wps1920.pdf
January 1998
EDUCATION AND EARNINGS IN A TRANSITION ECONOMY
THE CASE OF VIETNAM
Peter R. Moock, Harry Anthony Patrinos and Meera Venkataraman

Education and Earnings in a Transition Economy
The Case of Vietnam
Peter R. Moock, Harry Anthony Patrinos and Meera Venkataraman
ABSTRACT
The transition from a centrally planned to a market economy is likely to have a strong
impact on the labor market, on relative earnings and on the returns to education. Major economic
reforms in Vietnam since 1986 (the policy known as Doi Moi) have included a number of
measures to liberalize the labor market. It is too soon to assess the full impact of these reforms,
but this paper represents a first attempt to analyze the returns to education, on the basis of
earnings in 1992-93 (collected in the first Vietnam Living Standards Survey). It represents one of
the first countrywide analyses of the monetary benefits of schooling in Vietnam at a time when the
labor market was in transition. On average the estimates of rates of return are still relatively low,
which is to be expected since salary reforms were not introduced until 1993. Private rates of
return to primary and university education average 13 and 11 percent, but only 4 to 5 percent at
secondary and vocational levels. Returns to higher education are higher for females (12 percent)
than for males (10 percent), and public sector workers have higher returns on average than those
in the private sector. Evidence from other transition economies suggests that returns are likely to
increase as labor market reforms take full effect. The results support this hypothesis: returns for
younger Vietnamese workers are considerably higher than for older workers (14 percent
compared with only 4 percent). There are several policy implications: (i) It is important to
monitor future earnings and labor market trends, as updates of this analysis could provide more
robust estimates of the effects of transition on earnings and returns to education. (ii) At a time
when the Vietnamese government is reassessing its pricing policy, the fact that private rates of
return to higher education are relatively high suggests that there is scope for greater cost
recovery. (iii) Efforts to improve efficiency in secondary and higher education could increase the
rate of return by lowering costs.

Education and Earnings in a Transition Economy
The Case of Vietnam
Peter R. Moock, Harry Anthony Patrinos and Meera Venkataraman∗
INTRODUCTION
The most recent study of the benefits of education in Vietnam appeared almost 30 years
ago (Stroup and Hargrove 1969), and this covered only a part of what was then South Vietnam.
Since that time Vietnam has gone through a process of reunification and rebuilding, and now a
realignment of the economy towards a more market-oriented approach. Economic transition has
led to certain problems in the labor market. The education system that may have served the
country well in a command economy will need to be adapted to serve the needs of a market
economy.
There is a strong commitment to education in Vietnam. This is reflected in high levels of
schooling participation and adult literacy that exceed neighboring countries and even some of the
more advanced Asian nations (World Bank 1995a; Tan and Mingat 1992). In the late 1980s,
however, enrollments in pre-primary, primary and especially secondary schooling declined (World
Bank 1997). Opening up the economy has provided many new income-earning opportunities,

∗ This paper was written as background for the World Bank’s Vietnam Education Financing Sector Study.
A preliminary version of this paper was presented at the Microeconomics: Theory and Applications
Workshop at the University of Illinois at Chicago, April 11, 1997, where useful comments were received,
especially from Carmela Chiswick and Barry Chiswick. Comments also received from Christopher Shaw
and Maureen Woodhall.

thus increasing the opportunity cost of schooling and perhaps reducing rates of return to
schooling.
Education is an important determinant of earnings in market economies. The higher is
one’s level of education, the higher is that person’s starting salary and the steeper the rise in
earnings during the early working life. The evidence from centrally planned economies is
relatively limited, including from economies in transition. Nevertheless, it has been shown that in
centrally planned economies returns to schooling are usually low, for example, in China (see
Jamison and van der Gaag 1987), but they tend to increase as market reforms take place, for
example, in Slovenia (Orazem and Vodopivec 1995) and Hungary (Varga 1995). This paper
examines the returns to education in Vietnam in the early stages of transition (based on earnings
data fro 1992-93). It is hoped to update these estimates at a later date, to examine the effects of
market liberalization and education reforms.
KEY LABOR MARKET REFORMS
The Vietnamese government's strategy for development places high priority on investment
in human capital. But there are problems. There is a high population growth rate and a large
rural labor force that is mostly poor. In addition, there are large numbers of migrants, refugees,
disabled veterans and demobilized soldiers that must be integrated into the labor market. The
largely successful economic reforms have resulted in the retrenchment of up to 1.5 million
workers from government service and state owned enterprises (World Bank 1995b).
Unemployment has increased and real wages have declined by 30 to 50 percent (ILO 1994).
Moreover, modern labor practices are new to Vietnam. In 1995, a labor law was passed and a
2

labor code was published that covers a wide range of labor issues including labor contracts,
collective bargaining, social insurance, working conditions and training (for a full description, see
World Bank 1995b).
In 1990 the government began the process of dismantling the old public sector wage
structure. No longer would workers be paid mainly according to length of service, nor
guaranteed employment for life. A plan was announced to place all government workers on
contract, but this process has been slow (Hiebert 1993; Norlund 1993).
In 1993 the wages of public servants were partly monetized. Prior to this only a fraction
of the salaries of civil servants or employees of state-owned enterprises was paid in cash. Up to
90 percent was received in bonuses and in-kind payments (ILO 1994). State workers no longer
receive access to rationed goods at artificial prices, nor do new state employees receive fringe
benefits such as subsidized housing, health services, education and social insurance. The new
structure rewards workers according to skill, education, responsibility and job performance. The
government also announced plans to decompress public sector wages. The full impact of these
reforms may not be seen for years, however, since those hired prior to 1994 are largely exempted.
Private firms are now free to set wages without government interference. In the parastatal
sector, subsidies were reduced in 1989 and further privatization measures introduced in 1990.
Still, state owned enterprises continue to negotiate salary levels for employees with their parent
ministries rather than with workers directly (Lindauer and Haughton 1996).
3

Although labor mobility is not officially restricted, work permits are legally required in
most places, and those wishing to move their place of residence are supposed to obtain permits.
These regulations do not, in practice, prevent many rural inhabitants from ‘trying their luck’ in
urban labor markets. In the main cities job seekers queue each morning along central boulevards,
joining what has come to be known as the "labor bazaar" (World Bank 1997). These informal
labor markets emerged spontaneously in response to uncontrolled rural-urban migration, and they
have given rise to illegal employment services and charges of exploitation. The Ministry of Labor,
Invalids and Social Affairs estimates the number of rural job seekers at about 170,000 in Hanoi
and 800,000 in Ho Chi Minh City.
Although government sets an official minimum wage, it is only minimally effective given
the apparent reluctance to enforce this or any other labor regulation when it comes to domestic
firms. However, for foreign and joint ventures, the minimum wage almost certainly does have an
effect. Not only is it easier to monitor the relatively small number of foreign firms, but the
minimum wage is set much higher for foreign firms than for local firms. Between 1993 and 1996,
the minimum wage for all enterprises was Vietnamese dong (VND) 120,000 ($11) per month.
The minimum wage for firms with any degree of foreign ownership was $35 per month in Hanoi
and Ho Chi Minh City, and $30 elsewhere. A new minimum wage law took effect in July 1996.
It raises the minimum wage for foreign firms to $45 in Ho Chi Minh City and Hanoi; $40 in
Haiphong, Vinh City, Hue, Danang, Bienhoa, Can Tho, Hatay, Nha Trang and Vung Tau; and
$35 in other places.
BENEFITS OF SCHOOLING
4

Returns to investments in schooling are estimated here using the 1992-93 Vietnam Living
Standards Survey (VLSS). The VLSS was carried out jointly by Vietnam’s State Planning
Committee (now the Ministry of Planning and Investment) and the General Statistics Office, from
October 1992 to October 1993, with funding from the United Nations Development Programme
and the Swedish International Development Authority. It was the first nationally representative,
multipurpose household survey ever done in Vietnam covering a broad range of social and
economic indicators. Information was collected from about 23,000 people living in 4,800
households representing urban and rural areas in each of Vietnam’s seven geographic regions (for
more details, see Vietnam 1994). However, since many of the labor market reforms did not come
into effect until later (for example, salary reform did not take place until 1993), this analysis uses
data covering a period relatively early in the reform process.
Returns to schooling are estimated here using a variety of methods. This is necessary
given the limited sample size of the wage earning population in Vietnam. One method used is the
human capital earnings function (Mincer 1974):
lnY
2
i = α + βSi + γ
1EXPi + γ
2EXPi + εi
(1)
where S
2
i is the number of years of schooling of individual i, and EXPi and EXPi are years of
experience and its square. Hours worked per week is added as a compensatory factor. In this
semi-log specification the coefficient on S (β) is interpreted as the average private rate of return to
one additional year of schooling, regardless of the level of schooling.
The earnings function method is used to estimate average private returns to different levels
of schooling by converting the continuous years of schooling variable (S) into a series of dummy
5

variables representing the different levels of schooling. After fitting the extended earnings
function:
lnY
2
i = α + β1PRIMi + β2SECi +β3VOCi + β4UNIVi + γ
1EXPi + γ
2EXPi + εI
(2)
where PRIMi, SECi, VOCi and UNIVi are dummy variables indicating primary, secondary
academic, secondary vocational and university education completion by individual i, the private
rates of return to different levels of schooling are then calculated as follows:
r(PRIM) = β1 /SPRIM
(3)
r(SEC) = (β2 - β1 )/(SSEC - SPRIM)
(4)
r(VOC) = (β3 - β1 ) /(SVOC - SPRIM)
(5)
r(UNIV) = (β4 - β2 )/(SUNIV - SSEC)
(6)
where SPRIM=6, SSEC=7, SVOC =6 and SUNIV=4, the average number of years of schooling for the four
levels of education. However, it is incorrect to assume that primary school graduates forego
earnings for the entire duration of their studies. Therefore, only one year of foregone earnings are
assumed for primary school graduates, for example, SPRIM=1.
A more complete method is the net earnings profile method, where detailed age-earnings
profiles by level of education are used and the rate of return is computed as the discount rate that
equates the stream of education benefits to the stream of educational costs is computed:
n n
Σ Bt/ (1+r)t = Σ Ct/(1+r)t
(7)
t=0 t=0
where Bt and Ct are the average benefits and costs of the particular educational investment (for
example, completing secondary education following completion of primary education) in year t, n
is the expected life of the investment, and r is the rate of return.
6

The benefit of an educational investment is measured as the average additional earnings
received by (for example) the graduates of secondary school, over and above the graduates of
primary school. The cost of the investment comprises direct and indirect costs. The direct costs
for the individual include all expenditures related to school attendance, and for society, the full
resource costs of providing the educational service, including any subsidized costs not borne by
the individual or the individual’s family. The indirect costs are the average earnings foregone as a
result of the investment.
This method, while straightforward and requiring fewer “as if” assumptions than other,
more elegant methods used to estimate rates of return to educational investments, is data
demanding. One must have sufficient numbers of observations in all age-education level cells to
be able to construct well-behaved age-earnings profiles (that is, profiles that are noncrossing and
concave to the horizontal axis).
In the absence of good outcome data (and this is usually the case in a transition economy
where wages may not well reflect productivity differences), much can still be said with cost data
alone. For example, in considering the benefits of education, “switching values” are very useful in
the absence of reliable data on earnings differentials. The analysis, sometimes referred to as
“reverse cost-benefit analysis,” consists of estimating the productivity differential required in
order for an education investment to be acceptable in terms of its rate of return. The question is
framed in the following manner: given the cost of producing a graduate, what productivity
differential would be required for the investment to exhibit a 10 percent social rate of return
7

(assuming 10 percent to be the social opportunity cost of capital)? The reverse cost-benefit
method can be shown as:
DS/S-1 = 0.10[ tS(CS+ WS-1)]
(8)
where DS/S-1 is the required productivity differential between a person who has graduated from
schooling level S and one who has graduated from level S-1; tS is the number of years required to
complete level S (seven years, say, for secondary education, assuming no grade repetition); CS is
the per student direct resource cost of each year of the investment and WS-1 is the earnings
foregone during each year of the investment. This method provides guidance on investment in
education in the absence of reliable earnings information. It is up to the policymaker to decide
whether the estimated required differential is likely to be realized.
EMPLOYMENT AND EARNINGS DIFFERENTIALS
Vietnam is predominantly a rural country, and most Vietnamese make their living off the
land. Two-thirds of all workers are in agriculture. Other large sectors include manufacturing (12
percent) and wholesale and retail trade— including restaurants and hotels (11 percent) (World
Bank 1997).
Most workers in Vietnam (79 percent) are self-employed, and 11 percent are employed in
private firms. The public sector, broadly speaking, accounts for 9 percent of the labor force
(World Bank 1996). Estimates here of the returns to schooling are limited to the wage labor
force, thus excluding the large self-employed sector for which the VLSS does not collect earnings
information. The fact that this analysis relates only to 20 percent of the labor force must,
therefore, be kept in mind.
8

The Vietnamese wage labor force is relatively well educated, especially for a low-income
country (Table 1). The mean age of the sample is 31 years. The average number of school years
completed is 7.9, years of experience 17 and hours worked per week 46. About a quarter (22
percent) of the sample have no education or incomplete primary, and half have completed just the
primary cycle. Only 8 and 12 percent of the labor force have secondary academic and secondary
vocational education, and 7 percent have completed university. Over 40 percent of those in the
wage labor force are employed in the public sector broadly speaking (which includes government,
state companies, cooperatives and social organizations). Mean monthly earnings are about VND
152,000 ($14).
Table 1: Vietnam: Means of Selected Variables by Sex and Sector of Employment
All
Public
Private
Variable
Total Males Females Total Males Females Total Males Females
Age (years)
30.8
31.2
30.2
34.1
35.7
32.2
28.4
28.5
28.2
Years of schooling
7.9
7.9
7.8
10.5
10.5
10.4
6.0
6.3
5.4
Years of experience
17.0
17.4
16.4
17.7
19.2
15.8
16.4
16.2
16.9
Education level (0,1)*
No education
0.22
0.21
0.24
0.06
0.06
0.07
0.34
0.31
0.40
Primary
0.50
0.53
0.44
0.39
0.43
0.35
0.57
0.59
0.53
Secondary
0.08
0.08
0.08
0.12
0.11
0.12
0.06
0.07
0.04
Vocational
0.12
0.10
0.16
0.26
0.22
0.31
0.03
0.03
0.03
University
0.07
0.08
0.07
0.17
0.19
0.15
0.01
0.01
0.00
Earnings/month
152
170
124
167
188
143
140
159
107
('000s VND)
Public sector
0.42
0.38
0.48
1.00
1.00
1.00
0.00
0.00
0.00
Hours-worked/week 46.2
46.8
45.5
43.5
43.2
43.8
48.2
48.9
47.0
N
2,259 1,355
904
950
516
434
1,309
839
470
Source: VLSS 1992-93
* Highest level attained
9

Males are about one year older than females and their earnings nearly 40 percent higher
(VND 170,000 vs. VND 124,000). Men and women have nearly the same educational
attainment. Private sector workers are much younger than public sector workers and work nearly
five hours more per week. Educational attainment in the private sector is much lower than in the
public sector (more than four year’s difference). In the public sector, even small state enterprises
employ workers with at least ten years of schooling (Ronnas and Sjoberg 1995). This is a result
of the past policy of government’s allocating labor with more than primary schooling. Graduates
of secondary and tertiary education, including higher-level vocational graduates, were assigned
jobs in the public sector. Only primary school graduates and those without any education
operated in a labor market of sorts. A reliance on market signals to allocate educated labor was
announced, however, as Government policy at the 1991 Party Congress. Few primary school
graduates work in the public sector. Vocational school graduates work almost exclusively in the
public sector. Overall earnings are much higher in the public sector.
Economic restructuring policies, which began in the mid-1980s in Vietnam, accord a lead
role to the private sector in the transition to a market economy and in overall economic
development. Employment growth is highest in the private sector, which needs to absorb not only
new labor market entrants but also the labor shed by state enterprises. During 1989-1993, about
4.7 million new jobs were created in the private sector. Given a reduction of about 900,000 jobs
in the state sector, the net increase was 3.8 million additional jobs (World Bank 1995b). Private
sector workers are paid less than those are in the public sector. In the public sector, workers with
little or no education have relatively high earnings (Table 2). Remuneration rates in the public
10

sector are compressed, resulting in apparent overpayment of workers with less than a primary
education.
Table 2: Vietnam: Monthly Earnings by Economic Sector (VND per month)
Educational level*
Public
Private
No education
187,514
117,641
(59)
(446)
Primary
154,690
146,558
(375)
(747)
Secondary
157,871
182,930
(111)
(75)
Vocational
146,891
187,137
(246)
(34)
University
227,452
247,837
(159)
(7)
Overall
167,259
140,385
(950)
(1,309)
Source: VLSS 1992-93
Note: Numbers in parentheses
* Highest level attained
11

ESTIMATES OF THE RETURNS TO SCHOOLING
The reverse cost-benefit method (switching analysis) suggests that education was a
marginal investment in Vietnam in the early 1990s. Actual earnings differentials are presented in
Table 3. These show the absolute amount in VND, by which holders of a particular degree or
level of schooling surpass the earnings of those with the next lowest level of schooling. For
example, primary school leavers earn VND 276,000 more per year than do those with no
education. The next column presents the amount in VND by which primary school leavers would
need to surpass on average the earnings of those with no education in order for the investment to
realize a 10 percent social rate of return. Thus, primary school leavers already earn on average
more than what is needed to realize a 10 percent social rate of return. However, because of the
level of costs, required earnings differentials at higher schooling levels are very high, especially for
vocational education.
12

Table 3: Reverse Cost-Benefit Analysis: Earnings Differentials Required to Realize 10 Percent
Rate of Return
Variables in Equation (8)
(CS)
(WS )
(tS )
(DS )
(WS - WS-1) Observed as
Education level
Cost per
Average Assumed
Required
Observed
percent of
(and investment
student
earnings
cost
differential
differential
required
reference level)
year (VND'000)
periods
(VND'000)
(VND'000)
(VND'000)
(VND'000)
(years)
None
...
...
...
...
...
1,512
Primary
(vs none)
192
1,788
CS: 5
247
276
112%
WS: 1
Secondary
(vs primary)
436
2,016
7
1,557
228
15%
VOTECH
(vs primary)
1,331
1,824
5
1,560
36
2%
Tertiary
(vs secondary)
2,745
2,736
4
1,904
912
48%
Source: VLSS 1992-93.
The results of estimating a simple earnings function show education to be a marginal
investment (Table 4). Overall, the average private rate of return on another year of schooling is 5
percent. This is very different from the previous and dated analysis limited to rural South
Vietnam (Stroup and Hargrove 1969). An additional year of schooling at that time in South
Vietnam led to an increase in earnings of 16.8 percent in 1964.
13

Table 4: Vietnam: Earnings Functions
(dependent variable is the natural log of monthly earnings)
Variable
All
Males
Females
Constant
0.775
1.045
0.583
Years of schooling
0.048
0.034
0.068
(8.4)
(4.5)
(7.6)
Years of experience
0.064
0.059
0.065
(10.1)
(6.9)
(6.8)
Experience-squared
-0.001
-0.001
-0.001
(8.3)
(5.6)
(5.5)
Log hours-worked
0.726
0.722
0.697
(11.7)
(9.1)
(7.1)
R-squared
0.112
0.095
0.137
N
2,259
1,355
904
Source: VLSS 1992-93
Note: t-statistics in parentheses
When earnings functions are estimated separately, males experience lower returns to
schooling than do females. This is similar to the pattern found in China in the mid 1980s (Jamison
and van der Gaag 1987). In Vietnam, males receive a 3 percent increase in earnings for every
year of schooling; females experience a 7 percent increase.
The estimates for Vietnam are low compared with the returns to education estimated for
other developing countries. Worldwide, another year of schooling increases earnings by about 10
percent (Psacharopoulos 1994). But estimated returns to schooling are low in general in centrally
planned and transition economies. For example, in China the estimates of the returns to schooling
ranged from 1 to 5 percent in the mid 1980s (Jamison and van der Gaag 1987; Byron and
Manolato 1990; Meng 1995), and 3 percent in 1989 (Xie and Hannum 1996). Gregory and Meng
(1995), in a very useful analysis of rural industrial data, disaggregated their sample into two
groups: those who were allocated their jobs and those who found them on their own. Returns to
schooling for those allocated their jobs are small and not statistically significant. But for those
14

who found employment on their own, the returns are large and significant. Xie and Hannum
(1996) find that membership in the communist party in China accounts for a high proportion of
earnings and argue that this is a form of human capital in that workers use membership to gain
access to jobs allowing them to make use of their skills. Since in Vietnam the majority of workers
with secondary, vocational or university education were assigned to jobs in the public sector an
average rate of return of 5 percent to one additional year of schooling is entirely consistent with
these findings for China.
Low rates of return to schooling are found in other command economies: Hungary, 4.3
percent in 1987, and Poland, 2.9 percent in 1986 (Psacharopoulos 1994). Scarce, over-time
evidence, however, seems to be showing that successful reform will eventually lead to higher
returns. An empirical analysis of changes in the wage structure in Slovenia between 1987 and
1991 reveals that the returns to human capital rise dramatically during transition. Workers with
four years of university education gained the most in relative earnings, followed by those with two
years of university. The education group that gained the least, however, relative to the least
educated, were holders of vocational degrees (Orazem and Vodopivec 1995). In Hungary, the
private rate of return almost doubled in secondary education between 1971 and 1993, and there
was a three-and-a-half-fold increase in the returns to higher education (Varga 1995).
Workers in the public sector in Vietnam realize higher private rates of return to schooling
than do private sector workers (Table 5). This may not, however, signal that productivity is
better rewarded in the public sector. It may simply be a relic of the past policy of allocating
educated labor to public sector positions. Most of the workers in the private sector have primary
15

education or less, while public sector workers usually have at least a primary education.
However, the fact that in the public sector workers with no or very little education earned more,
on average, than primary, secondary or vocational school graduates (Table 2) suggests that there
are significant distortions in public sector pay. It is likely that factors other than education
(perhaps, as in China, membership in the Communist Party) had an impact on public sector pay in
Vietnam, as in other command economies.
Table 5: Vietnam: Earnings Functions by Sector of Employment (both sexes)
Sector of Employment
Public
Private
Constant
0.265
1.168
Years of schooling
0.062
0.039
(6.3)
(4.0)
Years of experience
0.046
0.072
(4.4)
(8.4)
Experience-squared
-0.001
-0.001
(3.3)
(7.3)
Log hours-worked
0.878
0.617
(10.5)
(6.7)
R-squared
0.145
0.085
N
950
1,309
Source: VLSS 1992-93
Note: t-statistics in parentheses
The earnings function is fitted using dummy variables for those who have completed a
particular level of education (Table 6). The results are reasonable, except that for males the
coefficient on primary education is not significant. Also it would appear that those with less than
a complete primary education earn higher wages in the public sector than do workers with a
complete primary education. The results from the earnings functions can be used to estimate the
earnings premium associated with each level of education and the private rate of return to each
level of schooling.
16

Table 6: Vietnam: Earnings Functions with Schooling Levels
Variables
All
Males
Females
Public
Privatea
Constant
0.870
1.181
0.617
0.845
1.220
Primary
0.134
0.035*
0.214
-0.278
0.233
(2.2)
(0.5)
(2.3)
(2.0)
(3.2)
Secondary
0.325
0.269
0.423
0.404
0.260
(3.8)
(2.4)
(3.1)
(3.7)
(1.9)
Vocational
0.207
0.240
0.259
0.276
---
(2.8)
(2.3)
(2.4)
(3.4)
University
0.437
0.414
0.488
0.429
---
(3.8)
(2.8)
(2.7)
(3.5)
Experience
0.064
0.057
0.066
0.053
0.072
(9.9)
(6.6)
(6.7)
(5.1)
(8.3)
Experience-squared
-0.001
-0.001
-0.001
-0.001
-0.001
(8.4)
(5.5)
(5.8)
(4.2)
(7.3)
Log hours-worked
0.750
0.732
0.751
0.888
0.618
(12.1)
(9.3)
(7.5)
(10.8)
(6.8)
R-squared
0.124
0.111
0.142
0.179
0.087
N
2,259
1,355
904
950
1,309
Source: VLSS 1992-93
Note: t-statistics in parentheses; all variables are significant at the 10% level or better, except where
indicated by *; the omitted category for level of education is no education
a Vocational and university coefficients not shown because of an insufficient number of observations
Large earnings premiums translate into relatively high private returns to schooling (Table
7). Overall university is a good private investment. For females primary and university education
are profitable investments. In the public sector, university is a good investment, while primary
education is not. The highest private rate of return is at the primary level for private sector
workers. Secondary academic and secondary vocational education are consistently poor
investments.
Table 7: Vietnam: Private Rates of Return to Schooling by Level of Education (percent)
Educational level
All
Males
Females
Public
Private
Primary (vs less than primary)
13
21
23
Secondary (vs primary)
5
4
6
6
4
17

Vocational (vs primary)
4
5
5
6
University (vs secondary)
11
10
12
11
Source: Based on the earnings functions results presented in Table 6
REGIONAL DIFFERENCES
Evidence of low returns to schooling in Vietnam is not surprising. Analysis of private
enterprises (Ronnas 1992) revealed a negative correlation between education and earnings. This
is attributed to large regional differences in general levels of education and income. The level of
education in the North is higher than in the South, and wage levels in the South are much higher
than in the North. Ronnas argues that the income differences are due to the generally higher
capital-labor ratio and better endowment with machinery and equipment in the South. Other
factors may also be at work such as differences in work discipline, infrastructure and transaction
costs.
The VLSS confirms that schooling levels are higher in the North. On average workers in
the North have about ten years of schooling, while in the South the average worker has less than
seven years of schooling (Figure 1). In terms of earnings, however, workers in the South earn
more than workers in the North: VND182,495 ($17) versus VND94,174 ($9) per month.
Therefore, workers in the South have a third less schooling but earn 50 percent more than
workers in the North.
18

Figure 1: Vietnam: Schooling and Earnings Differences by Region
Earnings ('000s VND)
Schooling (Years)
200
10
150
8
6
100
4
50
2
0
0
North
South
North
South
Source: VLSS 1992-93
Estimating earnings functions separately for workers in the North and the South results in
rates of return to schooling that are similar in both regions (Table 8). On average, an additional
year of schooling will increase earnings by 8 percent in both regions. Experience, however, is
rewarded more in the South than in the North. Also, hours worked have a slightly higher payoff
in the South.
Table 8: Vietnam: Earnings Functions by Region
Variable
North
South
Constant
0.515
0.782
Years of schooling
0.082
0.078
(6.8)
(12.0)
Years of experience
0.041
0.078
(4.1)
(10.2)
Experience-squared
-0.001
-0.001
(3.0)
(8.5)
Log hours worked/week
0.639
0.698
(6.7)
(9.4)
R-squared
0.113
0.177
N
788
1,471
Source: VLSS 1992-93
RECENT COHORTS
19

A cross-sectional sample may not be very instructive in a rapidly changing economy. For
this reason the sample is divided into two groups: recent labor market entrants and workers with
more than six years experience. Focusing on recent graduates allows one to reflect the fact that
reforms may not have affected all workers by 1993. Younger workers are expected to be more
affected by the recent changes as they enter directly into a free market wage economy. In fact,
those with less years of labor market experience receive higher returns to schooling: 14 percent
versus 4 percent for more experienced workers (Table 9). In China, too, the returns to schooling
for younger workers are higher than for more experienced workers, showing that in both Vietnam
and China the returns to schooling increase as newer cohorts enter the labor market (World Bank
1996).
Table 9: Vietnam: Earnings Functions for Recent Labor Market Entrants and
Experienced Workers
Variable
Recent Entrants
Experienced Workers
(less than 6 years)
(more than 5 years)
Constant
0.217
0.792
Years of schooling
0.144
0.044
(5.9)
(7.5)
Years of experience
0.204*
0.055
(1.3)
(6.5)
Experience-squared
0.006*
-0.001
(0.2)
(5.9)
Log hours-worked
0.489
0.759
(2.6)
(11.7)
R-squared
0.138
0.103
N
328
1,931
Source: VLSS 1992-93
Note: t-statistics appear in parentheses; all variables are significant at the 10% level
or better, except where indicated by *
AGE-EARNINGS PROFILES
20

The private rate of return to schooling is also estimated using the complete or discounted
earnings profile method. The age-earnings profiles estimated from the VLSS are not very well
behaved, a problem attributable to the small numbers in the VLSS data set. But after removing
outliers and smoothing out the data, estimation of the private rate of return to schooling results in
high estimates at the primary level and low estimates at the secondary and university levels (Table
10). In terms of pricing policy, were primary education free at the primary level, at least in terms
of direct monetary costs, then the returns to primary schooling would be very high. However,
parents do face direct monetary charges for their children’s enrollment in primary school. The
VLSS collects this information and it is used here to estimate the private rate of return to
schooling.
When the full private direct cost of schooling is added to the calculations, the private rate
of return to primary schooling decreases by 27 percent. Addition of the private costs reduces the
returns by 18 and 20 percent for secondary and university graduates. This reflects the fact that
higher levels of education are relatively more highly subsidized than primary education— an aspect
of pricing policy in Vietnam that benefits the better off at the expense of the poor, and therefore
raises serious issues of equity. Nevertheless, primary education remains a profitable investment
on average. But since the poor are likely to have more children, then the costs of schooling
become a significant burden. Secondary education does not appear to be a very good investment,
but the analysis masks one of the key benefits that individuals derive from secondary schooling—
it allows one to pursue studies at higher schooling levels. This is known as the “option value” of
secondary schooling (World Bank 1991; Weisbrod 1964).
21

Table 10: Vietnam: Rate of Return using Complete Method, 1992-93
Educational Level
Private
Private
Social
(no cost)
(private costs included)
(all costs included)
Primary (vs no education)
18.5
13.5
10.8
Secondary (vs primary)
5.5
4.5
3.8
University (vs secondary)
7.8
6.2
3.0
Source: VLSS 1992-93, using rate of return program (Psacharopoulos 1996)
Social rates of return are highest for primary education. In fact, the earnings data used in
these calculations suggest that only primary schooling was a worthwhile social investment in
Vietnam in the early 1990s. University education had a low social rate of return, but recent
evidence from a survey of higher education graduates indicates that earnings for graduates may
have increased in recent years. The difference between the social and private returns suggests an
over-subsidization of higher education. However, tuition fees were introduced at the university
level in 1993, and the government plans to increase cost recovery in higher education, so this gap
may be reduced in the future.
SUMMARY AND CONCLUSIONS
The returns to schooling in Vietnam are low by international standards. The estimates
come from earnings data for 1992-93. These are the best data currently available, but may be
distorted for a number of reasons. First, public sector pay reflects the past policy of assigning
secondary and higher education graduates to the public sector. Also, we have limited the analysis
to wage earners, while over 80 per cent of the Vietnamese labor force are self-employed and
many have multiple jobs. The estimates of social rates of return, as in most studies, do not take
22

into account externalities, and should therefore be treated as under estimates of the true social
impact of education.
There are signs, however, that the returns will increase as Vietnam’s market reforms take
hold. The high rates of return observed for younger cohorts are one sign. The evidence of other
former command economies that have completed the transition to strong market institutions also
point to higher education returns in Vietnam’s future.
This study has implications for the financing of education in Vietnam, even though it is
based on admittedly poor data. Primary education, the most profitable sub-sector judging from
the estimated social rate-of-return results, is much less subsidized than higher levels. In fact, the
high subsidy levels for higher and, especially, vocational education contributes to the low rates of
return for these sub-sectors. Family contributions to direct cost financing at the primary level are
a heavy burden, especially for the poor, and this is neither socially optimal nor equitable.
Vietnam’s policy makers may consider direct subsidization of poor primary school age children to
ensure that they enroll in school and remain enrolled. A policy of increased cost recovery in
higher education has already been introduced.
There is scope for reform in terms of remuneration policy and recruitment policy in the
public sector. Most university graduates are still recruited into the civil service, artificially
propping up the returns to public higher education. Evidence from the recently completed Higher
Education Graduate Tracer Study (World Bank 1997) suggests that change is occurring rapidly.
23

For example, more and more university graduates are taking positions in the private sector, in
contrast to the situation just three to four years ago, when hardly any graduates did so.
It is important to monitor future earnings and labor market trends. In fact it would be
desirable to update these estimates as soon as more recent earnings data are available. The
Higher Education Graduate Tracer Study conducted in 1996-97 suggests that the labor market is
changing rapidly. Future updates of this analysis, based on more recent data, could not only
provide more robust estimates, but also provide evidence on whether the impact of labor market
reforms is increasing over time, as suggested in this paper. The Vietnamese Government is
currently reassessing pricing policy. It has already begun to increase cost recovery at the higher
education level. Tuition fees were introduced in 1993 and the current policy is to increase cost
recovery, which is likely to reduce the gap between social and private rates of return. If costs
could be reduced— for example by better utilization of teachers and by reducing unit costs in
higher education through encouraging mergers to achieve economies of scale— this could reduce
costs and therefore increase the rate of return.
24

REFERENCES
Byron, Raymond P. and Evelyn Q. Manolato. 1990. "Returns to Education in China." Economic
Development and Cultural Change 38(4): 783-796.
Gregory, R.G. and Xin Meng. 1995. "Wage Determination and Occupational Attainment in the
Rural Industrial Sector of China." Journal of Comparative Economics 21: 353-374.
Hiebert, Murray. 1993. "Wage Revolution: Vietnam Makes Salaries Reflect Responsibilities."
Far Eastern Economic Review (28 September 1993): 95.
ILO. 1994. Viet Nam: Labour and Social Issues in A Transition Economy. Bangkok: ILO East
Asia Multidisciplinary Advisory Team (ILO/EASMAT), ILO Regional Office for Asia and
the Pacific Region.
Jamison, Dean T. and Jacques van der Gaag. 1987. "Education and Earnings in The People's
Republic of China." Economics of Education Review 6(2): 161-166.
Lindauer, David L. and Jonathan Haughton. 1996. "Economic Growth and the Labor Market."
In D. Dapice, J. Haughton and D. Perkins, eds., In Search of the Dragon’s Trail:
Economic Reform in Viet Nam
. Cambridge: Harvard University Press.
Meng, Xin. 1995. "The Role of Education in Wage Determination in China's Rural Industrial
Sector." Education Economics 3(3): 235-248.
Mincer, Jacob. 1974. Schooling, Experience and Earnings. New York: National Bureau of
Economic Research.
Norlund, Irene. 1993. "The Creation of a Labour Market in Vietnam: Legal Framework and
Practices." In C.A. Thayer and D.G. Marr, eds., Vietnam and the Rule of Law (Political
and Social Change Monograph 19). Canberra: Australian National University, Research
School of Pacific Studies, Department of Political and Social Change.
Orazem, Peter F. and Milan Vodopivec. 1995. "Winners and Losers in Transition: Returns to
Education, Experience, and Gender in Slovenia." World Bank Economic Review 9(2):
201-230.
Psacharopoulos, George. 1994. "Returns to Investment in Education: A Global Update." World
Development 22(9): 1325-1343.
Psacharopoulos, George. 1996. “RR Manual: A Program to Estimate the Rate of Return to
Investment in Education.” The World Bank. Washington, D.C.
Ronnas, Per. 1992. Employment Generation through Private Entrepreneurship in Vietnam.
Geneva: SIDA/ILO-ARTEP.
25

Ronnas, Per and Orjan Sjoberg. 1995. Economic Reform, Employment and Labour Market
Policy in Viet Nam. Labour Market Papers 9, Employment Department, International
Labour Office, Geneva.
Stroup, Robert H. and Michael B. Hargrove. 1969. "Earnings and Education in Rural South
Vietnam." Journal of Human Resources 4(2): 215-225.
Tan, Jee-Peng and Alain Mingat. 1992. Education in Asia: A Comparative Study of Cost and
Financing (Regional and Sectoral Studies). Washington, DC: World Bank.
Varga, J. 1995. “Returns to Education in Hungary.” Acta Oeconomica 47 (1-2): 203-16.
Vietnam. 1995. EDUKASIA 95: Market Brief (Viet Nam Chamber of Commerce and Industry
of Viet Nam and Ministry of Education and Training).
Vietnam. 1994. Vietnam Living Standards Survey 1992-1993. Hanoi: State Planning Committee
and General Statistics Office.
Weisbrod, Burton. 1964. “External Benefits of Public Education and Economic Analysis.”
Princeton: Industrial Relations Section Paper.
World Bank. 1997. Vietnam: Education Financing. (A World Bank Country Study).
Washington, DC: World Bank.
World Bank. 1995a. Priorities and Strategies for Education: A World Bank Review
(Development in Practice). Washington, DC: World Bank.
World Bank. 1995b. Viet Nam: Economic Report on Industrialization and Industrial Policy
(Report No. 14645-VN). Washington, DC: World Bank.
World Bank. 1991. “Decision and Change in Thailand: Three Studies in Support of the Seventh
Plan.” Asia Country Department 2, World Bank, Washington, DC.
Xie, Yu and Emily Hannum. 1996. “Regional Variation in Earnings Inequality in Reform-Era
Urban China.” American Journal of Sociology 101(4): 950-992.
26