japan, the role of artisans resembled more that of early
industrializers than that of late ones (Sumiya, 1974). In India, industries
that could use modern technology did so by way of import substitution, not
the destruction of craft production (the latter had expired much earlier). In
Korea, a craft tradition was weak, and those traditional skills that did
develop-furniture manufacturing, tailoring, and tanning, for example-
began to be undertaken in the “informal” sector. In the late seventies, the
informal sector was estimated to be about one-half to two-thirds the size of
the nonagricultural sector (Bai, 1982; Lindauer, 1984). No labor leadership
emerged from it, however.

Rising Real Wages

Despite repression, unlimited labor supply, the absence of international
migration, and a weak grouping of skilled workers, real wage rates in Korea
have soared, whereas they rose only desultorily in other late-industrializing
countries. Table 1 compares real wages in Korea, Brazil, Argentina, Mexico,
Turkey, India, and Taiwan during 1970-1984. The seventies were a period
of rising real wages in all of these countries, but in Korea wages rose
spectacularly: the index of real wages increased from a base of 100 in 1970
to 238 in 1979.6
Korea’s wage behavior appears to be unrivaled in earlier industrial
revolutions, although Japan’s is almost as impressive. For example, while it
took English workers 70 years to raise their real earnings by roughly 150
per cent, Korean manufacturing workers achieved a comparable gain in
about 20 years (from 1955 to 1976). In just one decade, 1969-1979, real
wages in Korea rose by more than 250 per cent (see Lindauer [1974] for
data on Korean wages; Lindert and Williamson [1983] for data on British
pay rates). Comparing the long-run trend in real earnings of workers in the

Industrialization, of course, began before 1970 in all of these countries. Wage increases
in Korea occurred faster than in India, Brazil, or Taiwan during the earlier period, as well.
Taking 100 as the base in 1957 (1959 in Taiwan), the wage index in 1969 reached 77 in
Brazil, 123 in India, and 156 in Korea (Amsden, 1989).
82 / ALlCE H. AMSDEN

TABLE 1
REALNONAGRICULTURAL
WAGE INCREASES IN SELECTED DEVELOPING
COUNTRIES, 1970-1984“

Year Koreah Brazil‘ Argentina Mexico Turkey India” Taiwan

1970 100 100 100 100 100 100 -

1971 102 110 105 103 100 100 -
1972 104 114 99 104 99 - 100
1973 119 119 107 104 98 106 107
1974 130 119 126 107 96 97 98
1975 131 127 124 114 116 110 110
1976 154 129 80 123 122 120 126
1977 187 134 76 125 146 116 138
1978 219 142 77 122 147 124 151
1979 238 134 87 121 155 130 163
1980 227 130 100 116 124 166
1981 225 118 91 119 130 171
1982 24 1 115 79 117 129 180
1983 26 1 97 97 86 130 188
1984 276 84 112 83 111 191

Base = 100; deflated by consumer price index.

Real earnings manufacturing sector.
Average wages for skilled workers in construction. Data are from the Central Bank.
Rupees per hour for industrial workers.
Source: As cited in Amsden (1989).

United States and Korea is equally revealing. Taking the United States’
index at its trough and peak-approximately 45 in 1865, just after the Civil
War, and about 130 in 1913, just before World War I-real wages increased
in the U.S. by 2.9 in roughly 50 years. In Korea, real wages increased even
faster, from trough to peak by 4.3 in roughly half as many years (Lindauer,
1984).
Japan’s best wage data are divided into two subperiods, 1900-1935 and
1950-1968, and refer to production workers in manufacturing establishments
of 30 or more production workers (Minami, 1973). Ignoring gender
differences, real wages for all production workers rose by a multiple of 3.6
in 35 years. Between 1949 and 1968, real wages rose by a factor of 2.7.
Again, however, real wages in Korea rose even faster-by a scalar of 4.3
between 1955 and 1980 and by one of 3.6 between 1966 and 1980 (Lindauer,
1984). Wage rates in Korea, moreover, are understated in comparison with
those in Japan. Whereas Korean data are restricted to firms with over 10
workers, Japanese data are restricted to firms with over 30. Yet smaller
firms in Japan are believed to have begun paying lower wages than larger
firms after the twenties (Minami, 1973).
Indirect labor, which includes managers, engineers , and technicians, is
 South Korea’s Record Wage Rates I 83
TABLE 2
IN WAGERATESBETWEEN
CHANGES PRODUCTION
AND PROFESSIONAL,
AND MANAGERIAL
TECHNICAL, WORKERS
Year Production Professional, technical,
workers and managerial workers

1965- 1970 12.8 6.6

1971-1 974 7.1 6.1
1975- 1979 16.8 15.3
198@-1984 5.3 2.5

Source: 1965-1979, Bai (1982); 1980-1984, MOL (Korean Ministry of Labor).

supposedly the scarcest type of labor in backward countries. Thus, one

would expect late industrializers to experience abundance and scarcity of
direct and indirect labor, respectively, with the wages of the latter pulling
up the overall average. The pattern of wage increases in Korea, however,
is unexpected. As Table 2 indicates, in four subperiods from 1965 to 1984,
the rate of wage increase was higher for production workers than it was for
professional, technical, or managerial employees, notwithstanding excess
supply, supposedly most characteristic of the direct labor market.
In the late seventies, Korea allegedly reached a “turning point” (Bai,
1982), its supply of surplus labor vanishing, with the outflow of workers-
mainly to the Middle East-amounting to as much as 27 per cent of the
male manufacturing work force (MOL, various years). Table 2 confirms
that wages increased in the late seventies at an especially fast rate.
Nevertheless, such scarcity proved short-lived with the collapse of the Middle
Eastern construction market and a slowdown in Korea’s growth rate. Thus,
in the 20 years spanning 1965 to 1984, a tight labor market was the exception
to the rule, yet real wages rose persistently.

The Demand Side

If Korea experienced faster real wage increases than did other late-
industrializing countries, this is partly because its GNP grew faster. A rapid
rate of capital accumulation provided the basis for rapid wage increases.
Nam (1980) estimates an equation to explain the rate of change of nominal
manufacturing wages in Korea during the period 1966-1977. He includes
variables to represent the demand for labor (the rate of GNP over the
previous year), the supply of labor (the lagged change in manufacturing
employment), and inflation (the lagged change in prices). The strongest
84 / ALICE H. AMSDEN

explanatory variable is the demand for labor, or the previous year’s change
in GNP.
The period in question, 1966-1977, was the Golden Age of Korea’s
industrial expansion. Output, exports, and productivity were all growing
very rapidly. Not surprisingly, wages also rose. Employers were under strong
pressure from the government to share their gains with labor. The rate of
real wage increase also tended to stay below the rate of labor productivity
growth (except in 1976-1979 and 1982) (Amsden, 1989). The wage strategy
of big business amounted to gainsharing.
Nevertheless, the impact of fast growth on real wage increases is
ambiguous.’ In Brazil, for example, it seems that the impact is weak. The
average annual growth rate of the Brazilian economy during the period
1950-1975 was as much as 6.7 per cent. Yet, while real wages increased for
the majority of the period, they did so very slowly (Amsden, 1989). Taiwan’s
economy grew about as fast as Korea’s, but its wages grew more slowly (see
Table 1).

The Supply Side

Several factors seem to account for the more rapid rise in real wages in
Korea than in other late-industrializing countries. These include: the low
base from which wages in Korea have risen; the structure of Korean
agriculture-a consequence of labor reform in the late forties; the unrivaled
length of the work week; and market segmentation-by gender, firm size,
and industry. Many of these same factors, moreover, were also operative in
Japan during its period of rapid wage gains. The factors operating in both
countries also go well beyond any obvious “cultural” explanation.

Korea’s low wage base. Korea’s wages grew fast, but they started from an
exceptionally low base. The significance of the extreme poverty characteristic
of Far Eastern economies in the sixties was three-fold (Smith, 1967). First,
“the backwardness of these countries-even by the standards of developing
regions-means that the industrial labor force tends to be unorganized and
in no position to force wage rises when the cost of living increases ...” (ibid.,
p. 27). Second, because of the small size of the industrial sector, industrial

’ Nam’s equation fails to explain why wages rose, notwithstanding the inclusion of a variable
intended to measure labor supply. According to Nam (1980, p. 79), “the fit of the wage
equation to the data is not excellent,” although a similar model provides better correlation
with the inclusion of a dummy variable for the period 1970-1977. Norton and Rhee (1980)
estimate a wage equation with a dummy variable (GNP) and a price deflator (but no supply
effect) and get an R2 of 0.88.
 South Korea’s Record Wage Rates i 85
expansion does not place excessive demands on agricultural output. Third,
the marginal product of labor in agriculture tends to be low because of high
population density, so rural-urban migration does not reduce agricultural
output. Thus, from a low wage base in Korea, it was easier for employers
to remain internationally competitive and yet raise wages, if only to improve
the physical stamina and psychic motivation of their workers.

Land reform and small-scale agriculture. Rapid increases in productivity in

Korean agriculture acted to drive up wages in modern industry. The
“implicit wage” of male workers in agriculture exceeded the average wage in
manufacturing until the late sixties (Ban et al., 1980). According to Bai’s
estimates (1982), the male wage rate in agriculture exceeded that in modern
industry until at least the mid-seventies.
Korean agriculture employed few proletarians (hence, the “implicitness”
of the agricultural wage rate). Instead, the agrarian reform of the late forties
invested land in the tiller, so a tenure system of small-scale, family farms
evolved, which required few hired hands except at harvest. Given sharp
increases in agricultural productivity as a consequence of government
support, rural-urban migration and downward pressure on manufacturing
wages can be assumed to have been less massive than it would otherwise
have been. The labor retentiveness of Korean agriculture by comparison
with less egalitarian land tenure systems is suggested by international
comparisons of the share of agricultural employment in total employment.
Despite the limited availability of arable land, Korean agriculture’s share of
the labor force in 1980 was as high as 34 per cent, compared with 30 per
cent in land-rich Brazil, for example. The Korean share was even higher
than that in Taiwan, although both countries had experienced land reforms
(World Bank, 1982). Consequently, both a lower-than-otherwise flow of
rural-urban migrants and a higher-than-otherwise rural household income
buttressed manufacturing wage rates.

Women workers. Nonetheless, Korea industrialized (and continues to do

so) with unlimited labor reserves. In explaining how agriculture may preserve
a society of family farms yet provide industry with enough labor to keep
wages at a socioeconomic subsistence level, Lewis (1954, p. 404) emphasized
the importance of female workers, observing that “The transfer of women’s
work from the household to commercial employment is one of the most
notable features of economic development .”
Not only has Korea set world records with its growth rate in wages, it
has also outcompeted other countries in its discrimination against women
workers, although in some years (1984, for example), this dubious distinction
86 / ALICE H. AMSDEN

fell to Japan. In 1980, the male-female wage gap was greater in Korea than
in any other country for which data are available (ILO, 1981).
Almost 60 per cent of the male-female wage gap in Korea has been
attributed to gender differences in human capital. It is estimated that 27.6
per cent of the gap is due to differences in education and 3 1.8 per cent to
differences in experience (traditionally, women have been forced to leave
paid employment when they marry). These two sources-“before labor
market” discrimination and discrimination in the labor market “are closely
interrelated” (Lee, 1983, p. 67). There are virtually no women in managerial
or even entrepreneurial positions in Korea’s primary manufacturing sector
(Grootaert, 1986).
The human capital explanation for the gender wage gap would lead one
to expect discrimination to recede as women invested in more education and
acquired more experience. According to Y. H. Kim (1986, p. 4), this is not
the case in Korea:

During the past few decades, gender inequality in educational opportunities

at all levels of schooling has . .. continuously decreased.. .. The participation
in the labor force of those with secondary and tertiary education increased
at a much faster rate for women than for men with [a] decreasing proportion
of primary school graduates.. .. Moreover, the proportion of women in [the]
25-29 age group, who have the burden of childbearing and childrearing,
has increased, which suggests improved women’s labor-force attachment.. ..
In contrast, the male-female earnings gap remains large, without reduction.

The length of the work week. In the length of its work week, Korea has
set still another world record. Even allowing for rapid increases in demand,
Korea’s work week length exceeds expectation and has underwritten large
increases in wage rates. For all persons employed in manufacturing in 1984,
73 per cent of men and 62 per cent of women worked at least 54 hours per
week (Grootaert, 1986). In other countries, the average for persons employed
in manufacturing was much lower (ILO, various years).
The length of Korea’s work week can be understood historically. The
Japanese colonial factory system was harsh, and Korean workers grew
accustomed to a work week similar to what prevailed in Japan and Germany
at the turn of the century. For example, “men in large-scale enterprises in
1939 [in Korea] worked on an average of 10 hours a day, and women and
children 10 hours and 15 minutes, and 10 hours and 20 minutes, respectively”
(Gradjdanzev, 1944, p. 184). What is extraordinary about Korea is that long
work hours have continued to persist well after the Second World War.
This may be understood in terms of both organized labor’s limited ability
 South Korea’s Record Wage Rates I 87
TABLE 3
PER CENT OF WAGEINCREASESIN SELECTEDSECTORSOF
MANUFACTURING”
(1965-197 1)
Sector Increase of
1971 over 1965

Food 336%
Textile 227%
Wood and cork 222%
Paper 288%
Leather 215%
Rubber 175%
Chemicals 354%
Petroleum & coal products 415%
Basic metals 222%
Metal products 296%
Machinery 208%
Electric machinery 363%
Transportation equipment 3 13%
Average 269%

Suingsan Yunbo (“Annual

Calculated from figures published in S U ? I ~ P
Report on Current Industrial Production Survey”), Seoul: Economlc
Planning Board, 1972, pp. 68-81.
Source: Ewing (1973).

to shorten hours for a given wage rate and labor’s willingness to take
advantage of rapid industrial expansion through higher wages for endless
toil.

Segmentation. Upward pressure on wage rates came especially from large-

scale firms in basic industry. Table 3 shows the origins of Korea’s segmented
labor market. The rate of wage increase during the labor-abundant period
1965 to 1971 varied according to industry, ranging from a low of 175 per
cent in rubber products to a high of 415 per cent in petroleum and coal
products. The upshot was a large dispersion in manufacturing wages.

Wage dispersion. Yet another world record that Korea holds is the degree
of wage dispersion in the manufacturing sector. In both 1973 and 1982, the
standard deviation of wages (in logs) was greater in Korea than in 13 other
countries examined (Krueger and Summers, 1986). Japan took second place.
Nor has such manufacturing wage dispersion become more compressed in
Korea over time (Richardson and Kim, 1986). The capital-intensive industries
have continued to pay more than textiles, wood, and miscellaneous
88 / ALICE H. AMSDEN

manufacturing. Wage dispersion also varies across firm size, especially when
disaggregated by gender.
The dispersion in interindustry wages is largely accounted for by blue-
collar workers (Lee, 1983). In part, blue-collar workers earn different wage
rates in different industries because they have different marginal products,
with those in the capital-intensive industries probably being higher. However,
it is unclear why capital-intensive firms pay a wage rate equal to their
workers’ marginal product rather than paying the all-manufacturing average.
Wage dispersion by industry and by firm size is present in most countries
that are now industrialized or that are in the process of becoming so. In the
former, segmentation arose in response to a complex of factors, including
union pressures (see Osterman [ 19841 for references). In Korea, union
pressures were weak but political pressures were strong. The Korean
government routinely intervened in industrial relations, in an effort to
prevent a resurgence of the political unrest and labor militancy that it
repressed at the time of the Korean War.8 In addition, big business, for
efficiency reasons, began to pay at higher levels. These reasons were more
evident in a country like Korea than in Taiwan because the bureaucratic,
big corporation was more predominant. Hence, wages have tended to rise
faster in Korea.

The Skills of the New Labor Aristocracy

The rise of segmented labor markets in a late-industrializing country like
Korea is best understood in the context of the labor policies of the modern
industrial enterprise-in Korea’s case, the diversified business group.
Diversified business groups may have paid their blue-collar workers above
the all-manufacturing average because the skills of such workers were in
scarce supply since they tended to be firm-specific. In the case of male
workers, one year of “inside” experience (with the same employer) tended
to raise wages on average by about 10 per cent, whereas one year of “outside”
experience (with a different employer) raised them on average by only about
3.8 per cent (Lee, 1983). The interesting question then becomes: What
firm-specific “skills” were being rewarded?
One must start by understanding differentiation among firms. Experience
in textiles dated back to the Japanese colonial period, so expansions and
modernizations in the fifties exposed textiles enterpreneurs to few problems.

In the sixties, labor affairs were placed under the jurisdiction of the Korea Central
Intelligence Agency.
 South Korea’s Record Wage Rates I 89
By contrast, even the huge, fixed-capital investments in the continuous-
process industries that relied initially on turnkey technology transfers-
cement, fertilizers, oil refining, and steel-created problems for managers,
of trying to bring under control a process that the foreign experts themselves
understood only imperfectly. Korean managers could never hope to manage
in a tight, “Taylorist,” top-down fashion, at least not initially, because no
one at the top knew enough about the process to do so. Under these
conditions, it was imperative to rely upon motivated workers, even if they
possessed little more than formal schooling, to exercise the most fundamental
skill of all-intelligence. In all of the new capital-intensive industries-
continuous process and, a fortiori, fabrication-assembly operations and job
shops of jumbo proportions-production workers were motivated with
relatively high wage rates, first, to get the product out the door; and later,
to improve quality. In short, they were motivated to adopt a reasonably
scientific approach to problem solving in a milieu whose technology was
tacit, implicit, and not yet procedurized.
To identify the skills of the new labor aristocracy more clearly, the
following section examines the personnel policies of one of Korea’s premier
producers, the state-owned Pohang Iron and Steel Company (POSCO).

Personnel Policies at POSCO

The internal wage structure of production workers at POSCO reflects the
external one at the national level. Production workers are divided into high-
paid and low-paid categories that are analogous to those of permanent and
transitory employees in Japan. POSCO managers are the highest paid; next
comes a sub-class of blue-collar workers, who are further subdivided into
two categories: “regular” and “contracted out.” In 1984, contracted-out
workers numbered 8,700, or 24 per cent of a total work force of 25,700.
Contracted-out workers are confined to menial tasks such as relining,
cleaning, packing, preparing ingot molds, scarfing, and treating slabs.
POSCO is estimated to save about 15 per cent in wages by contracting this
work rather than delegating it to regular employees (PaineWebber, 1985).
Regular employees are well-paid by Korean standards-about twice the
manufacturing average, depending on the annual bonus. Nevertheless,
POSCO claims that it has never faced a labor shortage.
In exchange for relatively high pay and the job security that comes with
employment in a company that is expanding rapidly, POSCO expects
attentiveness from its workers throughout their long hours on the job. A
century ago, steel-making required strenuous physical effort on the part of
production workers, as well as proficiency in the art of steelmaking on the
90 / ALICE H. AMSDEN

part of master steel-makers. The category of master steel-maker or “saint

technician” still survives in POSCO, but largely as a vestige of the past.
After a decade of operation, only one master technician and five submasters
exist. Instead, steel- or iron-making for production workers involves process
monitoring and control, as well as auxiliary operations such as overhead
crane driving. POSCO has 450 job categories, and the largest number of
workers can be found at data-collection stations positioned at well-defined
points in the process. Workers check sensors for temperatures in different
process zones, note the chemical composition of gases, and register flow
rates. For this, however, they must have a fairly good understanding of the
physical and chemical reactions involved in iron- and steel-making-in order
to ensure high quality-since steel production is not all in closed-loop
control, and the acceptable limits of materials change.
To increase productivity, POSCO provides its workers with extensive
training. In a single year, 1984, 9,924 workers out of 23,700 received some
form of training (Amsden, 1989). The aim is to instill in all workers a
general knowledge of POSCO’s operations and operating principles (25 per
cent of total training). Only to a lesser extent is training oriented toward
building in-depth specialized skills. Emphasis is placed on quality control
(17 per cent of total training), foreign languages (6 per cent), and training
abroad (10 per cent).
A normal work week at POSCO consists of eight hours per day, seven
days per week: 45 regular hours, and 11 overtime hours for which no
premium is paid. Every worker is entitled to only one day of leave per
month, plus one day for each year with the company. In 1977, labor turnover
at POSCO was 4.4 per cent; the national average in manufacturing was
approximately 5.1 per cent. In 1984, labor turnover had fallen to only 1.2
per cent, well below the national average of approximately 5.4 per cent
(MOL, 1985). Absenteeism is not a problem at the company, representing
only 0.07 per cent of labor days in 1984. Low absenteeism and turnover are
believed to be essential for high productivity. Although the overwhelming
majority of POSCO’s 23,700 workers cannot be described as possessing
highly developed craft skills, steel-making remains something of an art and
learning its technology makes in-house experience invaluable.
POSCO operates with the motto, “Resources are limited, but human
creativity is unlimited.” Limited resources, or scarcity, therefore, does
appear to have influenced POSCO’s wage policy by driving it to pay higher
than average in order to attract the “best” workers. “Best” should be
understood in a general sense, however, rather than with reference to specific
skills.
 South Korea’s Record Wage Rates I 91
Conclusion
Employers in most late industrializing countries have had to pay a work
force enough to induce it to exercise its intelligence, rather than any scarce
specific manual skills, in order to import foreign technology effectively. That
wages have risen especially rapidly in Korea, despite “unlimited” labor and
political and union repression, may be attributed to particularly intense
learning and special institutions, both of which also existed in Japan and
help to explain that country’s rapid rise in wages during learning-intensive
growth.
Korea’s special institutions gave rise to a very low wage level even by
backward country standards, allowing wages to rise rapidly without impinging
on competitiveness; a productive small-scale agriculture which exerted
upward pressure on urban wages; extraordinarily long hours of work that
compensated for higher annual labor costs; and extreme segmentation of the
labor market by gender and by manufacturing branch that has sustained
simultaneously the competitiveness of labor- and capital-intensive industries.
The intensity of learning in Korea (and Japan) arose from two sources.
First, Korea relied less than almost any other late industrializing country
on direct foreign investment as a source of technology transfer (Amsden,
1989). It behooved Korean-owned firms, therefore, to learn in order to
survive. Second, industrialization occurred in more of a spurt in Korea than
it did in India, Brazil, or Turkey, whose big businesses have a longer history
than do Korea’s. In Korea, it was necessary not only to accumulate experience
rapidly, but also to create organizations out of thin air; and this required
buying commitment on the part of both managers and labor.
It is noteworthy, however, that “implicit contracts” of higher wages in
return for greater effort and “loyalty” notwithstanding, Korea’s labor
aristocracy and disadvantaged workers both have engaged in massive strike
activity since 1987. The persistence of a low level of wages, along with
political and union repression, have offset 20 years of higher wage increases
as motivating factors, suggesting that de Tocqueville’s question of whether
good times or bad times stimulate social protest remains difficult to answer.
REFERENCES
Amsden, Alice H. Asia’s Next Giant: South Korea and Late Industrialization. New York and
Oxford: Oxford University Press, 1989.
Bai, Moo-Ki. “The Turning Point in .the Korea Economy,” The Developing Economics, XX
(1982)’ 117-140.
Ban, Sung Hwan, Pal Yong Moon, and Dwight H. Perkins. Rural Development. Cambridge,
MA: Harvard University Press [for the Council on East Asian Studies, Harvard
University], 1980.
Ewing, George Ogle. Labor Unions in Rapid Economic Development: Case of the Republic of
Korea in the 1960s. Unpublished Ph.D. dissertation, University of Wisconsin, 1973.
92 / ALICE H. AMSDEN
Grajdanzev, A. Modern Korea. New York: John Day [for the Institute of Pacific Relations],
1944.
Grootaert, Christian. “The Labour Force Participation of Women in the Republic of Korea:
Evolution and Policy Issues,” mimeographed paper, World Bank, Washington, D.C.,
1986.
International Labour Office (ILO). Yearbook of Labor Statistics. Geneva: various years.
Kim, Kwang-Suk and Michael Roemer. Growth and Structural Transformation. Harvard East
Asian Monograph. Cambridge, MA: Harvard University Press, 1979.
Kim, Sooyong. “Contract Migration in the Republic of Korea.” Working Paper MIG WP
4, International Labour Office, International Migration for Employment Branch,
Geneva, Switzerland, 1982.
Kim, Young-Hwa. Education and Male-Female Earnings Inequuliry in the Structured Labor
Markets: A Case Study of Korea. Unpublished Ph.D. dissertation, Stanford University,
1986.
Krueger, Alan B. and Lawrence H. Summers. “Reflections on the Inter-Industry Wage
Structure.” Discussion Paper No. 1252. Cambridge, MA: Harvard Institute of Economic
Research, 1986.
Kuznets, Simon. Modern Economic Growth. New Haven, CT: Yale University Press, 1966.
Lee, joung-Woo. Economic Development and Wage Inequality in South Korea. Unpublished
Ph.D. dissertation, Harvard University, 1983.
Lewis, W. Arthur. “Economic Development with Unlimited Supplies of Labour.” In A. N.
Agarwala and S. P. Singh, eds., The Economics of Underdevelopment. London: Oxford
University Press, 1958, pp. 400-449.
Lindauer, David L. Labor Market Behavior in the Republic of Korea: A n Analysis of Wages
and Their Impact on the Economy. World Bank Staff Working Paper No. 641.
Washington, DC: The World Bank, 1984.
Lindert, Peter and Jeffrey G. Williamson. “English Workers’ Living Standards During the
Industrial Revolution: A New Look,” Economic History Review, XXXVI (1983), 1-25.
Minami, Ryoshin. The Turning Point in Economic Development: Japan’s Experience. Economic
Research Series No. 14. Tokyo: Institute of Economic Research, Hitotshubashi
University, 1973.
MOL [Ministry of Labor]. Yearbook of Labour Statistics. Seoul: various years.
Nam, Sang-Woo. “The Dynamics of Inflation.” In Chong Kee Park, ed., Macroeconomic and
Industrial Development in Korea. Vol. 3 of Essays on the Korean Economy. Seoul:
Korea Development Institute, 1980, pp. 55-89.
Norton, Roger D. and Seung Yoon Rhee. “A Macroeconomic Model of Inflation and
Growth.” In C. K. Park, ed., Macroeconomic and Industrial Development in Korea. Vol.
3 of Essays on the Korean Economy. Seoul: Korea Development Institute, 1980, pp.
3-54.
Osterman, Paul, ed. Internal Labor Markets. Cambridge, MA: MIT Press, 1984.
PaineWebber. I’OSCO: Korea’s Emerging Steel Giant. World Steel Dynamics Core Report.
New York: 1985.
Patrick, Hugh and Henry Rosovsky. “Japan’s Economic Performance: An Overview.” In H .
Patrick and H. Rosovsky, eds., Asia’s New Giant: How the Japanese Economy Works.
Washington, DC: Brookings, 1976, pp. 1-62.
Reeve, W. D. The Republic of Korea: A Political and Economic Study. London: Oxford
University Press, 1963.
Richardson, Ray and Byung Whan Kim. The Structure of Labor Markets in LDCs: Overview
for South Korea, 1953-1 984. Discussion paper, report No. DRD 162. Washington,
DC: The World Bank, Development Research Department, Economics and Research
Staff, 1986.
Smith, Anthony D., ed. Wage Policy Issues in Economic Development. London: Macmillan,
1967.
 South Korea’s Record Wage Rates I 93
Sumiya, Mikio. “The Emergence of Modern Japan.” In Kazuo Okochi, Bernard Karsh, and
Solomon B. Levine, eds., Workers and Employers in Japan. Princeton, NJ: Princeton
University Press, 1974.
Woodruff, William. “The Emergence of an International Economy, 1700-1914.” In Carlo
M. Cipolla, ed., The Emergence of Industrial Societies, Vol. 2 of The Fontanta Economic
History of Europe. Fontma, CA: Collins, 1973, pp. 656-737.
World Bank. World Development Report. Washington, DC: various years.