How the japanese learn to work 2nd edition - part 3 ppt

Who goes where? 25
useful supplement to their pensions. The test marks were standardized on
the whole 14-year-old population of the prefecture in a generally understood
fashion—10 marks above or below 50 for each standard deviation—so that
only 5 per cent came above 70 or below 30. Since children’s effort input
tended to be fairly uniform, there was a relative stability in each child’s
scores—his or her hensachi range, or range of ‘standard deviation scores’,
tended to be relatively narrow. Since the success of last year’s pupils with
hensachi rating x getting into high school y was carefully documented by
the teachers—the scores provided a good basis for advising pupils on which
choice of high school to make. So much was this process the focus of parental
and teacher concern that ‘hensachi education’ became the boo-word par
excellence, summing up all that was stifling, uncreative and anti-educational
in the school system.
The growth of such criticism led, in the early 1990s, to a ban on the use of
such commercial tests in schools. According to the teachers of the middle
school senior classes, who used pupils’ records on such tests to advise them
as to which high school they should apply to, this has simply added to their
work. They still have to give advice; that advice still has to be given on the
basis of children’s prospective performance, relative to other children, in the
high schools’ competitive entrance examinations. The only difference is that
now they have to make up mock tests themselves, which means extra work
but no extra pay (although one suspects that the money which used to be
spent on buying the commercial tests is somehow channelled into teachers’
allowances). Increasingly, they are devising tests collectively—all the schools
in a town or wider area getting together—and in at least ten prefectures this
is already organized on a prefectural basis by the local education authorities.
(The prefectures have formal jurisdiction and budget for (public) high schools.

Primary and junior secondary schools are administered by the municipal/
rural district tier of government.) The only difference is a cosmetic one; only
raw scores on the tests are used; normalization and the calculation of a
‘standard deviation score’ are studiously avoided. So who can call it ‘hensachi
education’ now?
And meanwhile, of course, an estimated 50 per cent of third-year middle
school students—the 50 per cent who, or the parents of whom, are most
concerned about which high school they get to—are going to private after-
school juku where they still take the commercial tests, and still have their
standard deviation scores measured. It is highly unlikely that they refrain
from reporting these scores to their class teachers when the decision about
which high school they should seek to enter is being made.
The actual operation of the examination and selection system shows
considerable local variety; in some prefectures with large school districts—
a single catchment pool for eight or more schools, say—all the middle school
26 How the Japanes learn to work
teachers in charge of final year classes spend a horse-trading weekend where,
in the light of the range of scores of their charges they hammer out quotas,
first of all for the top school, in such a way as to equalize the scores of the
marginal lowest-scoring pupil from each school. Then they do the same for
the second-best school, and so on down the line. Over the next weeks they
advise their pupils and their pupils’ parents accordingly. An alternative is to
allow initial tentative applications, then for each high school to publish
statistics of applicant numbers, thus allowing the weaker candidates to be
withdrawn from over-subscribed schools, before definitive applications are
sent in.
How this works in practice in a rural area with a relatively low rate of
progression to university may be illustrated by the situation in Iwaki city in
Fukushima prefecture, which is the subject of Table 2.3. The details of the
following description relate to the mid–1980s, but there is no evidence that

the general picture has subsequently changed, except that the passage of the
second baby-boom—22 per cent fewer middle school pupils in 1994
compared with the mid–1980s peak, has left some of the lowest-ranking
schools short of pupils. The city constitutes a bigger-than-average catchment
area, where the ranking of high schools is unambiguous. At what one might
call Level 1, there is a single top public school for boys and one for girls
which between them take about 17 per cent of the age group. (Some of the
northern prefectures—unlike any in central and western Japan—still partially
hold out against co-education.) These two schools owe their pre-eminence
to the fact that they were the only pre-war selective secondary schools in the
district. Almost all of the boys and the majority of the girls who attend them
go on subsequently to college or university.
At Level 2, there is a third general academic-course high school which
takes both boys and girls and from which about a third of the leavers seek to,
and manage to, get into (lesser) universities. Also on Level 2 are two of the
vocational schools—an industrial technical school which takes predominantly
boys, and a commercial vocational school which takes predominantly girls.
They rank as of approximately equal prestige with the general-course high
school just mentioned, and are in fact slightly more difficult to get into—in
the case of the Data Processing Course at the technical school, much more
difficult. A third technical school in another part of the district has a slightly
lower entrance level.
Then, at Level 3 come three general high schools which divide up the
next slice of the age group (roughly from the 50th to the 80th percentile

of the ability range) primarily on geographical lines, and below them—Level 4—two
general schools which try to fill their places by a ‘second round recruitment’
— scooping up those who failed to get into their first-choice school higher
up the pecking order. Finally, at the bottom of the public school heap are two
vocational schools—one for agriculture and one for fisheries.

There are no elite university-preparatory private schools within commuting
distance of this area—a major difference from the situation in areas where
(at a rough guess) some 80 per cent of the population lives. There are just
two private high schools. They take children who cannot get into any public
high school even at the second-round recruitment, or children who could get
into a low-ranking public school, but whose parents would prefer them to be
educated with the children of other parents who can afford private school
Who goes where? 29
fees. This syndrome is stronger in the case of girls—what they learn is less
important for their life chances/marriage chances than who they learn it with.
The girls’ private school, in consequence, may have pupils who could have
got into Level 3 schools.
The only further complication: there is one school with all-prefecture
recruitment—the five-year College of Technology—in the prefectural capital
at some considerable distance and hence for boarders only as far as Iwaki is
concerned. The small number of Iwaki pupils who go there would have had
a good chance (in the case of those entering the machinery and construction
courses) or an almost certain chance (in the case of those entering the more
popular electricity or applied chemistry courses) of getting into the Level 1
high schools. But they are unlikely to be in that top five per cent of scorers
whose teachers would tell them they had a good chance of going on from
high school to a good national university.
It is obvious that it makes little sense to compare the average abilities of
general-course students and vocational-course students: some general courses
take the brightest, some the least bright children. One might summarize
Iwaki’s ability-distribution system as follows:

For a child diagnosed as being in the top 10 per cent of the ability
distribution, it will be difficult to resist the pressure of teachers’ urgings
that ‘of course’ they should go to one of the top high schools and on to

university—unless something in their connections or back-ground
makes the College of Technology alternative more attractive. (The
‘ability labelling effect’ makes the top schools attractive even for those
who might not be able to afford to go to a university.)
For pupils who fall in the 10th to the 20th percentile, there is a
choice between:
— Aiming at the top school, and being prepared to take a year in a
cramming school in order to get in at a second attempt.
— For those who cannot afford that, aiming at the top school and
risking relegation to Level 3 or 4 if the bid fails.
— Aiming at one of the Level 2 schools—the general school or one
of the 3 vocational schools—with near certainty of success.
30 How the Japanes learn to work
For those who fall in the 20th to the 50th percentile range (a rather
narrow range in performance terms) application to a top school is a
big risk. They are naturals for the Level 2 schools and provide the
bulk of the entrants to the vocational schools.
As to what determines choice as between general and vocational courses
in this ability range, there are few careful studies, it being, apparently, very
difficult for Japanese educational sociologists to do any work correlating
performance with social class characteristics except at the broadest ecological
level. (See e.g. Hata 1975–6.) Clearly important are: sex (it is more important
for boys to go to university, but, on the other hand, general courses have
higher general prestige and so are better for ladylike marriageability); the
family’s economic circumstances (keeping a child in a Tokyo university can
cost well over half of the average Iwaki family’s income); the general level
of social aspiration—white collar families are more likely to make sacrifices
to push a moderately bright child into the university-going bracket, blue
collar families to think that their child might as well learn something useful;
and parental occupation—families with small businesses want their eldest

sons to learn something of use to the business, private doctors’ and dentists’
sons aspire to enter father’s profession.
In a metropolitan prefecture like Tokyo where social aspiration levels are
higher and universities more cheaply accessible to those who can commute
from home, a much higher proportion of the high school population is bent
on taking a general course and keeping open their options for university
entry. Hence the vocational school students come from lower down the ability
range, as Table 2.4 shows. The ‘image’ of the vocational high schools is less
favourable, and employers’ eagerness to recruit from them is diminished—
quite apart from any doubts they might harbour about the usefulness of what
is taught in such schools or the quality of their teaching—doubts which might
apply equally in Iwaki as in Tokyo.
UNIVERSITY ENTRY
Vocational courses at the university do not suffer so obviously from any
‘ability-labelling’ problem, and the extent to which the ‘academic bias
problem’ is evident is limited. Japan has no Oxfords and no Cambridges. Its
elite universities have never been places where reverend clerks prepared young
gentlemen for a life of, hopefully cultured, indolence, or nobility-obliged
public service. They started off—just at the
Who goes where? 31
time when the feudal aristocracy was being pensioned off and winkled out
of its land rights—as meritocratic as the grandes écoles, and they were built
for a country which took industrialization, and especially manufacturing,
seriously. The cultural break with the past which the political upheaval of
the 1860s and 1870s brought, immeasurably weakened Confucian notions
about gentlemen keeping out of kitchens. Engineering as well as science
was an integral part of the first university foundations.
As between arts stream and science stream, therefore, there is, overall, no
great difference either in prestige, or in material prospects, though it seems
to be still the case that the law department of Tokyo University, established

as the meritocratic route into the governing class in the 1880s before there
were many scientists in Japan (nor politicians of any consequence to dilute
the bureaucrats’ power to govern) produces a higher proportion of the
directors of major manufacturing companies than the same university’s
science and engineering faculties. And in MITI the top universities’ science
graduates are more likely to be recruited as technologists than as generalists
with, consequently, lesser chances of reaching the top. But that is a special
case, and in general, children who are good at maths will have to be quite
single-minded to resist the assumption of teachers and parents that, of course,
they should try to get a university science place.
32 How the Japanes learn to work
Anyone trying for a place at a national university does not have to make
a definite choice of stream until he writes out his entrance application forms:
the first-hurdle national examination is in the five main subjects—maths,
science, social studies (history and geography), Japanese and English—for
all applicants. For the private universities, entrance requirements are more
diverse, and there has been more experimentation with admission by high
school record only. Some also require that applicants take the same first-
hurdle national examination; others have only their own examination. Most
of these require only three subjects (excluding either science and maths or
social studies and Japanese), so in the private high schools where pupils are
more likely to be able to afford the higher fees of the private universities,
there has generally been some specialization from the age of 16 or 17 since
the introduction of the broader national examination for the national
universities in 1979. Enough bright students have opted for these more
concentrated courses leading to the private universities to have had a
noticeable effect on the (hensachi-measured) ability levels of entrance cohorts
at Keio and Waseda, the two leading private universities, thus raising their
prestige, increasing the numbers choosing that route and hence the intensity
of competition, thus further raising ability levels, etc. Such is the respect for

the Ministry-ordained curriculum guidelines, however, that even the more
concentrated courses do not drop the non-central subjects entirely, so that a
change of stream is not too difficult at the age of 18.
The hensachi system permits some objective measure of the attractiveness
of different subjects and their consequent ability to recruit able students.
The advertising literature of the cram schools (which are attended by ronin
students making a second or third attempt at entrance exams as well as by
current high school students) rate each university department by the hensachi
score which should guarantee an 80 per cent chance of success in its entrance
examination. (The figures are produced by analysing the ‘average mock test
scores’ of the previous year’s applicants and the difference in scores between
those who passed and those who failed each department’s entrance
examination. The banning of the commercial tests mentioned earlier has
applied to public high schools, too, but has had even less effect on practice
than in middle schools since a much higher proportion of would-be university
students go to juku, and a high proportion are in private, not public, high
schools anyway.) From these figures it seems, if one takes the top university,
Tokyo, that entrance scores for the Law/Political Science Faculty and for the
Who goes where? 33
Science and Engineering Faculty are about equal, and both several points
below the Medical Faculty. (The intense competition to enter all medical
schools—in some of the less savoury reaches of the private university system,
competition in parental donations to the Building Fund as well as in marks—
is a reflection of the high incomes which doctors enjoy in Japan.)
Table 2.5 uses these ratings—as given by the brochure of one cram school
chain—to compare the national distribution by ability rankings of pupils in
economics on the one hand, and of various branches of science and
engineering on the other. It makes the point that the latter are on average
brighter, but in both subjects the dispersal is wide. There are a good number
of students graduating from private provincial engineering colleges whose

academic achievement levels are not much above average for the whole ability
range, and well below average for the university-going group. They are likely
to be recruited by local small building firms and manufacturing companies
for careers which would look more like technicians’ careers in Britain.
Likewise, or even more so, the lesser-university economists.
Within science and engineering there is certainly evidence of the ‘academic
bias’. But again it gets intermixed with the ‘ability labelling effect’. Where
physics is established as the prestige subject par excellence—as it was made
in Japan by the nation’s first two Nobel prizes—then that is where the cream
of the cream go. And getting into physics then becomes the only way of
publicly demonstrating that you are the cream of the cream.
Again there are objective indicators of differential values, albeit at the
local level. Tokyo University recruits all its science and engineering students
through two entrance examinations, one for the physical and one for the
biological sciences. Specialization starts at the beginning of the third year,
and students are allocated to their first-choice department according to their
marks in the first two years’ courses. Physics is, indeed, the elite department
which it is most difficult to get into, and it is some of the less prestigious
engineering departments like civil engineering which scoop up the left-overs.
But equally, some of the engineering departments—aeronautical engineering,
for example—are near the top of the tree.
And that is, after all, an elite university pattern—the pattern at a university
whose graduates can expect, if they survive among the quarter of the intake
who are admitted to the PhD course, to have no difficulty in pursuing an
academic career—albeit as teachers at a provincial university. At universities
further down the ability/prestige
36 How the Japanes learn to work
ladder—even among the less than a dozen universities to which PhD courses

are confined—this attraction of the academic, pure-science stream is less
evident. As will be seen from the overall distribution pattern shown in Table
2.5, although science and maths students do have better academic records
than engineering students, the difference is not very great.
It is also relevant that the science graduates of lesser universities and
those who do not make it into the graduate courses of the elite universities,
have little hesitation in going into industry after they graduate. The tendency
noted in other countries for finance houses, merchant banks and stock-broking
companies to snap up the best and the brightest of them was almost unknown
in Japan until 1987 when the internationalization of finance and the euphoria
of five years of bull markets finally produced yuppie-level salary offers from
the securities companies, and the newspaper leader writers began to deplore
the corruption of hitherto healthy Japanese capitalism. Are those they asked,
who make money finally going to have higher prestige than those who make
products? Is Japan finally treading the primrose path that led to Britain’s
decline?
A newspaper columnist joining in the general attack on the banks who
were demanding public funds to bail the banking system out of its bad debts
crisis, suggested that before they asked for public funds they should make
amends for past errors. First they should sell all the art works they bought at
inflated prices during the bubble, and second they should send all the scientists
and engineers they hired during the bubble into manufacturing industry where
properly they belong.
Higuchi has done some interesting calculations which demonstrate not
only the change in the industry-destinations of graduates overall, but how
the shift applies differentially to the high achievers and the low achievers
(Higuchi 1992:154). The HA/LA—high achiever/low achiever—co-efficient
of Table 2.6 is calculated from Higuchi’s figures. Divide all the graduates of
one year into three groups, those from universities which required on entrance
a hensachi of over 60 (one standard deviation above average), those where

the entrance cut-off was below 50, and those in between. The percentage of
the first group who entered a particular industry, divided by the percentage
of the second group entering the same industry is that industry’s HA/LA co-
efficient for that year. Although 1975 was a recession year and 1990 a boom
year, figures for intermediate years suggest that the difference between these
two years is more a matter of secular than of cyclical change. It will be
apparent that the machine industries are gaining in ‘quality’ at the expense
of steel and textiles, and so, also, are ‘transport and utilities’ —
Who goes where? 37
probably thanks largely to telecommunications. The other big gainer is,
indeed, finance and insurance.
LIFETIME EARNINGS
It may seem strange, to those used to orthodox analysis by economists of the
rates of return on investment in a university degree, that Japanese youth
should be so preoccupied with climbing the educational ladder when, as
Table 2.7 shows, the lifetime earnings of university graduates are not so
superior to those of high school graduates. The differential has always been
small—smaller, probably, than in any other country. According to Higuchi’s
calculations (1992:151), using the same source as Table 2.7, the differential
in totalled life-time earnings almost disappears—it narrows to less than five
percentage points—if one takes account of the number of years of earning
and, using a 10 per cent discount rate, calculates the present value at age 16.
His figures show that even on a discounted present value basis, differentials
were greater in the 1960s, but from 1975 to his latest date of 1989 they
remained stable.
38 How the Japanes learn to work
The figures of Table 2.7 suggest a long-term trend for the graduate
premium to increase at the early ages and decline after the mid-thirties. The
former may perhaps be ascribed to automation and the generally diminishing
importance of shop-floor workers, lowering the incentives for firms to try to

attract good high school graduates with attractive starting wages. The latter
may be explained by the diminishing ‘quality’ of university graduates as the
proportion of the age group entering university has greatly expanded.
University graduates had much stronger claims to be an intellectual elite in
the 1950s when the older cohorts of the 1976 column in Table 2.7 were
being educated. There does seem, however, to have been some acceleration
of the decline in the graduate premium for ages over 45. This is probably the
effect of what the Japanese call the dankai no sedai phenomenon—the ‘lump
generation’ of management-track recruits hired during the high-growth period
in expectation of continuous expansion which did not happen. It is plausible
to suppose that they have been more subject to taps on the shoulder inducing
them to take voluntary early retirement than high school graduates, and re-
employment is almost always at a lower salary. Also, in many firms among
those badly hit by the recession, managers have been induced to accept
exemplary—usually temporary—salary cuts to justify belt-tightening salary
freezes for union members, though whether that was sufficiently common to
be observable in the statistics is hard to judge.
Whatever the explanation, the fact is that calculations of the overall
graduate premium such as are contained in Table 2.7 are not often made, nor
generally discussed. And a major reason for that is that, even if 18-year-olds
Who goes where? 39
did make rational rate-of-return calculations as the human-capital theorists
tell them they should (or tell them that they do), how much the average
lifetime earnings of university graduates exceed those of high school
graduates would hardly enter into their choice of whether to go to a university
or not. What counts, everyone assumes, is which high school and which
university. This has been true for a long time: the first record of
institutionalized salary differentials dates from 1917, when NYK, the shipping
line, offered starting salaries for graduates of Tokyo Imperial University of
¥40 a month, the graduates of the two leading private universities, Keio and

Waseda, ¥30, and of lesser universities, ¥25 (Amano 1996:194). Postwar
egalitarianism forebade such overt discrimination, but which university one
graduated from became an enormously important determinant of whether
one got into firms like NYK in the first place. Internal career-track distinctions
among graduates of different ‘ranks’ of universities have recently been revived
in many large firms. ‘Headquarters recruitment’ is for the high flyers who
are likely to be posted to any part of the firm or its subsidiaries, and who are
eventual candidates for the top jobs; ‘local recruitment’ for those whose
academic record suggests that they are never likely to make senior
management anyway. In the mid–1990s one of the banks introduced a third,
intermediate career track for graduates.
Table 2.6, showing very different HA/LA co-efficients between graduates
in services or construction where pay is relatively low, and those in finance
and insurance where it is high, is one aggregated indication of the way
these mechanisms work. So are the firm-size differentials shown in Table
2.8. The importance of the latter is clear from another study which found
that in 1987, of men and women graduating from universities at the bottom
end of the hierarchy—those where the entrance-level hensachi was rated
at less than 45—only 7 per cent
40 How the Japanes learn to work
entered firms with more than 5,000 employees, whereas of those coming
from universities where the entrance rating was over 70, 70 per cent did so
(Takeuchi 1989). The wage differentials shown in Table 2.8 apply to graduates
as much as to other workers. The same source as Table 2.7 shows by how
much the wages and bonuses of university graduates in manufacturing firms
of more than 1,000 employees are greater than those of university graduates
in firms with 10–99 employees—11 per cent greater for 25–29 year-olds, 34
per cent for those in their late thirties and 66 per cent greater for those aged
50–54 (1994 figures).
The likelihood is, in other words, that any compression of the difference

between high school leavers’ and university graduates’ wages is a function
of changing inter-sectoral or inter-firm differentials and not a sign of an
overall shift to greater equality of income distribution. The Gini co-efficient
has been increasing rather than diminishing, and so, as in most other industrial
societies, has the dispersion of wages and salaries around the mean among
employees. It is unlikely that there should not be some correlation with
academic performance in such a credential-conscious society as Japan, but
that is not captured by a crude graduate/non-graduate distinction.
SOCIAL MOBILITY
Equality of outcomes and changes in the income distribution are one thing;
equality of opportunity is another. The grand social mobility studies tell one
very little about recent decades and use only the crudest distinctions of school
and university ‘quality’ (Ishida 1991). More informative is Higuchi’s recent
attempt (1992) to squeeze enlightenment out of official statistics and the
informal university rankings provided by the cram schools. He finds, first of
all, that the effect of parental income on the probability that a child will go to
a university at all has steadily diminished, even though university fees have
steadily increased in proportion to incomes (from a little over 1 per cent of
the annual income of the average 45–49-year old for national universities,
and 6 per cent for private universities in the mid–1970s, to 6 per cent and 10
per cent respectively in 1990). But second he finds that there was, for 1990
graduates, a considerable effect of parental income on the probability of
entering a university of high prestige, with difficult entrance examinations
usually attempted only by those who have had high hensachi ratings in their
high school.
It is, moreover, the impression of many observers, borne out by the parental
income figures given on scholarship application forms by students at the top
universities, that the correlation between parental income (and occupational
prestige category) and childrens’s academic performance is growing. The
general assumption is that the correlation is explained by economic factors—

Who goes where? 41
the much greater expenditure by well-off families on educational aids and
cram-school fees which is demonstrated clearly in the family budget surveys.
Educational sociologists who have looked into the matter, however, (for
example, Moriyama and Noguchi 1984) claim that large expenditures on
out-of-school tuition seem not to make much difference to performance,
and attribute the correlation more to cultural, rather than to economic,
factors—books about the home; the conversation over the breakfast table;
and high aspirations which might explain why children from higher status
families do well even in subjects like maths which are not often casually
discussed over the breakfast table. These differences in aspirations are,
however, not all that easy to detect in a society as culturally homogeneous as
Japan, and some brave spirits suggest that genetics might have something to
do with it—the effect of several generations of credential-meritocratic
mobility and assortive mating. The suspicion that this might be so make it all
too delicate a topic for any but the boldest spirits to tackle.
So one is reduced to anecdote, and there are many which support the
common sense perception that equality of opportunity increased massively
in the first three decades after the war and that occupational destiny has a
good deal more to do with native ability as compared with parental social
status than used to be the case. One might cite, for example, the observation
of a senior engineer on the success of the quality circles (work process
improvement groups) in his factory. He attributed a lot to the initiative, the
leadership, the sustained level of intellectual curiosity and the inventiveness
of some of his senior foremen: ‘Nowadays they would have gone to one of
the top universities, but they left school at fifteen—at twelve some of the
older ones. One never seems to get people of that calibre coming out of high
school into blue collar jobs these days.’
That can serve as a final reminder of the pervasive implications of the
mechanisms to the description of which this chapter has been devoted. The

way education and training systems sort and label people and thereby
influence their occupational destinations is a very important aspect of those
systems. It rivals and (for a society’s economic functioning at least) may
even surpass in importance the aspect to which attention is more frequently
devoted—the way schools seek to influence the intellectual (as well as the
aesthetic and moral) development of their charges.

3 Vocational streams in the
mainline formal education
system
THE SECONDARY LEVEL
In 1993 high schools which offered nothing but vocational courses made up
17 per cent of the 5,501 total. These schools are called industrial, commercial,
agricultural or music high schools, and so forth, according to their specialty.
In addition in 1993, there were another 32 per cent of schools with some
(minority) vocational streams, leaving just over a half of the total number of
schools with general academic courses.
Rarely do vocational schools offer a single type of course; some provide
up to ten different ones. Table 3.1 shows how the Ministry classifies them
and shows how the pupils are distributed among them.
The 1.25 million pupils on vocational courses in 1994 made up 26 per
cent of the total in high schools. (They had numbered 0.93 million in 1955
and 2.05 million in 1965—then, in both years, 40 per cent of