Pay differentials: the case of Japan Essay

In review of industrial relations research conducted during the
1970’s, James G. Scoville writes that, in both Japan and the United
States, size-of-firm wage differentials are explained by differences in
employees’ human capital. However, two recent studies suggest that
human capital differences do not completely explain the differential in
this country. Using data for 1979, Wesley Mellow found that wages in
firms of 1,000 more workers were 8 percent greater than those in firms
with fewer than 25 workers when a number of factors, including education
and experience, were held constant. Martin E. Personick and Carl B.
Barsky, who studied pay at various experience and responisbility levels
of professional, technical, and clerical occupations, reported
size-of-firm differentials for all but 1 of 25 job levels. Typically,
these were only for the largest corporations (more than 10,000
employees), where differentials were 10 to 15 percent for professionals
and 20 percent for clerical and technical occupations over pay in firms
with 500 or fewer employees.



If elements of human capital do not completely explain size-of-firm
differentials in the United States, is Japan a similar case? This
article explores that issue, and suggests an answer based on data from
the Chingin Kozo Kihon Tokei Chosa [Wage Structure Survey]. The
employment decision in Japan



The model employment relationship in Japan is that of Shushin Koyo
]lifetime employment]. Under this system, workers are initially
employed upon graduation from school. Once a worker is hired, the firm
goes to great lengths to provide continuous employment until the
individual retires, sometime between the ages of 50 and 60. In return
for the understood employer commitment to long tenure, the employee is
expected to devote himself fully to the firm and to allow management
considerable flexibility as to the type and geographical location of
work assignments.


Remuneration consists of a basic wage, various allowances, a
semiannual bonus, and a number of fringe benefits. The basic wage
depends upon the employee’s education, age, and job abilities. It
is increased annually based upon decisions made in collective
bargaining. The annual increase consists of two parts, one of which
recognizes an additional year of service to the firm, new job abilities,
and merit, and another that is a general increase in the base wage.



Given the employment opportunities and wage patterns faced by the
graduating student, what pecuniary variable should be used in making the
employment decision? Clearly, it is some subjective assessment of the
present value of future earnings with the various firms. Such a present
value calculation would incorporate growth of the firm relative to the
economy, the pattern of wages associated with long tenure, the pattern
of wages if tenure is short because of voluntary mobility or the
firm’s economic difficulties, and so forth. For the observer
trying to approximate such individual calculations, the most desirable
data would be those on wage and bonuses by worker age, education, and
length of service, and, for the question at hand, the size of the
employing firm. Fortunately, these data are available in the annual
Wage Structure Survey. It is thus possible to account for the principal
elements of huamn capital that economists believe are important for wage
determination, and to differentiate these among three size-of-firm
categories. (Of course, the individual graduate also considers other,
unquantifiable factors, such as his preference for risk, the prestige of
the firm, and subjective probabilities of advancement, in making the
final decision.) Differentials by size of firm



Table 1 presents monthly wage and wage-plus-bonus relationships by
size of firm and by workers’ age and educational attainment for
Japanese men who have been continuously employed by the same firm. (In
1980, about one-fourth of the regular private-sector labor force were
employed by firms of 1,000 workers or more, and another one-fourth were
in firms with 100 to 999 employees.) According to the table,
compensation is generally less in the smaller companies, regardless of
worker age or education. Monthly wages are about the same in the two
smaller size classes until workers are in their forties, when those in
the mediumsize firms begin to receive more. When bonus payments are
included as compensation, the differences between the largest and
smallest firms become more dramatic. In general, the higher teh level
of education, the larger is the wage gap by size of firm.


To more fully illuminate these relationships, table 2 presents
compensation relatives by industry for broad age categories of high
school and college educated men. Ata underlying the estimates relate to
individuals whose tenure suggests theat they have been continuously
employed by the same firm since graduation. Thus, only a few of all
possible matched age-tenure pairs are shown, but these represent core
groups in the economy. Two distributions are presented, one for monthly
wages and one for monthly wages plus one-twelfth of annual bonus
payments. Again, the inclusion of bonuses tends to increase the income
differences among the three size-of-firm classes, and the benefits of
working for the larger firms increase with age and tenure.



The pay relatives suggest little in the way of systematic variation
by industry, although those for transportation and communication tend to
be quite high in samller firms while those in finance and insurance are
comparatively low. The indices of each industry’s differentials
were ranked and compared to rankings by union penetration and proportion
of total employment in large firms by industry. Neither comparison
indicated any systematic relationship with size-of-firm differentials.



Except for occupations that require substantial training–airline
pilots, construction crafts, and so forth–occupational distinctions are
weakly, if at all, correlated with wages in Japan. Hence, while table 3
shows significant occupational wage differentials by size of firm, these
results may be less meaningful than estimates based on other variables.



The data in table 2 do suggest capital investment with its own
rewards. Yet the greater opportunity to achieve long tenure which
characterizes employment in large firms should also be seen as an
additional benefit to such employment, unless the individual worker has
a positive taste for risk. New graduates are quite aware that their
prospects for long tenure with a large firm are more promising than with
a smaller firm. For example, in 1981, 19.4 percent of all college
educated men age 45 to 49 were employed in firms with 1,000 workers or
more had worked 20 or more years for their current employer. The figure
for those in firms with 100 to 999 workers was 54.5 percent, and for
firms with 10 to 99 workers, it was 31.7 percent. Earnings data for
50-to 54-year-old high school educated men suggest that workers do not
have to pay a compensation premium for the greater probability of long
tenure: Among those with 30 or more years of tenure, wages plus bonuses
in large firms are 17 percent higher than in middle-size firms and 31
percent higher than in small firms, while the comparable figures for
similary aged workers at all levels of tenure are 25 percent and 40
percent.



Employment opportunities for women, especially at high-level jobs
and with the major employers, are markedly different from those for men,
although there have been changes toward equality during the postwar years. In particular, men’s wages increas more with age: In 1981,
the 50- to 54-year-old high school educated male with 1 to 2 years of
firm tenure had a monthly wage that was 56 percent higher than that of a
similary educated 18- to 19-year-old. Among women, the worker age 50 to
54 received only 17 percent more than her younger counterpart. Yet,
firm-specific tenure appears to be relatively more valuable for older
women than for older men. This is probably because women with brief
tenure are likely to have been in the labor market for only a short
time, which is not typically the case for men. Yet, when the
compensation of high school educated workers with at least 30
years’ tenure was compared by size of firm, the patterns for men
and women were quite similar. Women’s wages plus bonus in firms
with 1,000 workers or more were 18 percent higher than in firms with 100
to 999 workers, and 26 percent higher than in small firms. Again, there
is no compensation premium paid by workers for the probability of long
tenure in larger firms: At ages 50 to 54 for all levels of tenure, wages
plus bonus for women in the largest firms were 36 percent higher than in
middle-size firms and 44 percent higher than in the smallest firms. The
puzzle


It seems clear in Japan, as in the United States, that the standard
human capital variables of education and experience do not completely
explain, if ever they did, size-of-firm differentials. In addition, it
is evident that the Japanese differential is much larger after age 40 or
when bonus income is included. Any explanation, therefore, must be
consistent with the age pattern demonstrated and the concentration of
the differential in the bonus portion of compensaton.



It is possible that a more exhaustive test of work characteristics
would reduce the size of the differential. We know, for example, that
the most able students enroll in the very best schools, from which the
larger, more successful corporations seek employees. Yet it seems
unlikely that such difficult-to measure characteristics of employees
could explain wage differentials of the magnitude shown in the tables.



Widening differential with age. Some recent studies of
compensation by age include variables for implicit contracts,
experience, risk, incentives, and so forth, that may explain the
Japanese pattern. One approach incorporating a variety of these
concepts was presented in a 1982 article by Milton Harris and Beng
Holmstrom.



According to the authors, there are four possible reasons why
compensation increases with age: a) firms learn about individual
abilities and are better able to match workers to jobs; b) workers begin
to pay employers lower implicit premiums to guarantee their ability to
do acceptable work; c) employees learn productivity-enhancing skills;
and d) pay levels are a particulary important means to motivate
employees in a world of lifetime employment security. The first two of
these, while consistent with a general widening of the wage
differentials over time, do not imply a rapid shift after the age of 40.
The second two appear to be more relevant.



In the larger firms, there is more physical capital per worker,
which could yield greater productivity, and thus justify higher wages.
It also is probable that the interaction of higher quality employees
with similar employees and with higher levels of physical capital
generates greater increases in human capital in the larger firms. The
development of productivity enhancing skills with additional tenure may
well be an important element in the ability of large firms to pay high
wages. Indeed, in the context of a technology-specific skills model,
Hong Tan has argued that such gains over a working life are key
determinants of Japanese wage patterns. A somewhat similar argument was
made by Kazuo Koike, who hypothesized that the more developed system of
internal training in large firms provides a greater range of
technologically related positions than is true in smaller firms, which
in turn contributes to wage differentials by size of firm. However,
even if enhanced skills are an important factor, there remains the
problem of timing. Why should so much of the differential be
concentrated in the years after age 40?



The last element, motivational allowances, may best explain the
time pattern. As is well known by the organizers of games of chance,
large prizes and prizes that are ever in the future seem to have
disproportionate power to motivate participants when compared to their
discounted value. Many of the new employees in major Japanese firms wll
not be there to collect their “prize” at older ages, but the
promise of greater compensation is a constant motivating factor. Thus,
the firm saves money compared to paying an annual motivational allowance
to each employee. In a sense, the firm also has received an
interest-free loan from the employee, who has tacitly agreed to defer a
portion of compensation to later worklife. In a rapidly growing
economy, such an arrangement is highly advantageous to the firm, but
even in less dynamic times an interest-free loan has value.



Japanese institutions. There are two institutional factors unique
to Japan which also have significantly affected the time pattern of the
differential and its size. The concept of a living wage based upon
family needs has long been important in Japan. It is rooted in the
nation’s history, but has become more prominent since World War, I,
and particularly since the economic difficulties of the post-World War
II years. The concept provides that wages should increase over a
worker’s life to accommodate marriage, the birth of children, the
high costs of private college, and savings for early retirement from the
primary employer. The latter two factors would suggest significant wage
increases after the age of 40.



The second institutional consideration is that the nature of the
large corporation in postwar Japan is quite different than in prewar years. Formerly, corporations were uniquely capitalistic, owned and
controlled by wealthy individuals. However, share ownership in postwar
Japan has tended to be diluted into the hands of other firms and banks.
There is a high proportion of capital in the form loans and internally
generated funds, and an almost complete absence of outside directors.
These changes, in conjunction with Japanese historical patterns and
moral visions, have persuaded many scholars that today’s large
firms are essentially collectives of employees who hire high risk-high
gain capital from shareholders and low risk-fixed gain capital from
banks. If the assumption that the Japanese corporation is a collective
of employees which hires capital rather than a collective of owners of
capital which hires workers (including senior managers) is valid, it is
hardly surprising that economic rents are shared among the members of
the collective–the employees.



The extensive use of bonus payments as the mechanism to pay out
significant portions of the higher income received by employees in large
firms is more complicated to explain. The payment of a semiannual bonus
is a very old Japanese practice which was intended to provide employees
with sufficient funds to meet the extra needs associated with certain
cultural and religious practices. The bonus also served to provide a
measure of equity and motivation in the form of profit sharing.
However, with the democratization of employment in the postwar years, a
significant bonus, which to an extent had been reserved for white-collar
and management employees, was extended to all workers.



While extensively used by all Japanese employers, the bonus tends
to be relatively larger in the larger firms, while smaller firms compete
for labor on the basis of regular monthly wages. The emphasis small
firms give to wages as opposed to bonuses seems to be attributable to
two factors: First, the firm wants to provide a monthly wage to cover
the necessities of life, and second, a somewhat less rosy employment
future gives any “promised” bonus made by a smaller employer
less value than an equivalent promise by a large employer.
Consequently, one would expect that smaller firms would first meet
competitive levels in monthly wages, and only later meet those of the
bonus. Patterns over time



There is no simple measure of the degree of wage difference by size
of firm because the wage ratios between alternative matched pairs do not
all move together. To describe movements over time, I chose to examine
wages for 35- to 39-year-old male high school graduates with 15 to 19
years of tenure who worked as production workers in manufacturing (table
4). In addition, data on wage dispersion are provided for selected
years (table 5). According to table 4, size-of-firm differentials that
were quite wide in 1955 closed somewhat, reaching near equality in 1964.
The 1960’s were a period of generally tightening differentials as
the labor market became much more competitive, and the productivity
levels of small firms approached those of large firms. After 1967, the
differential gradually widened until a second period of near equality
occurred during the oil-shock years 1973-74. This second narrowing was
undoubtedly related to inflation, for employment growth in manufacturing
had leveled off, turning negative by 1972.



The estimates in table 4, which have been standardized for
industry, general type of work, age sex, education, and firm tenure,
suggest that size-of-firm wage differentials have remained relatively
constant since 1975. However, the figures in table 5, which exclude
bonuses and include data for workers at all levels of education and
years of firm tenure, show a continuing narrowing of the dispersion of
wages within the three size classes. Both tables imply that there has
been a greater narrowing of differences between firms of 10 to 99
employees and those with 100 to 999 employees than between the latter
and firms of 1,000 and more employees. A comparison with the United
States



Recent estimates of size-of-firm differentials in the United
States, cited earlier, permit some limited comparison. Wesley
Mellow’s estimate of an 8-percent pay advantage in firms of 1,000
workers of more over firms with fewer than 25 employees appears
relatively modest compared to most of the differentials for Japanese men
shown in tables 1 and 2. In the United States, as in Japan, the
large-firm differential was greater when specific firm tenure was not
considered, and the differential existed across all major industries,
although the U.S. diffierential appeared to be greater in manufacturing
than in nomanufacturing.



Personick and Barsky’s study of professional, technical, and
clerical occupations revealed as typical 10- to 15-percent differentials
for professionals and a 20-percent gap for clerical and technical
occupations between firms of 10,000 or more employees and those with 500
or fewer employees. Although these estimates are for for quite
different firm-size classes, they do approximate the differentials
reported in table 1 for younger Japanese high school and college
graduaes, but they are smaller than those for older college educated
males. Interestingly, the U.S. size-firm differential seemed to be
larger for workers with less than a college education. Also, the U.S.
differentials were larger for entry-level positions than for higher
levels of experience. Again, this is the opposite of the Japanese case,
in which differentials widen at older ages. These differences between
the two countries are consistent with a situation in which large firms
pay above-market prices in oder to pick and choose among applicants
whose employment potential has not yet been established, but in which
one economy embraces the norm of continuous tenure from graduation while
the other anticpates considerable interfirm mobility at younger ages.