-
Economics of Education Review, Vol. 9. No. 4, pp. 351-W. 1990.
0272-7x7/90 $3.00 + 0.00 Printed in Great Britain. 0 1590 Pergamon
Press plc
Vocational Education and Productivity: A Case Study of the
Beijing General Auto
Industry Company
WEI-FANG MIN* and MUN CHIU TSANGPS
*Institute of Higher Education, Peking University, Beijing
100871, China; and tCoUege of Education, Michigan State University,
East Lansing, MI 48824, U.S.A.
Abatraet -This paper examines the economic rationale for the
current vocationahxation of secondary education in China by
comparing the performance of vocational education graduates and
general education graduates in the workplace. It uses a model that
recognixes the importance of work effort and group dynamics on
production, employs a more direct measure of productivity, and
presents an empirical study in the context of the current
educational and workplace reforms in China. Based on data from the
Beijing Auto Industry Company, China, the study finds that
vocational-technical education graduates engaged in factory work
relevant to their previous trainidg were more satisfied with their
job and were more productive than general education graduates. It
also found that a workers education had a lesser impact on
productivity when the worker performed in a more co-operative
environment.
I. INTRODUCTION
SINCE THE Third Plenary Session of the Eleventh Congress in
1978, the Peoples Republic of China (China) has pursued a national
policy of modemiz- ation. This policy has resulted in a major shift
in the purpose and function of the education system. Instead of
being used for political indoctrination and as an instrument for
class struggle, the school is trying to inculcate the young in both
the skills and values needed by a changing economy and to be the
strategic foundation of the modernization pro- gram. In the late
1970s and the early 198Os, the education system underwent several
significant changes; including reinstating national exami- nations,
re-establishing key schools and universities, re-focusing on skill
training, and sending students and faculty for training overseas.
The call for changes in education reached its peak in May, 1985,
with an official pronouncement of a large-scale reform of the
education system (Peoples Press, 1985).
One of the most important components of the 1985 reform plan was
the vocationalization of
education at the upper-secondary level. The plan calls for a
rapid development of secondary vocational and technical schools so
that enrollments in these schools would constitute about 50% of the
total enrollment in upper-secondary education in the near future, a
dramatic increase from the 5% level in the early 1980s.
Since modernization policies were formulated in the late 197Os,
a large-scale industrialization pro- gram has been implemented. As
industrialization accelerated, serious shortages of skilled
workers, semi-skilled workers and middle-level technicians have
emerged. In many factories, some of the low- and middle-level
technical work has to be performed by high-level technicians and
engineers. Also, workers recruited from general senior high schools
were found to be inadequately prepared for tech- nical work.
However, in the late 1970s in China, less than 5% of senior high
schools had a vocational track. Most of the senior high schools
focused on college preparation, but the transition rate from these
schools to colleges was less than 4% (National Bureau of
Statistics, 1986). Most of these senior high school graduates had
to work in industry as
*To whom correspondence should be addressed.
351
-
352 Economics of Education Review
skilled or semi-skilled workers. Besides, there is increased
emphasis on economic efficiency. In the belief that
vocational-technical education can better prepare young people in
the skills needed for employment in industry than general
education, vocationalizing senior high schools is seen as a
strategy that contributes to increased efficiency in educational
investment. Thus, industries and government agencies appeal for
more upper- secondary vocational-technical schools (Ministry of
Education, Ministry of Labor and Personnel, Ministry of Finance,
State Planning Commission, 1983).
productivity, presents a model of education and productivity
that highlights the role of work effort and group dynamics in
production, and constructs several testable hypotheses to be tested
in the empirical study.
Economic Relevance of Vocational Education
A key assumption in the economic motive behind this secondary
education reform is that vocational- technical school graduates are
more productive than general school graduates in factory work. How-
ever, there exists no systematic empirical study to test the
validity of this assumption in the present day workplace in
China.
The linkage between education and productivity has been the
subject of numerous studies in both the economics and sociology
literatures. Many studies have examined how different levels of
education and different types of education influence worker pro-
ductivity. In particular, a significant number of studies exist
which compare the productivity of vocational school graduates with
that of general education graduates.
This paper attempts to examine the relationship between the
vocationalization of secondary edu- cation and economic development
in China by studying the impact of vocational education on
productivity in the specific institutional context of China. By
presenting a case study conducted at the Beijing General Auto
Industry Company, it com- pares the performance of
vocational-technical education graduates and general education
gradu- ates in the workplace. The object is to examine whether
secondary vocational-technical school graduates are more satisfied
with their jobs and more productive than senior high general
education graduates. The empirical analysis has some origi- nality
in applying a model of productivity that recognizes the importance
of work effort and group dynamics on production, and in employing a
more direct measure of productivity.
Most of these studies use earnings as a measure for
productivity. A few use alternative proxies such as a worker
efficiency rate (Fuller, 1970, 1976; Godfrey, 1977). Most of the
studies are cross- sectional in design. Some have examined the
over- time differences in the relative effectiveness of vocational
and general education (Hu et al., 1971; Gustman and Steinmeier,
1982; Meyer and Wise, 1982; Bolino and Uri, 1982; Tannen,
1983).
The rest of this paper is divided into three sections. Section
II provides the conceptual treat- ment that guides the empirical
study presented in Section III. Section IV summarizes the findings
of the study and makes recommendations for further research.
II. VOCATIONAL EDUCATION AND PRODUCTIVITY: EMPIRICAL EVIDENCE
AND
CONCEPTUAL ISSUES
The findings from these studies are mixed, with some showing
higher earnings or productivity for vocational graduates (Hu ef
al., 1971; Freeman, 1974; Corazzini, 1968; Fredland and Little,
1980; Li et al., 1981; Bolino and Uri, 1982; Gustman and
Steinmeier, 1982) and others showing no significant difference or
negative results in comparison with general education graduates
(Taussig, 1968; God- frey, 1977; Grass0 and Shea, 1978; Rumberger
and Daymont, 1984; Meyer and Wise, 1982; Psacharo- poulos and
Loxley, 1985). Some investigators have found that the economic
effects of vocational edu- cation vary with programs, and the
gender and ethnicity of students (Gustman and Steinmeier, 1982; Li
et al., 1981; Meyer, 1982); and the effec- tiveness of vocational
training depends upon the nature of the training received and type
of insti- tutions providing it (Tannen, 1983). Also, the estimated
results can change significantly with the model employed and the
time period studied (Gustman and Steinmeier, 1982). In general,
there is no consistent and convincing evidence in support of the
presumed productivity advantage of vocational education over
academic education.
This section provides a brief critical review of the Besides the
mixed evidence in the existing litera- existing literature on
vocational education and ture, there are some issues to consider in
applying
-
Vocational Education and Productivity 353
the methods of these studies to the Chinese situ- work
experience, and other on-the-job training ation: variables.
First, most of the studies do not measure pro- ductivity
directly but assume that productivity is reflected in individual
wages. While the use of wages as a proxy for productivity is
questionable for capitalist economies characterized by imperfect
competition, it is even more problematic for China. China has a
centrally planned economy; wages are tightly controlled and set by
the government. In such a case, wages would not be a valid measure
of productivity.
Second, most of the studies have not taken account of the impact
of work effort on productivity. The degree to which workers exert
their effort on their jobs could make a significant difference in
productivity; and work effort is an important inter- mediate
variable linking education and productivity (Pencavel, 1977; Levin,
1984; Tsang and Levin, 1985; Tsang, 1987). Without taking account
of the work effort variable, the linkage_ between education and
productivity would not be properly revealed.
Third, most of the studies have taken the indivi- dual worker as
the unit of analysis and have ignored the potential interaction
among workers that might affect productivity. In modem production,
however, a large proportion of work is group work which can
influence how an individual worker behaves (Alchian and Demsetz,
1972). In China, especially, many incentives are given at the group
level. The impact of group dynamics on individual productivity
should be taken into account.
Recent research on the relationship between education and
productivity has focused on work effort as an important
intermediate variable relating the education of workers and their
productivity (Tsang and Levin, 1985; Tsang, 1987). Work effort is
seen as a function of worker characteristics, job characteristics,
and the degree of match/mismatch between these two sets of
characteristics. Worker characteristics include worker skills and
abilities, type and level of education, worker needs and
expectations, experience, age, and gender. Job characteristics
include both physical aspects such as the nature of job and its
skill requirements, the reward structure and promotional
opportunities, and the physical environment; and social aspects
such as peer-group interaction, supervision and worker-supervisor
relationship, atmosphere for co- operation, etc. Two types of
match/mismatch can be identified: match/mismatch in skill and
match/mis- match in occupational expectations. If a worker finds
his skills to be underutilized in his job or his occupational
expectation unfulfilled, he may be- come dissatisfied with his job,
exert lower level of effort, thus resulting in lower productivity
(Vroom, 1964; Sheppard and Herrick, 1972; Quinn and Mandilovitch,
1975; Srivastva ef al., 1977).
Fourth, vocational education might have different impacts on
productivity in different social, economic and institutional
contexts. An empirical study on China is needed to examine the
economic rationale for the vocationalization of secondary education
within Chinese contexts. This study attempts to resolve these
issues.
Education can affect worker characteristics by its influence on
an individuals skill, needs, and expec- tations; it can also affect
the job characteristics of a worker by influencing the kind of jobs
available to him/her as a result of his/her education. Through its
influence on both worker characteristics and job characteristics,
education can affect the degree of match/mismatch between these two
sets of charac- teristics, and thus work effort and productivity.
Education can have an indirect effect on pro- ductivity besides a
direct effect.
Education, Work Effort and Productivity Most of the empirical
work on the relationship
between education and productivity is based on human capital
theory (Schultz, 1961; Becker, 1964; Mincer, 1974). According to
this theory, education can raise the productivity of an individual,
primarily by increasing the individuals cognitive skills. Thus,
there is a direct and positive relationship between education and
productivity. This relationship is usually estimated by relating
earnings to human- capital variables such as years of formal
schooling,
This model of work effort has been applied to the study of the
impact of the level of education on productivity (Tsang, 1987;
Tsang et al., 1991). It was found that workers with more education
than their jobs require (i.e. underutilization of education) tend
to exert lower work effort (proxied by job satis- faction) and have
lower productivity. It is conceiv- able that the type of education
(vocational or general) that workers receive can be related to
differences in productivity among workers. To the degree that
vocational education and general edu- cation socialize individuals
with different skills,
-
3.54 Economics of Education Review
needs and occupational expectations, the type of education may
be correlated with work effort and productivity.
Individual Behavior and Group Dynamics It should be noted that
the Tsang-Levin model of
work effort is based on indi~duafs, not groups. Because of the
possible interactions among indivi- duals in a group setting
(Forsyth, 1983; Shaw, 1981; Schmuck and Schmuck, 1983), an
individual may perform differently under different group settings.
The impact of the interaction among workers in a group on the
relationship between education and productivity has to be taken
into account. Consider the situation in the workplace in China.
On the one hand, since most of the work is group work and many
incentives (bonuses) are given at the group level, co-operation
among group members is likely. For example, in a co-operative work
environ- ment those with more knowledge and skills may share that
information with those with less knowf- edge and skiffs. This may
equalize skiff proficiencies among workers and may obscure the
differences in productivity due to differences in the skill back-
ground of workers.
On the other hand, since 1979, the promotion of workers to a
higher wage level within groups has been partially based on the
evaluation of individual performance both in the production process
and in the tests of technical knowledge and skiffs. Only a small
proportion of workers is promoted each year. In the unified
eight-level wage system, the ratio of the Iowest wages to the
highest wages is 1:3. This promotion policy is supposed to
encourage workers to improve themselves technically, but it may
also prevent workers from helping each other, thus maintaining
individual differences in skills and productivity. Since both
group-based incentives and individual-based incentives exist in the
workplace, both group interactions and individual behavior are
assumed in this study. Some indicators and hypoth- eses related to
group dynamics will be developed and included in the Tsang-Levin
work effort model. This study provides a preliminary analysis of
the impact of group dynamics on the relationship between education
and productivity in China, a previously unexplored area of research
on China.
Hypotheses to be Tested Based on the above discussion the
foIlowing
hypotheses, to be tested empirically, have been generated.
Hypothesis (1): Secondary vocational school graduates as factory
workers holding jobs more closely related to their training tend to
exert higher work effort in their jobs than general senior high
school grad~tes who hold the same jobs. This result might arise
because vocational education graduates have better matched
occupational expectation in China. Since secondary vocational
schools prepare students for employment after graduation, students
in these schools expect to be factory workers. General senior high
school graduates are prepared for college. What they want is to be
college students, but college cannot enroll them all. Most of them
are forced to work as factory workers. They do not have the
psychological preparations for the jobs they are going to do. Thus
with respect to occupational expectation, vocational education
graduates are better matched and they thus have higher work effort2
Vocational education graduates may also have higher work effort
than general education graduates because their skills acquired in
school better match those required in the workplace.
Hypothes~ (2): Secondary vocationat school graduates as factory
workers hoIding jobs closely related to their training tend to be
more productive than general senior high school graduates who hold
the same jobs. This result may arise because vocational school
graduates exert higher work effort in their jobs than genera1
senior high school graduates, and the skills they have acquired
from the vocational programs better meet the skill re- quirements
of their jobs.
Hypothesis (3): A workers educational back- ground has a less
significant impact on productivity in a more co-operative work
group than that in a less co-operative one. This result may arise
because, in a group with a high level of co-operation, individuals
with more skills may share their knowledge with individuals with
fess skills. This will equalize skill proficiencies among members
of the group, thus obscuring differences in productivity due to
differ- ences in the educational background of the group. The
impact of vocational education on pr~ucti~ty in a workplace with a
higher level of co-operation may be fess significant than that with
a lower level of w-operation.
-
III. VOCATIONAL EDUCATION AND and other programs. A larger
proportion of these PRODUCTIVITY: A CASE STUDY OF THE training
programs are programs for training skilled
BELjING GENERAL AUTO INDUSTRY and semi-skilled workers and
middle level tech- COMPANY nicians because of the serious shortages
of these
technical personnel. This study will focus on evalu- Methodology
ating the programs for training skilled and semi-
This includes the specification of the population skilled
factory workers. to be studied, and the discussion on the
definition of variables, data needs and collection, and methods of
Defining the variables. According to the frame- analysis. work and
the hypotheses, the following variables are
included in this study: Specifying rhe population. There are two
tracks in Job satisfaction (work effort). Job satisfaction of
upper-secondary education in China at the present workers is
measured by a five-item scale with a value time: of one for the
lowest level of job satisfaction and a
(1) General senior high schools in which students value of five
for the highest level and is taken as a mainly have a
college-preparatory curriculum. The proxy for work effort (see
Tsang, 1987). Accord- transition rate to college was below 4%
before 1980. ingly, job satisfaction is a function of worker Only
those students who are from a few key-point characteristics and job
structures and the degree of senior high schools have a good chance
of entering match/mismatch between these two sets of charac- the
college. Most of the senior high schools gradu- teristics. It is
also an important intermediate vari- ates go to work as
semi-skilled or skilled workers. able relating education and
productivity. It is ex- This study considers general senior high
school pected that job satisfaction is positively related to
graduates who did not go to college but are working productivity.
in the factory. Work efficiency. Instead of using wages as a
(2) Secondary vocational schools in which measure of
productivity, this study uses an alter- students spend half time
studying academic subjects and half time learning specific skills.3
There are two
native measure: work efficiency (Fuller, 1970). Work efficiency
reflects the amount of goods and
kinds of secondary vocational schools. services produced by a
worker during a period of There are senior high schools with a
large propor- time and is defined as:
tion of vocational courses. These are schools that switched
recently from a general track to a Work efficiency = vocational
track. Their programs are being devel- oped. Some of these schools
are still short of actual output of a worker per period
vocational teachers, funding, and equipment (Li et production
quota for a worker per period al., 1985).
There are also company-affiliated vocational The production
quota in the above ratio is a function schools. These schools have
a relative long and of labor, equipment, technology, materials and
stable history. Compared with the other kind, they time. It
reflects the physical conditions and inputs to are much better
funded and managed. Students in production. Thus work efficiency
measures the these schools spend half of their time studying
performance of the worker based on a given set of academic courses,
and half of their time learning inputs and physical conditions of
production per specific skills needed by the companies with which
period. their schools are affiliated. Generally speaking, they The
advantages of using worker efficiency as a will be assigned jobs in
the affiliated companies. The measure of productivity are: (1) it
is a direct vocational school graduates in this study came from
company affiliated vocational schools.
measure of worker productivity, and (2) since quotas are
designed to take account of factors that
There are various training programs in secondary might affect
worker output other than the workers vocational schools, such as
skilled and semi-skilled contribution, work efficiency holds
constant factors worker training programs, middle-level technician
that might affect worker performance, but are training programs,
clerical staff training programs, extraneous to the worker, such as
machine capacity, cook training programs, barber training programs,
availability of tools and quality of materials. Thus it
Vocational Education and Productivity 355
-
356 Economics of Education Review
enables us to concentrate on the performance of a worker.
Variables related to worker characteristics: (1) Type of
education. This is .the major explanatory variable to be
considered. The types of education are secondary
vocational-technical education and secondary general education. It
has a value of one for vocational-technical graduates and a value
of zero for general education graduates. The rest are control
variables. (2) Levef of education. It is measured by the number of
years of schooling. (3) Experience. It reflects abilities acquired
on the job and is measured by the number of years of work
experience. Because some workers had on-the-job training and others
did not, an on-the-job-tra~nf~g variable (measured in months) is
also included. (4) ~ocfoeconumic status. It is measured by parents
level of education. (5) Gender. It has a value of one for females
and a value of zero for males.
Variables related to job characteristics: (6) Salary. Salary is
measured by an eight-Ievel wage scale. Other relevant job variables
include (7) promotional opportunities, (8) level of autonomy in
work, (9) feader-member relationship, (10) peer relatfon~hj~, and
(11) co-operution among workers.6 Variables (7) to (11) are
measured by a five-item scale with a value of one for the lowest
level and a value of five for the highest level.
Data needs and data collection. Data were col- lected in 1986
from the Beijing General Auto Industry Company,7 which currently
has 12 factories and 59,369 employees. There are nine vocationai-
technical schools affiliated with this company. These schools
enroll 600 students each year. Students are admitted to these
v~ational-techni~l schools according to their own applications,
test scores and their academic performance at junior high schools.
Upon finishing 3 years learning at these vocational- technical
schools, the graduates are assigned to jobs in the company.
Selection bias does not appear to be a problem.
566 workers were selected based upon a clustered sampling
strategy from two of the factories within the company. The average
level of education is 10.8 years and 41% of the workers are
females. Among the 566 workers, 150 are vocational-technical school
graduates, and 90% of them consider them- selves as holding jobs
related to their training.
Of the 566 workers selected from the cumpany, this study focuses
on the 413 workers who are under
40 years old. Since this study is concerned with the
productivity impact of the current educational re- form, the older
workers who have undergone a different educational process from the
younger workers are excluded from our analysis. Among the 413
younger workers, 114 are vocational-technical school graduates, and
93% of them consider them- selves holding jobs related to their
training. 288 of the 4I3 workers had 12 years of schooling. Both
the 413-worker and the 288-worker samples are analyzed.
The production records were examined and a questionnaire was
designed by us in collaboration with the staff members of the
company. The questionnaire was administered by the Workers
Education and Training Center of Beijing General Auto Industry
Company through each of the fac- tories Education and Training
Office for Workers. Reliability and validity of these measurements
were tested in a pilot study including 137 individual workers.
I0
Methods of analysis. Our model of productivity consists of a
two-equation recursive system. The first equation is a job
satisfaction equation in which job satisfaction is a function of
the type of edu- cation, and other worker and job characteristics.
The second equation relates work efficiency to job satisfaction,
type of education, and other worker and job characteristics. The
type of education is expected to have a direct effect and an
indirect effect (through job satisfaction) on work efficiency.
To examine the impact of the level of co- operation among group
members on the reiation- ship between education and productivity,
two sub- samples are considered: individuals working in groups with
a high level of co-operation (co-oper- ation among workers has a
value higher than three), and individuals working in groups with a
low level of co-operation (co-operation among workers has a value
not higher than three).
Empirical Results Table 1 presents the regression results for
the 413
workers. The dependent variable in Equation 1.1 is job
satisfaction. The estimated coefficient for the variable type of
education is both positive and significant, showing that
vocational-technical school graduates tend to be more satisfied
with their jobs as factory workers than genera1 education
graduates, thus supporting hypothesis (1). The
-
Table
1. R
esu
lts o
f re
gre
ssio
n analy
ses fo
r w
ork
ers
under 4
0 years
old
Job
sati
sfact
ion
Work
er c
hara
cteri
stic
s:
Leve
l of
educa
tion
Type of
educa
tion
On
the jo
b
train
ing
Exp
eri
ence
Exp
eri
ence
sq
uare
d
Farh
er
s educt
ion
Moth
er
s educa
tion
Gender
Job c
hara
cteri
stic
s:
Sala
ry
Peer
rela
tionsh
ip
Leader-
mem
ber
rela
tionsh
ip
Pro
moti
onal
opporl
unit
y
Work
auto
nom
y
Co-o
pera
tion
am
ong
work
ers
Equati
on 1
.1
JOB
SA
TIS
FAC
TIO
N
-
0.0
51
(1
.5)
Equati
on 1
.2
Equati
on 1
.3
Dependent v
ari
able
W
OR
K
EFF
ICIE
NC
Y
WO
RK
EFF
ICIE
NC
Y
0.0
38
0
.03
3
(5.2
)
(3.2
)*
0.0
11
(2
.9)*
0
.00
6
Equati
on 1
.4
WO
RK
EFF
ICIE
NC
Y
0.0
34
(4
.6)
2
0.2
8
(2.6
)*
0.0
69
(5.4
)
0.0
53
(0.8
) 0.0
12
(2.5
)
0.0
80
0.0
22
(1.6
) (2
.0)*
(5
.3)
::
2:.
0.0
08
(4
.7)
0
.00
6
0.0
24
(2.0
)
0.0
08
9
0.0
25
B
0.0
16
(0.9
) (6
.6)
0.0
27
(3.9
)
- -0
.00
1 (-4
.6)
-0
.00
1
(3.6
).
-0.0
01
(5
.1).
-0
.00
4 (-0
.1)
(-2.7
)
(-3.0
)
z -
- -
-0.0
35
(-0.6
) :
-0.0
15(-
1.4
) -0
.00
3(-
0.1
) -
f!
-0.0
49
(-0.5
) -0
.02
2 (-1
.6)
6
3
%
0.1
3
(2.5
)
0.0
88
(1.4
)
;.O
&
;;2
;*
0.0
19
:
(1.4
) 0
.01
s (2
.0)
a
0.0
32
(0.5
) 0.1
6
(2.5
)
0.0
13
(0.9
) 0.0
13
(0.5
) %
0.0
61
(1.1
) 0
.01
7
0.0
17
0
.01
6
a.
(2.5
)
0.1
5
(2.1
)
0.0
38
(4.4
)
- (2
.0)
-
(2.0
).
$.
Const
ant
0.5
8
0.4
5
0.5
9
R-s
quare
d
0.5
3
0.3
1
0.5
8
0.4
2
N
0.5
6
41
3
39
8
15
5
24
3
Note
: *Sig
nific
ant a
t 0.0
5 le
vel.
Num
bers
in p
are
nth
ese
s are
t-valu
es.
-
358 Economics of Education Review
coefficients for all the other independent variables were
statistically insignificant except for the job characteristic
variables for salary promotional opportunity, and co-operation
among workers which were statistically significant and
positive.
The dependent variable in Equation 1.2 is work efficiency.
Equations 1.1 and 1.2 constitute the recursive model for the
413-worker sample. From Equation 1.2 we can find that the type of
education is positively and significantly related to work ef-
ficiency, indicating that vocational-technical gradu- ates as
factory workers holding jobs closely related to their training are
more productive than general education graduates. Job satisfaction
(as a proxy for work effort) is also positively and significantly
related to work efficiency. Thus the type of edu- cation has a
direct and indirect effect on pro- ductivity, supporting hypothesis
(2). From Equations 1.1 and 1.2, the average total productivity
advantage of vocational education relative to general education is
0.078 (= 0.069 + 0.28 x 0.033). The direct effect of the type of
education (0.069) is much larger than the indirect effect (0.009 =
0.28 x 0.033). Except for gender and leader-member relationship,
the other variables are all significantly related to work
efficiency.
Equations 1.3 and 1.4 are the work-efficiency equations for
workers in more co-operative groups and less co-operative groups,
respectively. From the R-square values of these two equations, we
can see that our individual-based model explains pro- ductivity
significantly better for less co-operative groups than more
co-operative groups. It is also interesting to compare the
coefficients in these two equations. We find that the human-capital
variables reflecting individual differences in skills, such as type
of education, level of education, on-the-job training, and
experience are less significant and less strong in more
co-operative groups. In particular, the coefficient for level of
education is 0.012 and is significant in Equation 1.4; but it is
only 0.006 and is insignificant in Equation 1.3. Also the
coefficient for type of education in Equation 1.4 is 50% larger
than the one in Equation 1.3. This result is consistent with
hypothesis (3). Thus, the level of co-operation among group members
appears to affect the re- lationship between productivity and the
skills of workers.
Table 2 presents the regression results for the 288 workers who
are under 40 years old but also with 12 years of schooling.
Equations 2.1 and 2.2 are the
estimated equations for the recursive model for these 288
workers. The results indicate that, at the upper-secondary level,
vocational-technical gradu- ates are both more satisfied with their
jobs and are more productive than general education graduates. Work
effort, proxied by job satisfaction, is a significant determinant
of work efficiency. The total productivity advantage of
upper-secondary vocational-technical education relative to upper-
secondary general education is 0.096 (= 0.087 + 0.24 X 0.037). The
direct effect of type of education is again much larger than the
indirect effect.
Equations 2.3 and 2.4 are respectively the esti- mated work
efficiency equations for more co-oper- ative groups and less
co-operative groups in the 28& worker sample. The results are
supportive of hy- pothesis (3). In particular, type-of-education
has a much larger coefficient in Equation 2.4 than 2.3. The
coefficient for on-the-job training in Equation 2.4 is twice as
large as that in Equation 2.3; it is significant in Equation 2.4
and insignificant in Equation 2.3. In general, skill-related
variables appear to be more important factors for explaining
productivity differences in less co-operative groups than they are
for more co-operative groups.
IV. CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER RESEARCH
Our empirical study found that secondary vocational education
graduates holding jobs closely related to their training are more
satisfied than secondary general education graduates with their
jobs as factory workers. A plausible reason is that vocational
school graduates have better preparation for their jobs. The
expectations and skills they acquire in vocational schools better
match the job characteristics of factory work in China.
Vocational education graduates holding jobs closely related to
their training are also more productive than general education
graduates. This may be explained by the higher level of job
satisfaction of vocational education graduates and the specific
skills of these graduates.
Moreover, there is significant interaction between education and
the level of co-operation among group members. In a work
environment which is more co-operative, the education or skill
back- ground of workers becomes less important. This may be due to
the sharing of knowledge and mutual
-
Table
2. R
esu
lts o
f re
gre
ssio
n analy
ses fo
r w
ork
ers
under 4
0 years
old
and w
ith 1
2 years
of sc
hoolin
g
Equati
on 2.1
JOB
SA
TIS
FAC
TIO
N
Equati
on 2.2
Equati
on 2.3
D
ependent v
ari
able
W
OR
K E
FFIC
IEN
CY
W
OR
K E
FFJC
IEN
CY
Equati
on 2.4
WOR
K EF
FICIEN
CY
Job
sati
sfac
tion
W
ork
er c
hara
cteri
stic
s:
Lev
el o
f ed
ucat
ion
Typ
e of
edu
cati
on
On
ihe
job
trai
ning
E
xper
ienc
e E
xper
ienc
e sq
uare
d Fa
ther
s ed
ucdo
n M
othe
rs e
duca
don
Gen
der
Job c
hara
cteri
stin
: S&
ry
Pee
r re
lati
onsh
ip
Lea
der-
mem
ber
rela
rion
ship
P
ro~
tion
~i
oppo
rtun
i~
Wor
k au
tono
my
Co-
oper
atio
n am
ong
wor
kers
-
0.24
- (2
.2)*
;.g
g.3
-0.0
11(-0
.2)
-0.0
56
(-0.8)
-0
.029
(-0
.2)
0.1
5
(2.6)
0.
056
(0.7)
0.
038
(0.5)
0
.18
(2
.4)*
0.
012
(0.2)
0.
22
(2.6)
0.03
7 (4
.7)
0.0
87
(6
.1)*
0.0
1;
(1.6
) o.O
G
(1.3
) 0
.00
3 (0
.3)
0.0
04
(0
.2)
0.0
06
(0
.8)
0.0
16
(1
.1)
0.0
33
(3
.1)*
-
0.04
0 (2
.7)
0.0
61
(2
.3)*
0
.00
4 (1
.2)
0.0
27
(3
.2)
-0
.00
1 (-2
.0)
-
-OH
&-0
.1)
0.0
39
(4
.0).
3
0.1
0;
(6.0
).
I:
P.
0.0
08
(3
.0)*
0
.02
2 (3
.2)
f
-0.0
01
(-1
.3)
G
-
P
-0.0
18(-
l.*)
a
p
f:
Const
ant
1.2
0
.64
0
.68
0
.72
R
-square
d
0.3
3
0.5
5
0.4
1
0.5
2
N
27
0
27
9
11
3
16
6
Note
: l Sig
nific
ant a
t 0.0
5 le
vel.
Num
bers
in p
are
nth
ese
s are
r-valu
es.
-
360 Economics of Education Review
assistance among workers that reduce individual differences in
skills.
Furthermore, the results show work effort (proxied by job
satisfaction) to be a significant factor affecting productivity. It
is an important intermediate variable relating education and pro-
ductivity. Thus, education can affect productivity directly through
its human-capital capacity; it can also affect productivity
indirectly through the work effort variable. This indirect effect
depends on the degree of match/mismatch of worker characteristics
and job characteristics. If education socializes students in the
skills, expectations and needs that are consistent with the jobs
they find in the workplace, then education has a positive and
significant indirect effect on productivity. If there is a
significant level of mismatch in skills and expec- tations, the
indirect effect can be negative (Tsang, 1987). Since the indirect
effect of education depends on a complex set of factors, the
relationship between education and productivity is necessarily
intricate, and is likely to vary under different conditions. This
may partly explain the mixed evidence on the relationship between
vocational education and pro- ductivity in the literature. From the
findings of this study it might be implied that at Chinas current
stage of social and economic development, com- pany-affiliated and
job-directed secondary vocational education, not general education,
would be the appropriate way for preparing factory workers for
Chinas industrialization. Such a recommendation, however, is
premature, for several reasons.
First, our study relates to only one company of the auto
industry. Further research is needed to deter- mine if the findings
are generalizable to other companies in the same industry and other
industries in China. It would also be revealing to study the
productivity of graduates of vocational schools not affiliated with
a company. In fact, our study has considered user graduates of
well-established company-affiliated vocational schools only, non-
user graduates from non-affiliated vocational schools might not
perform as well. Our finding may represent the upper bound of the
productivity advantage of vocational-technical education rela- tive
to general education. Further research should be conducted to
examine the cases for non-users and non-affiliated schools.
Second, even if the findings are generalizable to other
companies and industries at a point in time,
we need to examine the education-productivity relationship over
time. It is conceivable that, in the long run, general education
graduates can catch up and even surpass vocational education
graduates in productivity. Because of their general training,
general education graduates may be more flexible in handling a
range of tasks and are more adaptive to technologically induced
changes in the workplace than vocational education graduates
trained in specific skills. Time and rates of renewal of technol-
ogies of production will play a critical role. General education
graduates may be more productive than vocational education
graduates in a dynamic en- vironment after they have made an
initial adjust- ment to their job.
Third, our study compares the relative pro- ductivity advantages
of two types of education, without considering their respective
costs. To in- form decisions on efficient allocation of scarce
resources, a cost-benefit approach should be adopted. Further
research should thus examine the costs of different types of
education.
Finally, our analysis has been confined narrowly to a comparison
of economic returns to different types of education; but economic
rationality is not the sole basis for national policy-making for
edu- cation in China in general and the vocationalization of
secondary education in particular. Political and social
considerations are also relevant. The vocationalization of
secondary education in the broader context of the current
educational reform in China can be seen as part of an effort to
rally popular support for the modernization policy of the new
leadership, thus carrying both economic and political legitimation
implications. In China as well as in other countries, vocational
streaming is a form of social stratification. It is also a commonly
used strategy for reducing the social demand for higher education.
Thus, the vocationalization of secondary education reflects a
reassessment of the social function of education in China. These
various considerations should be properly balanced in arriv- ing at
an educational policy13 to promote national development.
Acknowledgements - An earlier version of the paper was presented
at the Comparative and International Education Society Conference
in Washington, D.C., March, 1987. The authors contributed equally
to the paper, and they gratefully acknowledge the helpful comments
of Henry Levin, Edwin Bridges, Russel Rumberger, and an
-
Vocational Education and Productivity 361
anonymous referee. The authors would like to thank the General
Auto Industry Company for its cooperation and Workers Education and
Training Center of Beijing support for the field work.
NOTES
1. There are various kinds of training programs in vocational
schools, such as training programs for semi-skilled and skilled
workers and middle-level technical personnel (vocational-technical
pro- grams), as well as training programs for cooks, barbers, and
other non-technical personnel. The emphasis in the current reform
is on vocational-technical programs to prepare young people for
factory work.
2. Note that this hypothesis refers to employed vocational and
general education graduates. This study does not consider graduates
who are seeking employment.
3. There are significant differences in curriculum between the
two tracks. For example, general senior high school graduates have
a total of 400 h in Chinese language and 465 h in mathematics while
vocational education graduates have 114 h in Chinese language and
194 h in mathematics. General education students have more courses
in humanities and sciences; vocational education students have more
technical subjects.
4. In recent years, more and more vocational schools have
affiliated themselves with companies. For some areas,
company-affiliated vocational schools represent a large proportion
of secondary vocational schools. For example, in Chongqing City,
among the 115 secondary vocational-technical schools, 105 are
affiliated with companies (Chongqing City Bureau of Labor, 1986).
Given the problems of non-affiliated vocational schools, it is
likely that more and more of such schools will seek affiliation
with companies. Thus this study has focused its attention on
company-affiliated vocational schools.
5. Industrial psychologists usually use job satisfaction as a
proxy for work effort and assume that higher levels of job
satisfaction produce higher levels of work effort (Vroom, 1964;
Srivastva et al., 1977). Following a practice often adopted by
studies on job satisfaction (Kallenberg and Sorensen, 1973; Quinn
and Mandilovitch, 1975) this study has treated the job satisfaction
scale as a continuous scale. The job satisfaction variable was
based on the response to two questions (How well do you like your
job? On the whole are you satisfied with your job?) and a statement
(I am bored with my job).
6. For the relevance of these variables, see Porter and Steers
(1973) March and Simon (19.58). Slavin (1983) and Chew (1983) Price
and Mueller (1986). In this study, these variables were constructed
from responses to questions or statements in a survey. Promotionof
opportunities was based on the responses to What do you think is
the possibility that your wage level will be raised?, Think of the
opportunities for promotion that you have now. Which of the
following statements best describes your situation? and the
promotion responses to How much does your job allow you to take
part in decisions that affect you?, How much autonomy do you enjoy
as to how you do your work? In the production process, my job
allows me to make a lot of decisions on my own. Leader-member
relationship was based on the responses to Does your group leader
usually give you recognition for work well done? and I like my
group leader very much and We are encouraged to speak our mind,
even if it means disagreeing with our group leader. Peer
relationship was based on responses to Do you like your fellow
workers in your work group?, Are there opportunities to develop
close friendships among workers in your group?, and A harmonious
and friendly atmosphere prevails among the workers in my work
group. Co-operation among workers was based on responses to Workers
in our group always cooperate with one another and help one another
in our work, Workers with knowledge and skills in our group always
share their knowledge and skills with those without the knowledge
and skills, and When you run into problems in your work, can you
get help from your fellow workers in your group?.
7. This company was chosen because it met the conditions for our
study: it has semi-skilled and skilled workers engaged in
industrial work, it has a good representation of general education
graduates and vocational-technical education graduates among these
workers, and it has vocational-technical schools with welI
established programs. The access to, and co-operation from, the
company were also important considerations.
8. The issue of selection bias should be considered seriously.
In particular, the general education graduates and vocational
education graduates in the company may have different ability so
that the impact of type of education on productivity may reflect
the influences of both skill training and ability. It is thus
desirable to include measures of ability in the analysis to control
for selection bias. However ability measures such as IQ or test
scores are not available. Nevertheless, interviews with the
-
362 Economics of Education Review
administrators of the vocational technical school affiliated
with the Beijing General Auto-Industry Company indicate that there
should not be significant differences in the ability levels of the
two groups of graduates in our sample. The great majority of the
workers in our sample graduated from junior high schools before
1980. At that time, there were very few vocational-technical
schools at the upper-secondary level, most senior high schools
(over 95%) offered a general education curricuhtm. Only 4% of these
general education graduates succeeded in getting into college. The
vocational- technical school afftiated with the company selected
junior high school graduates whose test scores in the graduating
examination were comparable to those who enroll in general
education senior high schools. The mean of the test scores of the
students enrolled in the vocational-technical school affiliated
with the company was about the same as that of the students
enrolled in the general education senior high schools. Since only a
very small percentage of high school graduates (presumably the most
able) proceeded onto college, the test scores of the general
education graduates in our sample (who did not attend college)
should be comparable to those of the vocational-technical education
graduates in our sample. Thus it has assumed that the two groups of
graduates in our sample are of similar ability. The impact of rype
ofeducarion on productivity can then be attributed to the type of
skill training. Finally, the lack of data on the background of
individuals does not permit a probit analysis of self selection in
secondary education (Willis and Rosen, 1979).
9. It is the policy of the company to assign graduates to jobs
directly related to their training. In the questionnaire, graduates
were asked if they held jobs related to their previous training.
Ninety percent of the vocational-technical school graduates
responded in the affirmative. A distinction can be made between
vocational graduates holding jobs related to their previous
training (users) and graduates holding jobs not related to their
previous training (non-users). For a given job, it is possible that
a user performs better than a non-user. Since graduates from
company-affiliated vocational schools are usually assigned to jobs
in the company related to their training, a high percentage of such
graduates are users. But graduates of non-company affiliated
vocational schools may not be able to find jobs related to their
training; such schools will have a higher percentage of non-users.
While our study has focused on users, non-users can be the subjects
of further research.
10. Reliability and validity tests were conducted on the
measures for job satisfaction, peer relationship, leader-member
relationship, work autonomy, perceived promotional opportunities,
and group co- operation. For the pilot study, both the Cronbachs
alpha and the inter-item correlation coefficient had values well
above 0.7. Also, intra-scale correlations are stronger than
inter-scale correlations. Thus the measures in this study are both
reliable and valid.
11. In the regressions, the five variables for peer
relationship, leader-member relationship, promotional opportunity,
work autonomy, and co-operation among workers were measured by a
continuous scale from 1 to 5. This measurement procedure is
analogous to the one commonly used for job satisfaction. Separate
equations were also estimated using categorical variables for these
job characteristics. The results were similar to those in Equations
1.1 and 1.2. Also, additional statistical analyses indicated that
there was very low (and insignificant) correlation between the
error terms of equations 1.1 and 1.2 and the exogeneous variables
in these two equations. This justified the use of
ordinary-least-squares estimation of the recursive system.
12. According to interviews with the administrators of the
vocational-technical school affiliated with the Beijing General
Auto Industry Company, the average cost of a vocational-technical
school student is about 1000 yuan per year, while that for a
general senior high school student is about 210 yuan per year; and
the average net profit produced by each worker in the factories (in
which our sample was drawn) is 7500 yuan per year. Assuming that
vocational-technical school graduates are 8% more productive than
general senior high school graduates and this productivity
advantage lasts for 8 years, and assuming a discount rate of 5%,
one can derive .a benefit-cost ratio of 1.62 to 1, showing that
vocational-technical schools affiliated with companies are more
cost-effective in economic terms. It should be recognized that this
is only a simplified simulation. In subsequent research, the costs
of different types of vocational-technical schools should be
studied.
13. It should be added that company schools are a policy option
that is limited by the number of large companies that have large
training needs. Small-medium sized companies may have to use CO-
operative training efforts or free-standing schools with closely
managed linkages.
REFERENCES
ALCHIAN, A. and DEMSETZ, H. (1972) Production, information
costs, and economic organization. Am. Econ. Rev. 62.777-795.
BECKER, G. (1964) Humun Cupiraf. New York: Columbia University
Press. BOLINO, A.C. and URI, N.D. (1982) Vocational education: a
human capital/productivity nexus? J.
Behuv. Econ. 11, l-32.
-
Vocational Education and Productivity 363
CHEW, R.G. (1983) A proposed model of the turnover process for
staff registered nurses in a hospital setting. (Dissertation
submitted to the Stanford University School of Education.)
Stanford, CA: Stanford University.
CHONGQING Cm BUREAU OF LABOR (1986) Developing
company-affiliated technical schools. Paper Dresented at the
National Conference on Vocational-Technical Education, Beiiina.
China.
C&AZZINI, A.J. (1968) The decision to invest in vocational
education: an analysis df c&t and benefit. 1. Hum. Resour.
(Supplement) 3, 88-120.
FORSYIH, D. (1983) Inrroducrion ro Group Dynamics. Monterey, CA:
Brooks/Cole Publishing Co. FREDLAND, J.E. and Lrrrt.~, R.D. (1980)
Long-term returns to vocational training: evidence from
military sources. 1. Hum. Resour. 15,49-66. FREEMAN, R.B. (1974)
Occupational training in proprietary schools and technical
institutes. Rev. Econ.
and Srurisr. 310-318. FULLER, W. (1970) Education, training, and
productivity: a study of skilled workers in two factories in
South India. (Dissertation.) Stanford, CA: Stanford University.
FULLER, W. (1976) More evidence supporting the demise of
pre-employment vocational trade training: a
case study of a factory in India. Camp. Educ. Rev. 20, 30-41.
GODFREY, M. (1977) Education, training, productivity and income: A
Kenyan case study (Discussion
Paper No. 253). Nairobi, Kenya: Institute for Development
Studies, University of Nairobi. GRASSO, J. and SHFA, J. (1978)
Vocarional Education and Training: Impact on Youth. Carnegie
Council
on Policy Studies in High School Education. GUSTMAN, A.L. and
STEINMEIER, T.L. (1982) The relation between vocational training in
high school
and economic outcomes. Indusr. Lab. Rev. 36, 73-87. Hu, T.W.,
LEE, M.L. and STROMSDORFER, E.W. (1971) Economic returns to
vocational and
comprehensive high school graduates. J. Hum. Resour. 6, 25-50.
Hu, T.W. and STROMSCXJRFER, E.W. (1979) Cost-benefit analysis of
vocational education. In Handbook
of Vocarional Educarion Evaluation (Edited by ABRAMSON, T.,
TITLE, C. and COHEN, L.), pp. 195-215. Sage Publications.
KALLENBERG, A. and SORENSEN, A. (1973) The measurement of the
effects of overtraining on job attitudes. Sociological Methods and
Research 2, 215-238.
LEVIN, H.M. (1984) Improving productivity through education and
technology (Project Report No. 84- A25). Stanford, CA: Institute of
Research on Educational Finance and Governance, School of
Education, Stanford University.
Lt. C., LAN, R. and LIN, M. (1985) Several problems in the
development of secondary vocational and technical education. Educ.
Econ. December. 29-32. In Chinese.
LI, W.L., SEITZ, P., LEWIS, M. and MER~ENS, D. (1981) Vocational
Educarion Graduare in rhe Labor Marker Experiences of Young
Workers. Columbus, OH: The National Center for Research in
Vocational Education, Ohio State University.
MARCH, J.G. and SIMON, H.A. (1958) Organizarions. New York: John
Wiley. MEYER, R.H. (1982) An Economic Analysis of High School
Vocarional Educarion. The Urban Institute
and Harvard University. MEYER, R.H. and WISE, D.A. (1982) High
school preparation and early labor force experience. In The
Yourh Labor Marker Problem (Edited by FREEMAN, R. and ELLWOOD,
D.), pp. 277-287. Chicago: University of Chicago Press.
MINCER, J. (1974) Schooling, Experience, and Earnings. New York:
Columbia University Press. MINISTRY OF EoucanoN, MINISTRY OF LABOR
AND PERSONNEL, MINISTRY OF FINANCE, STATE PLANNING
COMMISSION (1983) Informorion About rhe Developmenr of
VocarionallTechnical Education. Renmin- ribao (May),Beijing,-
China. In Chinese. . _
NAnONAL BUREAU OF STAnsncs (1986) Starisrical Yearbook, 1986.
Beijing, China: Statistics Press of China. In Chinese.
NATIONAL PEOPLES CONGRESS OF CHINA (1986) Document of rhe Fourrh
Session of rhe Sixrh National Connress of rhe Peoples Republic of
China. Beiiine. China: Peooles Press.
PENCA~EL, J.-(1977) Gork efiort, on-ihe-job scredn&, and
alteriative methods of renumeration. In Research in Labor Economics
(Edited by EHRENBERG, R.), Vol. I. pp. 225-259. Greenwich, CT: JAI
Press.
PEOPLES PRESS, CHINA (1985) Documenti on rhe Reform of rhe
Educarional Sysrem, pp. 1-18. Beijing, China: author. In
Chinese.
PRICE, J.L. and MUELLER, C.W. (1986) Handbook of Ornanizarional
Measuremenr. Marsfield. MA: Pit&an Publishing Inc. .
, I
PORTER, L.W. and STEERS, R.M. (1973) Organizational, work, and
personal factors in employee turnover and job absenteeism. Psychol.
Bull. 80, 151-176.
PSACHAROWULO$ G. and LOXLEY ,- W. (1985) Diversified Secondary
Education and Development: Evidence from Columbiu und Tanzuniu.
Baltimore, MD: John Hopkins University Press.
-
364 Economics of Education Review
QUINN, R. and MANDILOVITCH, M. (1975) Education and Job
Satisfaction: A Questionable Payoff. Ann Arbor, MI: Survey Research
Center, University of Michigan.
RUMBEIGER. R.W. and DAYMONT, T.N. (1984) The economic value of
academic and vocational training acquired in high school. In Youth
and the Labor Market: Analyses of the National Longitudinal Survey
(Edited by BORUS, M.), pp. 157-191. Kalamazoo, MI: W.E. Upjohn
Institute for Employment Research.
SCHMUCK, R. and SCHMUCK, P. (1983) Group Processes in Classroom.
Iowa: W.C. Brown Publishing Co. !&IIJLTZ, T. (l%l) Investment
in human capital. Am. Econ. Rev. 51, 1-17. SIGW, M.E. (1981) Group
Dynamics, , The Psychology of Small Group Behavior. New York:
McGraw
Hill. SHEPPARD, H. and HERRICK, N. (1972) Where Have All the
Robots Gone? Worker Dissatisfaction in the
1970s. New York: Free Press. SLA~IN, R.E. (1983) Co-operative
Learning. New York & London: Longman, Inc. Smtv~sna, S. et al.
(1977) Job Satisfaction and Producrivity. Kent, OH: Comparative
Administration
Research Institute, Kent State University. TANNEN, M.B. (1983)
Vocational education and earnings for white males: New evidence
from
longitudinal data. Southern Economic Journal SO. 369-384.
T~ussrc, M.K. (1968) An economic analysis of vocational education
in the New York City high schools.
J. Hum. Rex&. (&pplement) 3,59:87. _ -
TSANG, M.C. (1987) The impact of underutilization of education
on productivity: a case study of the U.S. Bell companies. Econ.
Educ. Rev. 6, 239-254.
TSANG, M.C. and LEVIN, H.M. (1985) The economics of
overeducation. Econ. Educ. Rev. 4, 93-104. TSANG. M.C.. RUMBERGER.
R. and LEVIN, H. (1991) The impact of surolus schooling on
productivity.
lndu&ial delations (forthcoming). . _ .
VROOM, V. (1964) Work Motivation. New York: John Wiley. WILLIS,
R. and ROSEN, S. (1979) Education and self-selection. J. polit.
Econ. 87, (part 2), S7-S36.