学校代码： 10246

学 号： 16210680467

硕 士 学 位 论 文
（学术学位）

双城记：越南两个最大经济城市的教育回报率的研究
TALE OF TWO CITIES: THE STUDY OF RETURN TO
EDUCATION IN THE TWO BIGGEST ECONOMIC HUBS
OF VIETNAM

院 系： 经济学院
专 业： 应用经济学（中国经济）
姓 名： Ngoc Ha Truong
指 导 教 师： 李怀露
完 成 日 期： 2018 年 3 月 20 日
 UNIVERSITY CODE: 10246
STUDENT ID: 16210680467

TALE OF TWO CITIES: THE STUDY OF RETURN TO EDUCATION

IN THE TWO BIGGEST ECONOMIC HUBS IN VIETNAM

MA THESIS

Presented in Partial Fulfillment of the

Requirement for the Master Degree by the
Graduate School of Fudan University

By
Ha, Truong

Fudan University
2018

Defense Date: May 18, 2018 EMA Chinese Economy Program

Advisor List

Prof. 陈钊 (CHEN ZHAO)

Prof. 罗长远(LUO CHANGYUAN)
Prof. 奚锡灿(XI XICAN)
CONTENTS

摘 要 .......................................................................................... 1
ABSTRACT ................................................................................. 5
ACKNOWLEDGEMENTS ..... Error! Bookmark not defined.6
INTRODUCTION ....................................................................... 7
1.1 Background ....................................................................................................................... 9
1.2 Scope & structure .............................................................................................................. 12

LITERATURE REVIEW ............................................................ 13

Previous research on Human Capital ...................................................................................... 14
2.1 Mincer equation ................................................................................................................ 16
2.2 Extend Mincer equation .................................................................................................... 16
2.2.1 Ability bias ..................................................................................................................... 16
2.2.2 Selection bias ................................................................................................................. 18
2.3 Return to education in Vietnam studies ............................................................................. 23

DATA & METHODOLOGY ..................................................... 25

3.1 Data ................................................................................................................................... 25
3.2 Methodology ..................................................................................................................... 26
3.2.1 Dependent variable & Independent variables ................................................................ 26
3.2.2 Control variables ............................................................................................................ 29
3.2.3 Bias issues ...................................................................................................................... 30

EMPIRICAL FINDINGS .......................................................... 33

4.1 Descriptive results ............................................................................................................. 33
4.2 Return to education in Vietnam: Hanoi vs. Ho Chi Minh City ......................................... 39

CONCLUTION.......................................................................... 43
LIMITATIONS........................................................................... 46
REFFERENCES ........................................................................ 47
FIGURES ................................................................................... 52
APPENDICES ........................................................................... 55
 摘 要

关键词: 教育回报率, 越南教育体制, 河内, 胡志明市, 越南劳动力市场

研究背景: 在过去的数十年中，劳动力从农业部门的大量流出给越南带来
了生产力的巨大增长。但是，这样的发展模式将在未来几年难
以为继，要想保持经济的增长，越南需要寻找新的发展路径。
目前已有很多关于越南不同时期教育回报率的研究，但目前尚
未有研究专门关注河内市(首都)、胡志明市(最大经济体)的教育
回报状况。这两个城市的教育与发展水平同越南其他地区之间
有巨大差异。
研究目标: 本文试图调查研究越南两个最大且最具影响力城市的教育回报
率情况。两个城市间的对比及它们与越南全国水平的对比，在
以往研究中已有过论述。
研究方法: 本文将会利用扩展的明 Mincerian 模型展开研究。本文的研究
数据来自越南国家统计局的居民生活水平调查报告。通过使用
不同的工具变量，本文用 Heckman 两阶段模型去解决选择性偏
差。
研究结论: 河内市与胡志明市相比于越南其他城市，显然有更高质量的劳
动力供给。与以往基于全越南的研究结论不同，在这两个城市，
通识教育并没有提高劳动者收入，工人需要至少获得一个职业
学位以显著提高他们的工资。
关于这两个城市之间的差异，统计结果显示，胡志明市总体上
提供更高的工资（4.45%）。此外，首都为额外的一年教育支付
更多（1.95%），而年轻的城市对经验更有价值，尽管不是很大
差距（0.66%）。Heckman 的修正模型确实表明，Mincer OLS 高
估了教育的回报。然而，偏差的大小并不大（通常少于 10%），
这与之前的一些研究结果是一致的。
描述性和统计结果提供了政策影响;例如，提高职业教育的推
广，以及对资本的需求，以建立一个更加开放、灵活、不那么
依赖的雇佣制度，特别是在公共部门。
本文的发现一方面为政策制定提供了启示，另一方面表明研究
越南特定经济区域的教育回报率的必要性。

中图分类号：F276.6
 ABSTRACT

Keywords Return to schooling, Vietnamese education system, Hanoi, Ho Chi

Minh City, Vietnamese labor market
Background In the last decades, Vietnam has enjoyed significant growth in
productivity thanks to the massive structural change of labor moving
out of the agricultural sector. This will no longer be the case in the
next years; and the country will need to look into other solutions,
most prominently the education system, to keep the economic
growth going. Return to schooling has been studied in Vietnam for
different periods, but none so far has had the focus on Hanoi (the
capital) and Ho Chi Minh City (biggest economy) – the two cities
with very much different education and development level compared
to the rest of the country.
Purpose The thesis attempts to investigate the return to schooling of the two
biggest and most influential cities in Vietnam; comparing and
contrasting between these two cities, as well as the country-level
results presented in previous studies.
Method In this thesis, the author will make use of an extended version of the
classic Mincerian human capital equation. Data from the latest
Vietnam Household Living Standard Survey (2016) from the
General Statistics Office of Vietnam is employed. Heckman two-
stage method is also conducted to address selection bias issue, using
different appropriate instrumental variables.
Conclusion It is apparent that Hanoi and Ho Chi Minh has a very different
overall labor market compared to the country’s average, both in
structure and labor quality. There is no premium wage for people
with general education, meaning that workers need to earn at least
a vocational degree to increase their wages significantly compared
to people with no education. The results are in contrast to previous
studies’ findings using national-level data, where all educational
level pay incentives.
Regarding the differences between the two cities, statistical results
reveal that in general Ho Chi Minh City offers higher salaries
(4.45%). Furthermore, the capital pays more for an additional year
of education (1.95%) while the younger city is more rewarding
toward experiences, though not by a huge gap (0.66%). Heckman
correction model does show that the Mincer OLS overestimates the
return to education. The size of the bias, however, is not large
(mostly less than 10%), which is consistent with results from some
previous studies.
Descriptive and statistical findings offer policy implications; such
as improving the promotion of vocational education, along with the
demand for the capital to develop a more open, flexible, less –
depending - on - credential hiring system, especially in public
sector. This study also raises the need to study return to education
in more specific economic regions in Vietnam.
Chinese Library Classification Code: F276.6

2
 ACKNOWLEDGEMENTS

I would like to thank my advisor – Professor Li Huailu for her constant and valuable
guidance, as well as commitment in the completion of this thesis.

I would also want to thank Mr. Pham Duc Duong at the General Statistics Office of
Vietnam, who has been supporting me throughout the process of my writing.

Lastly, sincere thanks to all my families and friends for giving me unconditional
encouragement and love. This thesis would not be finished without your kind
involvement.

Ha Truong
Fudan School of Economics, 2018

3
 INTRODUCTION

1. Introduction
According to a report by McKinsey consulting firm (Sustaining Vietnam’s growth: The

productivity challenge, 2014), in the next years to come, Vietnam will need to improve

its overall labor productivity growth by more than 50 percent if the economy is to meet

the government’s target of 7 to 8 percent in annual growth by 2020. Without this push,

the country’s growth is likely to decline to between 4.5 and 5 percent. For the past few

decades, the Southeast Asian country has experienced and enjoyed significant benefits

from structural transformation (especially the labor force’s transition out of agriculture).

However soon this instrument of economic growth will no longer be as rewarding as it

once was, and the country must find ways to step up its productivity performance

without relying heavily on structural changes.

Like in many developing countries, Vietnam has looked at its education system as the

top prioritized measure to answer the question for skilled workers. In particular, return

to schooling, one of the central policy parameter (Becker, 1962), has been studied and

estimated in a number of researches, especially after the market reform in the 1990s.

The purpose of this thesis is also to identify the wage earnings attributable to education

in Vietnam. However, the uniqueness of this study comes from the specific focus on the

two most significant economic hubs in Vietnam: Hanoi and Ho Chi Minh City; and the

usage of latest dataset from the 2016 Vietnam Household Living Standard Surveys.

Previous studies on return to education in Vietnam have all focused on the aggregate

level of the country as a whole. While this is often the case for studies of this kind, a

line of work from Backman & Bjerke (2009) in Sweden, Hering & Poncet (2010) in
4
China, Fally, Paillacar & Terra (2010) in Brazil and Cieslik & Rokicki (2016) in Poland

have chosen to take the direction of finding individuals’ incomes’ heterogeneity as

related to regional disparities. These studies are based on the theory of New Economic

Geography, which suggests regional competitiveness are closely related to economic

geography and regional economy. Their empirical findings all confirmed the

differences in return to education existed between regions within the same country.

What cause these differences? According to Roback (1982, 1988), regional wage

differences can be explained largely by local attributes; in particular quality of life, or

amenities (such as infrastructures, pollution level, climate and other natural

endowments). Furthermore, Hanushek & Kim (1995) assert that “quality” of the labor

force is the ingredient that can explain differential growth within regions. Conducted at

country level, the study used cognitive skills in mathematics and science (through

comparative tests for children) as a measure for labor quality to conclude the

importance of investments in education and training. Finally, with the rise of creativity

as a fundamental economic force, vibrant, better communities built with the focus on

“people climate” have also become locations that can attract companies and higher

quality labor, especially the “creative class” – a new socioeconomic class identified by

American economist Richard Florida which can spur regional economic growth

through innovation.

Like siblings in all of nature, Hanoi and Ho Chi Minh city have and will always be

measured against each other in every single aspect; and often are being compared to the

similar rivalry of Beijing and Shanghai, Madrid and Barcelona, or Sydney and

Melbourne (Nielson 2009). Undoubtedly both are the ultimate driving forces behind

the dramatic growth that transform the country and a special kinds of urban in Vietnam.
5
Both have the highest concentration of colleges and universities – in number and quality

(Nguyen & Nguyen, 2008), and are the leading destinations of migrant workers. Thanks

for their significant roles in the development of Vietnam, the two cities are oftentimes

studied separately or in comparison with the rest of the country. Yet, as the nation was

still divided into Northern (with Hanoi as the capital) and Southern (with Ho Chi Minh

city, or Saigon at the time, as the capital) Vietnam until 40 years ago; each city does

possess her own characteristics, culture and business environment that can potentially

influence the labor market.

As a result, estimating and understanding the differences in return to schooling of Hanoi

and Ho Chi Minh city is, first of all, to provide an insight into the situations of these

two prominent cities, which might be different from the overall picture of the country.

Moreover, like many developing country, there is a great deal of provincial variation –

even ones within the same region in Vietnam. Hence, it is highly important to study and

set up perceptive strategy at province level (Dapice, 2003). The second purpose of this

thesis, therefore, is to hopefully offer some implications in customizing a better-fit

framework of education, training and labor for each city.

1.1 Hanoi & Ho Chi Minh City: The tale of two cities

Vietnam, for the last century, has been a tale of two cities. Hanoi is the capital, a long-

time political, economic and cultural center; whereas Ho Chi Minh is the largest city

by population and the leading economic, social and technological hub of the country.

In term of economic significance, the two cities are by far the biggest contributors to

the state budget revenue, with Ho Chi Minh city (HCMC) reached $15.32 billion in

2017, topping and outperforming all other cities and provinces in the country (General
6
Satistics Office, 2017). As a developing economy with rapid urbanization rate, Vietnam

also suffers from a widening gap between urban and rural poverty. Big hubs like Hanoi

and HCMC that can offer better wages, education and overall lifestyle attract a large

number of migrant moving away from their rural hometowns to the cities. According

to the 2015 National Internal Migration Survey, of those aged 15-59 living in Hanoi

and HCMC, migrants account for 16.3% and 20.7% (higher than nation’s average)

respectively. Most migrants are employed is places of destination, while the majority

of the rest migrate for education purposes. Migrants also tend to get married later and

have higher education (either professional/technical qualifications or college/university

level) compared to non-migrants. All of these factors have an considerable impact on

education, health care services and employment in the migrant destinations, especially

regions with a high level of young migrants such as the Red River Delta and Hanoi.

After the fall of Saigon in 1975, Vietnam was formally reunified. Peace was restored,

and the country could finally move forward and develop as a whole. Nevertheless,

differences in natural endowments and history have helped establish certain

distinguishing characteristics (sometime borderline stereotypes) of Northerners and

Southerners, with Hanoi and HCMC as the main representatives. Hanoi - the capital

with a long history, rich culture, distinct extreme cold and hot seasons - is somewhat

more traditional, cultured, formal and introverted. Young HCMC, on the other hand, is

an international port with the tendency of openness and more favorable climate

endowments; hence much more dynamic, vibrant and casual. These differences in

history and culture have shaped a relatively contrasting mindset of Vietnam’s two

largest cities when it comes to business. Home to the ruling elite since ages, Hanoi is

the city with hierarchical society where strong political connections tend to go a long
7
way (Cooper, 2016). Due to the development path and position as the capital, Hanoi

also has more state-own enterprises running, which might reinforce this tendency. On

the contrary, HCMC has always been a welcoming destination for traders and

entrepreneurs with great business mentality. It is therefore generally more “fluid,

multicultural and outward-looking” compared to the capital (Cooper, 2016). For the

past decade, Hanoi has been and is still trying to refresh and rebrand its image as a more

open, nurturing place for talents. Still, the differences are there.

In addition, influenced greatly by Confucianism due to the long history of Chinese

domination, Hanoi also tends to be more concerned about status and qualifications. The

capital was the home to the very first university in the country and until today is still

leading in many indicators in education. Despite a smaller population (7.3 million vs.

8.3 million, 2016), Hanoi has a higher number of school (1,546 vs. 950) and students

(1,240,038 vs. 1,183,792) in general education (primary, lower secondary, upper

secondary); as well as number of teachers (23,948 vs. 17,189) and students (610,872

vs. 509,532) in universities/colleges compared to HCMC (General Statistics Office,

2016). The labor force also tells a similar story: Hanoi has a higher percentage (42.7%

vs. 34.8%) of trained employed workers and the highest percent of migrants with

professional or technical qualifications in the country – 46.7% (General Statistics

Office & UNFPA, 2015). Noted that even though the numbers are lower for HCMC,

they still generally outperform the rest of the country. An online article (Quek, 2017)

about the start-up scene in Hanoi and HCMC practically summed up the essence of

these subcultures that influence the way business work in the two cities when stating

that Hanoi has more in number of qualified engineers who are also generally more

talented; whereas HCMC creates much better business-minded entrepreneurs.

8
All in all, considering these facts in mind, one might make the predictions that Hanoi

is more rewarding to people who spent longer time in school. Whereas a more flexible,

fast-moving boomtown like HCMC might be more responsive to factors other than

qualification, such as experience. This thesis, therefore, will take the route of exploring

the return to schooling in these two most prominent cities of Vietnam; compare and

contrast between the two, as well as the results of the whole country from previous

studies.

1.2 Scope and structure

This paper attempts to utilize the data available in Vietnam Household Living Standard

Survey conducted by the General Statistics Office of Vietnam to estimate the return to

schooling in Hanoi and Ho Chi Minh city. Other unobservable factors, such as the role

of labor union, efficiency wage, school quality et cetera cannot be discussed and

incorporated within the scope of this thesis.

The paper is structured as follows: Sector 2 reviews literatures on human capital

theories, as well as the returns to schooling in both international scale and the country

of Vietnam. Section 3 discusses in detail the data and methodology of the study,

followed by descriptive and econometrical findings in section 4. Last section will

content concluding remarks.

9
 LITERATURE REVIEW

When it comes to return to schooling, there have been an overwhelming number of

studies, conducted in various geographic locations and different time periods,

supporting the positive correlation between education and the level of success of

individuals in the labor market. Better-educated people are more likely to earn better

job, enjoy higher wages and less likely to suffer from unemployment (Card, 1999).

Compiling studies such as the like of Psacharopoulos (1985, 1994, 2002) all show a

relatively consistent pattern across the globe throughout the years. More recently,

Montenegro & Patrinos (2014) have carried an ambitious paper which assembles

empirical findings from 139 countries and 819 household surveys – all with comparable

estimates, database using the same specification, estimation procedure et cetera. The

results once again reinforced findings from previous studies: Private returns to

schooling are generally positive but declines as the investment for education are

expanded, women receive more profits in obtaining higher education than men; and

estimated return to education seems to be higher in low or middle income countries.

Though in another study specifically targets developing countries, Peet, Fink & Fawzi

(2015) concluded that they found no evidence suggesting returns to schooling in these

countries are systematically higher than those in developed economies. Recent studies

have also revealed a significant reversal result from previous findings; and that globally,

private returns to tertiary education is now the highest, followed by primary and

secondary schooling (Patrinos, 2016). This is mostly due to the fact that primary schools

have become compulsory and universal in a big portion of the world.

10
One thing for sure, returns to schooling is not a single parameter; it is a variable

influenced greatly by different characteristics and background of an individual (Card,

1999). More will be discussed in the following section.

2.1 Mincer equation

How to accurately measure the causal effect of education on earnings, on the other hand,

is a topic of great debate and is still being developed, refined until today. Nevertheless,

a large number of modern studies in earnings attributable to education have adopted the

classic Mincer’s earning function (1974), which models the natural logarithm of an

individual’s earnings into a function expressed by her number of completed school year

and experience as follows:

Ln wi = α + β*Si + β1Xi + β2Xi2 + ε

Where
w: individual’s earnings
S: individual’s years of schooling
X: individual’s years of experience

And ε is the disturbance term presenting other unobserved factors that

determine wages but cannot be measured.

The use of log transformation serves several purposes. For one, this specification allows

interpreting in percentage term. Moreover, the distribution of log earnings is very close

to normal distribution; and it helps exclude negative earnings values – which happens

when certain individuals only receive “income” in form of government subsidy and do

not have to pay tax (Friedman, 1962). This, however, is not really a concern in this

paper. Various studies use different earnings measure – annual, weekly and most

popularly, hourly.
11
Mincer equation depicts a linear relationship between education and earnings, which

assumes that each additional year spent in school will increase wages in a linear trend

– constantly and proportionally the same throughout the year. This is not always the

case. Additionally, this use of measuring education soon showed some flaws when

adopted to countries where students can have very different school tracks (and thus

cannot be slumped into one same years-of-schooling unit) depending on their future

plans (either go to college, vocational school or start work right away after high school)

(Card, 1999). As a result, new studies soon preferred the use of dummies variables for

education level that one completes as an alternative measure of education. This intuition

was developed and dubbed as “Sheepskin effect” by Hungerford & Solon (1987) and

Belman & Heywood (1991), which hypothesizes that the obtainment of an actual degree

is more rewarding than the same amount of year in school without a certificate. Thus,

more surveys also ask for the degrees that people have acquired, besides how many

years of school they had completed. Nevertheless, Harmon et al. (2000) demonstrated

that the assumption of a linear return to schooling, despite being a strong assumption,

is “remarkably hard to reject”. For this thesis, the author will use both methods of

measuring education for appropriate regressions.

Lastly, a quadratic form of experience is used in the equation to express the concave

relationship between age and earnings. In other words, the benefits of education can

behave differently for different age group: Typically, it grows relatively faster at the

beginning of one’s career and slows down, even declines when she approaches

retirement (Murphy & Welch, 1990). Since it is hard to obtain accurate information

regarding a person’s relevant experience; most studies construct a “potential experience”

variable (proposed by Mincer), which typically equals to an individual’s age minus

12
number of completed schooling year and 6: X = A – S – 6 (assuming that she starts

school at the age of 6 and begins to work immediately after graduation).

Throughout the years, the Mincer equation has received criticism for its many

shortcomings, most prominently are its potentials for biases. Regardless, the human

capital earnings function has laid foundation and become a solid framework for a

majority of studies in return to education – thanks to its ease of use, simplicity yet

powerful intuition. In fact, several papers (Card, 2001; Montenegro & Patrinos, 2015)

have summarized results from different studies and noted that basic Mincerian model

is more stable than may have been expected. In the next sections, we will discuss about

some of its most debated issues, and later, how we can overcome these limitations

within the constraint of this thesis.

2.2 Extended Mincer equation

2.2.1 Ability bias

One of the most and earliest discussed drawbacks of Mincer equation is the extent to

which schooling is endogenous due to people’s unobserved abilities (Griliches, 1977).

Economists realized that ability and schooling were likely to be correlated (most likely

positively); and if ability affect earnings, the estimated return to education will

potentially upward biased (Lang, 1993). Put it differently, higher earnings might

mistakenly credited to higher level of education, while in fact it is owing to higher

abilities. Individuals with superior abilities are more likely to do well in job interviews,

and generally perform better in their jobs (hence, higher earning outcomes) – not

necessarily because they have higher level of education.

There have been several approaches to tackle this issue. One way is to include a proxy
13
variable that can represent inherent ability. IQ test result (Altonji, Dunn, 1996; Belzil

& Hansen, 2002; Gensowski, 2018), for example, has been said to be a strong predictor

of education (Card, 2001). Though indicators of ability are in general not perfect, if

even present at all in regular dataset. Another approach is to use data of twins and

siblings (Ashenfelter & Zimmerman, 1997; Ashenfelter & Rouse, 1998; Bronars &

Oettinger, 2006; Arias, Hallock & Sosa-Escudero, 2001) who share the same family

conditions and sometimes abilities to eliminate omitted ability bias. After the bias is

removed, the differences in earnings between the siblings and twin will then be given

credited to education. Some other studies use panel data or repeated observations over

time (Hausman & Taylor, 1981; Chatelain & Ralf, 2010) to cross out the unobserved

abilities (since the effects to returns to education of these latent variables are constant

over time).

A final and quite popular method to deal with ability bias is the use of instrumental

variables (IV) for schooling, which has raised a considerable body of work. Follow the

requirement of instrumental variables, a valid instrument should be correlated with the

exogenous variable – education; but not ability, and thus not earnings. Various studies

have adopted different instrument variables in both supply (or institutional features)

and demand side. Proximity to the nearest high school/university (Kane & Rouse, 1993;

Card, 1993; Conneely & Uusitalo, 1997); minimum school leaving age (Dickson &

Smith, 2011; Harmon & Walker, 1995) and school quality (Hogan & Rigobon, 2010)

are examples of institutional features. Whereas on the demand side, studies have used

quarter of birth (Angrist & Krueger, 1991; Staiger & Stock, 1997); family background

such as parents/fathers’ education (Ichino & Winter-Ebmer, 1998; Meghir & Palme,

1999) and incomes as instrumental variables. The use of IV method has its limitation,
14
especially the issue of weak instruments – which leads to asymptotically biased towards

the corresponding OLS estimates (Card, 1999). Spence (1973) also pointed out that

individuals’ latent abilities might already be reflected through the acquirement of

educational degree by itself, as people with higher abilities know this is an effective

way to communicate, or send signal to their potential employers. In other words,

education can be interpreted as the mean to enhance productivity, or indication of

existing inherent productivity. Spence also argues that firms tend to believe that

credential is positively correlated with better quality candidates, and relies on it to make

recruitment decisions.

Nonetheless, Becker (1964), Griliches (1977), Becker & Tomes (1986), echoed each

other and came up with a similar conclusion that ability bias is overestimated in their

comprehensive review of literatures; and such biases were small with ability could only

explained 1% of returns to education (Griliches, 1977). According to Griliches, when

schooling is treated symmetrically with ability measures, it also allows itself to the

subject of measurement error and be correlated with the error term in the earning

function. Biases caused by unobserved ability variables and measurement error will

then tend to cancel each other out. Borrowed the words from Heckman et al (2005), the

study of endogeneity of returns to schooling “is far from settled”.

2.2.2 Selection bias

The problem of selection bias of the Mincer model was first raised and discussed by

James Heckman (1979). The argument here is quite simple: With this earning equation,

we can only capture the representative information of the wage-earner population (a

subset of our actual population in study), foregoing the people who are unpaid or
15
unemployed. This causes non-randomly selected sample and consequently, downward

bias in the OLS estimates using the Mincer equation. In this thesis, the author will

employ the 2 stage method named Heckman correction to address the issue of selection

bias. Statistical software packages like Stata offer very simple mean to execute this two-

step estimator; follow is a brief of the intuition behind.

First, we have the basic wage equation:

Wi = Xiβ1 + u1i (1a)

Where W is an individual’s earnings, Xi is a vector of her schooling, experience,

gender … relating to her productivity. We can only observe W when the person enters

the waged labor market. For others who do not, it is likely because their offered wage

is below their reservation wage.

We then have a second equation relating to employment, that is the lowest rate at which

an individual would be willing to accept a particular job offer:

E*i = Ziβ2 + u2i (1b)

Where Zi is a vector of explanatory variables (schooling, experience, gender, parents’

income and/or education… which can explain her constraint). E*i = Wi – E'i is the

difference between the wage and the reservation wage E'i. Equation (1b) can be a probit-

type of selection model, describing whether a person will decide to work. If the wage

is below the reservation wage, they choose not to work. We observe only an indicator

variable for employment defined as E =1 if E*i >0 and E=0 otherwise.

Further, we also assume that the error terms u1i and u2i are normally distributed with

mean 0 and positively correlated. (u1i, u2i) are independent of both sets of explanatory

variables X and Y:

u1i ~ NID (0, δ2) and u2i ~ N (0,1)

16
 cov(u1i, u2i) = ρ12
ρ12 is the correlation coefficient.

Heckman’s 2 step model is to first estimate a so-called inverse Mills ratio (or

Heckman’s lambda) by a way of a Probit model:

λi = ϕ(Ziβ2) / Φ(Ziβ2)

Where ϕ denotes the standard normal density distribution and Φ is a normal cumulative

distribution function. In the second step, λi will then be used in the original wage

equation (1a):

Wi = Xiβ1 + δ12λi + ηi

As long as u2i is normally distributed and ηi is uncorrelated (or independent of) λi, the

Heckman model is consistent (Puhani, 2000).

One of most challenging task in employing Heckman 2 step model is to find a good

identification strategy. All explanatory variables used to predict the likelihood of

employment are also based on the human capital theory. Weakly identified variables

can worsen or show mislead results in the attempt to correct the sample selection bias,

producing large size distortions.

Finally, there are other factors besides latent abilities variables that might affect a

person’s earnings as well. Gender, marital status, job industry, job sector (Mann &

Kapoor, 1988; Lall & Sakellariou, 2010) et cetera are variables that have been suggested.

The original Mincer equation, therefore, will be extended for this thesis as follow:

Ln wi = α + β1Zi + β2Xi + β3Xi2 + ε

In which:
Ln w = individual’s log of monthly/hourly wage
Z = Education level dummies
X = Age/Potential experience
E = Other controlling variables

The regression will also be run with the Heckman correction model, using some
17
appropriate instrumental variables to exam for selection bias (More details in the

Methodology section).

2.3 Studies on Return to education in Vietnam

The study of return to schooling has been a topic of interest for researchers in Vietnam,

with most results show a positive relationship between education and earnings with

different magnitudes.

A few studies chose to focus on the periods after the Doi moi (English: Renovation)

economic reforms initiated in 1986, with the goal of abolishing the command economy

and creating a “socialist-oriented market economy”. In particular, Moock, Patrinos &

Venkataraman (2003), using earnings data from 1992-93, estimated that the rates of

return was relatively low compared to other developing countries during that time.

Return to primary and university averaged 13 and 11% respectively, but only 4 to 5%

at secondary and vocational levels. Gallup (2002) reported a similar results, stating that

the payoff for one additional year in school, though doubled from 1993 – 1998, was

still low at 5%.

Liu (2006), however, observed a considerable shift in demand toward more educated

workers between 1992 – 1998, with earnings for people with tertiary education had

significantly increased in comparison with workers with below-primary education. And

in a more recent paper with updated data from 1998-2014, Doan, Tran & Le (2016)

again estimated a higher rise during this same period compared to the results of Glewwe

& Patrinos (1998), Gallup (2002) and Mock, Patrinos & Venkataraman (2003). The

authors also concluded that returns continued to increase up until 2008 when it appeared

to fall between 2008 and 2014 – which they credited the rapid expansion of education
18
and the slow-down of economic growth and transformation as possible explanations.

Furthermore, one of the most common findings in these papers is the higher returns to

education for female compared to male workers in the labor market (which is similar

with the results found across the globe). Studies about wage inequality such as Gallup

(2002), Liu (2004), Pham & Reilly (2007) all found a sharp reduction gender pay gap

disparities for the wage employed population associated with the Doi moi reform. Using

the decomposition method, however, the authors found that the convergence of the

Vietnamese gender gap in the 1990s hided an adverse change in discrimination. Similar

to China, women still faced a substantial wage discrimination due to employer’s

preferences and marker competitiveness of female employees (Liu, 2004).

As a transitional economy, Vietnam has experienced a somewhat distorted labor market.

Coxhead & Phan (2013), using data from 1993 to 2008, explored a more interesting

dimension of inequality: households with and without access to jobs in the state sector.

Economic policies biased toward state-owned enterprises (SOE) have yielded certain

privileges for public servants. State sector jobs provide stable income, promotion and

overall longevity; they are also generally less demanding in time and productivity. This

leads to fierce competition to land a bureaucrat position, despite relatively low salaries.

And since the number of available positions are limited, organizations will tend to favor

people with connections based on family ties. According to Coxhead & Phan, as a result,

household with these types of “relationships” (or, quan he) tend to invest more in

education for their children. Whereas families who see little incentives from studying

or getting credentials, encourage their offspring to opt out of school early and

sometimes use that investment money to bribe for jobs instead (Doan, Tran, Le, 2016).

Such phenomenon has caused severe corruption in the Vietnamese labor market, even
19
popularized the term “Con ong chau cha” (Similar to the idiom “Born with a silver

spoon in mouth”) which refers to people who are hired and promoted due to their

parents/grandparents’ connections. Since the beginning of the economic reform

however, Vietnam has made efforts to privatize SOEs, promote private firms in specific

industries and attract foreign-invested firms. The situation, therefore, might have more

or less changed since then.

Lastly, some other studies took on different routes such as ethnic inequality (Pham &

Reilly, 2009) or returns for private vs. public schooling system (Glewwe & Patrinos,

1999). So far no studies have put regional differences as the focal point of their paper;

or done analysis for the two biggest cities separately. Though Gallup (2002) did

estimate in his paper that workers in Hanoi and Ho Chi Minh city receive a substantial

(50% higher) wage premium over the rest of the country in the period between 1993

and 1998.

Studies of returns to education in Vietnam often faced the difficulties of data

unavailability and insufficiency. Most research in this topic also used the data from

Vietnam Household Living Standard Survey conducted by the Vietnam General

Statistics Office, which until 2002 was still criticized for employing a relatively small

sample sizes. Moock et al., (2003) recognized a limitation in their analysis; as, due to

the survey, could only include wage-earners when during the time of the study 80% of

the Vietnamese labor force is self-employed. Moreover, most studies, except for more

recent one like Doan (2011) and Doan et al., (2016), did not consider proxies for

individual ability and motivation from family resources; even though investments for

education in Vietnam face high liquidity constraint (Glewwe & Jacoby, 2004; Glewwe

& Patrinos, 1999). Some acknowledge but did not include selection bias in their
20
analysis (Moock et al., 1999; Pham & Reilly, 2007).

This thesis, despite overcoming several limitations of previous studies, also has to deal

with similar shortcomings related to data. In the next section, we will talk in details

about the chosen dataset, as well as the methodology that the author will implement.

21
 DATA & METHODOLOGY

3.1 Data
The Vietnam Household Living Standards Survey (before that, Vietnam Living

Standard Surveys - VLSS), or VHLSS, was one of the two most important household

survey developed in Vietnam during the 1990s. The first two VLSS were implemented

in 1992-93 and 1997-98 with the technical assistance from the World Bank and funded

by the UNDP and Swedish International Development Cooperation Agency (Sida).

Despite being high quality and contains a wide variety of topic, the survey was

criticized for its relatively small sample and implementation frequency of every 5 years

– too low to satisfy the need for up-to-date information for social monitoring. As a result,

the Vietnam General Statistics Office had developed a strategy for the second wave of

this survey in the next 10 years (2000 – 2010); combining it with the other big but

similar survey – Multi-purpose Household Survey - and turning it into an once-every-

two-years and rotating module household survey. The new VHLSS, though more

simplified, still includes intensive sections regarding demographic characteristics,

education, employment, income, expenditures and so on of the Vietnamese population.

Though there was no re-interview of households from the original VLSS panel.

This thesis will be utilizing the results from the latest VHLSS in 2016 (which to the

best of my knowledge, the first one to do so), containing information of 2,482

individuals from Hanoi & 2,262 from HCMC (over 15 years old and currently are in

the labor force). The data provides quite sufficient variables for all analysis purposes

(More on that later).

22
3.2 Methodology
As stated earlier, the thesis will make use of an extended version of the Mincerian

equation to estimate the causal effects of education and labor market outcomes in the

two cities Hanoi and HCMC:

Ln wi = α + β1Zi + β2Xi + β3Xi2 + ε

Where ln w is the log of the individual’s earnings, Z is a dummy for educational

attainment level, X is his/her age and E is the vector of other controlling variables.

3.2.1 Dependent & Independent variables

Income (or Wage)

In the traditional Mincerian framework (1974), the dependent variable can be either

income or wage; with both are equally valid alternatives. In this thesis, the person’s

earning is calculated as the average monthly wage, plus total bonus in that job divided

by 12 months (if any). Also, monthly wage seems to be a more appropriate choice than

hourly, as all salaries in Vietnam are negotiated and calculated on a monthly basis. All

monetary amounts are expressed in Vietnam Dong (hereafter VND).

Age vs. Potential experience

Most studies using the Mincerian equation construct a “potential experience” variable

by subtracting the average years to obtain the educational degree and 6 (the age when

a person typically starts getting formal education) from the individual’s age. This is to

assume that the person stays in school completely until she finishes her degree, which

is actually not the popular case for postgraduate-holders in Vietnam. Here, people tend

to keep working while getting their master and PhD (taking night, after-work and

23
weekend classes). Sometime it is the companies that send their staffs to get the degree

and allow flexible schedule for them to study while remaining in the job. The only case

when the person most likely spends her entire time at graduate school is when she goes

abroad, which there is no available information and this subpopulation only accounts

for a small portion of the labor force. By using a potential experience variable, we might

potentially dismiss 2 years of experience for master holders, and another 3-4 years for

people who have a PhD. Also in Vietnamese context, aside from work experience,

“experience of life”, relationship, connections acquired through longer years of being

in the labor market are almost as important. So for these reasons, in this thesis, the

author will use Age as the independent variable. According to Harmon et al. (2000) who

compiled the OLS estimates of different European countries, using “age” tends to

produce the lowest returns.

We will also use an age squared variable to capture the concave relationship between

age and earnings, as explained in the previous section.

Educational level

Vietnam’s education is arranged on national level by the Ministry of Education and

Training, following the typical education system ranging from primary school to

doctoral degree (for more details, see Appendix A). Only the five years of primary

education are mandatory, with literacy rates very high at 97.3% (World Bank).

The country also has a vocational system, which consists of Elementary vocational

school, Middle-level vocational school and Professional school. This vocational track

can start either after Higher secondary, Lower secondary or even Primary education;

and last for 6 months up to 2, even 3 years - depending on the courses and programs.

24
The VHLSS records in detail an individual’s education level from “No qualification”

to “Doctoral degree”, as well as the grade they are currently in (if any). It also provides

separate information for person who earns both secondary, postsecondary education and

vocational education. The author will make use of both highest educational level and

estimated year of schooling variables in the regressions for this thesis.

An estimated year of schooling makes it more convenient and effective for the

comparison of average rate of return to an additional school year between the two cities,

as well as for other data descripting purposes. If a person earns both higher education

and a vocational degree, her years of schooling should be the sum of both degrees.

Highest educational level obtained is used as well. With people who have both her

regular and vocational degree listed, there is no way to identify when and in what order

the person gets her credentials. But for the sake of analysis, the author decides that if a

person obtains a secondary education (or lower) and a vocational degree, then the

vocational credentials will be counted. Otherwise, if she earns both a vocational degree

and a tertiary education (college, university and up), the tertiary education will be listed

as her educational level. This is based on the assumption on which credentials might be

more relevant when people apply for jobs. Say for example, if a person has a high school

degree and a vocational certificate; while she would list both in her resume, the

vocational education is likely to be more related to the job and hence has more value to

the employer. Conversely, while a vocational degree might be useful in a person’s

profile; if the person takes extra time to earn a higher education, it is most likely that

she wants to land a better job which she cannot get with only vocational credentials.

25
3.2.2 Controlling variables

Gender, Ethnicity, Migrant (whether the person was born from another province or not),

Area (urban/rural) are factors that might have an impact in an individual’s earnings and

will be included as controlling variables.

As for the employer side, Economic sector, Firm type and Job level (occupation

category) are constructed as employment variables. The VHLSS categorizes

participants into 99 detail small economic sectors, which will be grouped into larger

sectors for analysis conveniences. Firm types include dummies for: Public, Private,

Individual household, FDI.

Occupation category is divided into 8 small categories from low-skilled workers to

leaders/managers. Different job level might have different attitudes toward credentials.

For example in order to become a director, one is usually faced with many requirements

in experience and skill sets, education is not entirely relevant if the person has a good

working record. Whereas at lower / entry-level positions, companies tend to look at the

person’s educational level to make assessment. The author decided to exclude Skilled

laborers in agriculture, forestry and fisheries (due to very low number of observation in

Hanoi & HCMC) & Members of the Army (due to sensitive issues) in the Occupation

category1.

Both Firm type and Job level help us include the informal job sector2 to the dataset and

1
World Bank Survey Manual describing how to sort observations into different Occupation category:

http://microdata.worldbank.org/index.php/catalog/2729/download/39053. This is linked instead because

the General Statistics Office does not provide an online instructions of the survey.
2
ILO Guidelines Guidelines concerning a statistical definition of informal employment:
http://ilo.org/wcmsp5/groups/public/---dgreports/--

stat/documents/normativeinstrument/wcms_087622.pdf However do note that Vietnam, to some extent,

26
to some extent, overcome the limitation in some previous studies as they were forced

to exclude this sub-population due to the nature of the survey. Indeed, the informal

economy is an important sector for a developing country like Vietnam (Moock et al.,

2003; Kazuhiro, 2007) and studies have shown that there were evidence of education

premium for high skilled workers in the informal sector (Arbex et al, 2010). With the

new VHLSS survey, informal sector is now covered (or partly covered) in forms of

Individual household business (in Firm type category) and Low-skilled workers /

Manual labor (Job level category).

For a detail list of variables and their category values, please check Appendix B.

3.2.3 Bias issues

In a paper exploring different sets of exogenous instruments on both supply and demand

side to explain the variation in schooling in Vietnam, Arcand et al., concluded that only

parental education (along with HT matrix of instruments) satisfy the necessary

requirements. Unfortunately, the way that the VHLSS is designed does not allow us to

capture such information. In particular, the survey is conducted at household level,

interviewing household head. Therefore, with people who do not live with their parents,

there is no way to obtain their education information. Furthermore, some people live in

the same household are son/daughter in law or cousins, whose parents that influence

their education choices live in a different resident. Constructing parental education

variable, therefore, will diminish the sample size dramatically. Likewise, attempts to

extract similar family background information like parents’ income, job sector

has its own definition of this sector.

27
(private/public) et cetera also face the same problem. Moreover, fixed effects using

panel data is also not practical as the VLHSS survey is not intended to survey the same

sample over time. The returns to schooling in this thesis, therefore, are assumed to be

unaffected by unobservable individual abilities. This is one of the limitations in using

the VHLSS data.

As for explanatory variables that are used for the Selection equation in Heckman

correction model, the author will use Household size, dummy for being the Household

head or not, Marital status and Household non-wage income as instrumental variables.

The rationale behind is that while the number of people in the household, whether a

person is married (especially female) and a household head or not, do have a certain

influence on a person’s decision to enter the workforce in Vietnam; employer are

unlikely to take these facts into consideration when hiring someone. A variable for

number of children might potentially be a more effective variable than marital status;

however this information cannot be efficiently obtained due to the survey structure as

explained earlier. Additionally, Doan et al., (2016) also used household non-wage

income as instrumental variable, reasoning that families with this type of additional

income (either from family business, agriculture activities, loan interests…) will have

higher reservation wage, and are less likely to send their members outside to work.

The International Labour Organization (ILO) has a detail definition of different

employment status1, including unemployment. In this thesis, the author adopts a relaxed

1
Resolution concerning statistics of the economically active population, employment, unemployment

and underemployment, adopted by the Thirteenth International Conference of Labour Statisticians:

http://www.ilo.org/wcmsp5/groups/public/---dgreports/---

stat/documents/normativeinstrument/wcms_087481.pdf
28
understanding of unemployed people, being who answered (1) “Cannot find job” and

(2) “No need to work”, (3) “Family reason” & (4) “Health reason” to the question:

“Reason for not working in the past 30 days” in the survey.

Basic Mincer equation will be run first, then with personal characteristics and other

controlling variables. Finally, regression with Heckman correction model will be

reported alongside with the Mincer OLS results. All models are tested with estat vif

afterward to check for variance inflation factors (VIFs) of the independent variables

and passed. Survey (sampling/probability) weight variable is included in all regressions,

indicating how the observations represent the number of people in Hanoi and HCMC’s

population (selection probabilities).

Next section will present in-depth the descriptive and regression results.

29
 EMPIRICAL FINDINGS

4.1 Descriptive findings

Table 1 provides an overview descriptive comparison of Hanoi and HCMC in terms of

demographic and employment characteristics.
The two cities have almost the same average age (37) and the share of women in the
labor force (45.37% vs. 46.7%) – which is slightly lower than the country’s number
(labor force at 15 years of age and above). Workers in Hanoi, however, are much less
likely to live in the urban area (51.5% vs. 83.1%), even though this number is still
significantly higher than the national average – 32.1% (General Statistics Office, 20161).
That might also partly explained why a notably higher number of people in Hanoi are
married – besides the fact that the capital is rather traditional and slow paced than the
other city – people tend to settle down earlier. Differences in demographic can also be
found in the percentage of ethnic minority and people who were born in another
province (“migrant” here is used rather loosely since some people were given birth to
in another area but spent most of their life in their respective current city and/or do hold
Hanoi/HCMC’s “hokou”). This reflects the fact that HCMC has a relatively more
diverse labor compared to the capital, with more labor belong to ethnic minority and/or
are migrants. In addition, from the data we can also see that Hanoi only attract workers
from 31 other provinces, mostly from the Northern and Central regions; whereas
HCMC has people from almost all over the country (52/58 provinces) migrating here
to work.

1
All data cited are taken or calculated based from numbers published in General Statistics Office of
Vietnam’s website: http://www.gso.gov.vn/Default_en.aspx?tabid=766
30
 Table 1 - Variables description of Hanoi & HCMC
(Numbers other than percentage are average values)
Variable Hanoi Ho Chi Minh City
Age 36.97 36.79
Female 45.37% 46.7%
Ethnic minority 2.01% 7.96%
Urban area 51.5% 83.1%
Married 76.87% 62.72%
Migrant (born in another 16% 33%
province)
Higher education share 41.48% 31.5%
Vocational education share 18.6% 13.36%
Years of schooling 12.5 10.6
Monthly income (thousand 7025.96 7211.55
VND)
Public sector 28.32% 18.18%
Informal sector 38.72% 35.97%
Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

As expected, the capital does have more educated people participate in the workforce;
in particular people with higher education (41.48% vs. 31.5%) and vocational education
(18.6% vs. 13.36%) (note that some people might have both). When converting into
years of schooling, Hanoi also a higher average (12.5 vs. 10.65).
Again, it is not surprising that the capital has a considerably higher number of labor
working in the public sector (28.32% vs. 18.18%). The number one sector that recruits
the most workers in both Hanoi and HCMC, though, are the private firms (34.69% and
49.27% respectively). We can also argue that the overwhelming number of private firms
in HCMC partly reflect the fact that the city has a more encouraging and nurturing
business environment for entrepreneurs to open and pursue their own enterprises.
Lastly, Vietnam’s labor market is characterized by a very low level of unemployment –
1.87 % in the third quarter of 2017 (Report on Labour Force Survey, 2017) due to the
fact that as a developing country, a majority of the population cannot afford to stay idle
and prefer to involve in any kind of job – even ones with low income and unstable,
indecent working conditions. Also remarkably, a dominant share (75.1%) of

31
unemployed young labor (15-24) is formed by people who graduated from college or
above (Report on Labour Force Survey, 2017). The situations are a bit different in Hanoi
and HCMC, however. Unemployment rates are higher, around over 2% according to
the data used for this thesis. The two cities also appear to be more effective in absorbing
higher educated individuals when most of their young laborers without a job are upper-
higher secondary graduates (college graduate also accounts for a considerable number
of unemployed youth, but no way near the country’s number).

Education

According to the General Statistics Office, in 2016, only 11.7% of the Vietnamese
workforce obtained a college, graduate and above degree; 8.9% has been received short
to long-term vocational training. The pictures in Hanoi and HCMC, however, are very
much different; with Hanoi having one third of its workforce owns an university degree,
and only 1.49% has no qualification whatsoever.
Table 2 is a breakdown of educational level attainment in Hanoi and HCMC, in total
and by gender. In line with previous discussions, Hanoi does show that it has a generally
more educated workforce than HCMC – who has over 20% of its labor gets no
qualification or only primary education. Interestingly, the gap between male and female
who receive tertiary education is very small in both cities, with female workers do
slightly better. A lot more men, especially in HCMC, obtain vocational training –
perhaps reflecting the fact that the courses/programs offered at vocational school are
not so female-friendly.

32
Table 2 – Education attainment in Hanoi & HCMC, total and by gender (in %)
Educational level Hanoi Ho Chi Minh City
Total Male Female Total Male Female
No qualification 1.49 62.2 37.8 5.42 56.2 43.8
Primary 7.94 16.4
Lower secondary 14.83 57.9 42.1 17.38 52.9 47.1
Upper secondary 16.73 16.67
Elementary 6.97 61 39 5.69 63.8 36.2
vocational
Middle vocational 3.99 7.44
3-year college 5.97 4.84
University 30.19 48.4 51.9 24.55 49.6 50.4
Master 3.99 1.52
PhD 0.93 0.09
Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

Furthermore, figures 2 and 3 display the distribution of education by types of firm

ownership in Hanoi and HCMC respectively. Both show a relatively similar pattern:
The public sector absorbs the most educated labor; whereas people with no to general
education are mostly likely to take a job in either a household business or a private firm.
Though it seems that in HCMC, people with lower education do have more chances to
work across all firm types, as their composition between different ownerships are less
extreme.

33
Figure 1 – Education attainment in Hanoi, by firm ownership

8.8% 4.1%
4.1% 6.1% 9.8% 7.8%
5.9% 15.6%
17.6% 23.2% 25.1% 14.9% 28.3%
48.4%
32.8%
73.5% 43.6%
68.6% 62.8%
39.8%
36.7%
18.3%
4.0%
No education Primary Lower secondary Upper secondary Vocational College and
education above

Individual household Private Public FDI

Figure 2 – Education attainment in HCMC, by firm ownership

5.4%
2.7% 10.5% 12.0% 10.9% 8.7% 9.8%
6.4% 12.2% 14.5% 18.6%
33.0% 31.2%
42.7%
48.7% 56.0% 49.7%
58.9% 55.5%
40.4%
27.1% 18.6% 23.0%
3.4%
No education Primary Lower secondary Higher secondary Vocational College and
education above
Individual household Private Public FDI

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

Another particular point that needs to be mentioned is that despite Hanoi having an
overall higher number of qualified workers, it is HCMC that has the most competitive
labor market. According to a report from Vietnamworks (the biggest recruiting firm in
Vietnam), job competition in HCMC is 48:1 (for all job sectors); whereas in Hanoi is
39:1. This is not necessary due to size of the population, since many provinces with
much lower number of residents (such as Binh Duong or Da Nang provinces) have the
same or even higher “competitiveness” rate.

Earnings

Finally we take a closer look at the income of the workforce in Hanoi and HCMC.
The two cities do have quite comparable monthly income averages: VND7,025,960 in
34
Hanoi and VND7,211,55 in HCMC (Table 2). Furthermore, figures 3 and 4 (Appendix
2) show that income increases as we go up the educational level. This relationship also
appears to be concave since the wage gaps tend to be larger at the higher end of the
educational distribution, picking up especially at the university level. Notable case is
PhD holders in HCMC actually earn less than master or even university graduates.
Though the number of observation is too low to draw a interpretation. Male workers
earn more than female no matter what attainment level. HCMC’s average salary are
higher than Hanoi’s at all educational levels.
In terms of firm type, figure 5 and 6 (Appendix 3) display the income distribution across
different ownerships in the two cities. The take-away here is that even though the public
sector absorbs the most highest-educated labor, it cannot pay these labors as well as the
private and FDI sectors. Foreign companies are especially rewarding toward higher
educated workers.
Figure 5 (Appendix 4) shows the average monthly income in both cities by job sector.
In general, jobs in information & communication; finance, banking and international
organizations/agencies (e.g. non-profit organizations) are the most well-paid. It is also
interesting to note is that while 41.9% of Vietnamese workforce are still currently
engaging in different agricultural activities; there is almost no representative of this
sector in Hanoi and HCMC. Sectors that employ the most number of workers are
manufacturing, construction and different business activities related to motorbike.

Lastly, there are some interesting findings that cannot discussed in details. Such as
migrant workers tend to be more well-educated, more likely to work in a foreign firm
and averagely paid more than local labors. The reasoning is that both Hanoi and HCMC
are big cities with specially competitive working environment. Therefore it is possible
that only the “bests” of migrant workers can come and survive in these cities. And even
then they still might have to work extra hard in order to compete with the locals due to
migrant discrimination, their lack of connection or city household registration.

35
4.2 Returns to education in Vietnam: Hanoi & HCMC
General situation in Hanoi and HCMC

Table 3 (Appendix C) are the regression results running for the whole dataset using City
and educational level dummy variables to have an overview look of the 2 cities.
Regression are run at 3 levels: (1) basic Mincer model, (2) with personal characteristics
variables and (3) with all controlling variables. R squared are higher in models (2) and
(3), suggesting that those added variables have explanatory ability toward earnings
variation. In general, HCMC offers a higher level of monthly salary than the capital, in
particular 4.45%. Most independent variables show a relatively similar patterns to
previous study using national - level data set. The most notable finding, however, is that
general education (primary, secondary and high schools) might not bring wage
premium (coefficients are insignificant) to employees in these two cities – in contrast
with results from previous studies using nationwide data where it shows all educational
levels offer incentives. Specifically, an additional year spent in elementary vocational
rewards a 2.47% of wage premium; the numbers are 1.85% for higher vocational/
professional school, 2.29% for 3-year college, 3.78% for university, 4.47 for MA/MSc
and 5.56% for PhD. Accordingly, after general education, the higher credentials (except
for 3 year college) that one can obtain, the more payoff she will get in the labor force.
In term of year measurement, rate of return to education is 18% for every year of
schooling (regression table results are not reported here). This is higher than the global
average of 10% of private returns, with consideration to costs incurred for attending
school (Montenegro & Patrinos, 2014).

Wage premium for education and experience

The main focus of our study are presented in the regressions in Table 4 and 5 (Appendix
C), where Year of Schooling and Age variables alternately are put to interact with City
dummy. This is to find if one city offers wage premium for an additional year of
schooling and experience over the other city or not.

36
From regression (3) in table 4, we can see that indeed Hanoi has a wage premium of
1.95% for an additional school year compared to HCMC - even when controlling
variables related to individual characteristic and employment information. To check
whether or not this is due to the PhD holders in Hanoi skewed the results, the author
also runs a separate regression (3*) for people with university degree and below. With
this subgroup, Hanoi still offers a wage premium of 1.75% over HCMC.
One of the reasons that can explain the premium return to schooling in Hanoi is the
higher number of SOEs. Vietnamese state-owned firms do not have the soft or hard
infrastructures to hire people based on multiple criterions; and instead rely heavily on
credentials as one of main factors not just to recruit but also to standardize and
categorize their officers. The author did a quick regression and found that based on this
dataset, SOEs offer over 4% for an additional schooling year – whereas for the rest of
the firm types (Private, Individual business household, FDI), this number is only less
than 1%. Still, as the regressions already include a dummy variable for firm type, the
1.95% premium might lie on other reasons. Perhaps of the observations that the capital
is still somewhat more conservative and put greater significance in education as
discussed earlier in the Introduction? Another problem with SOEs is that they actually
pay less than any other types of ownership, even individual household businesses
(results from Table 3), implying that the formulated wage scale in the public sector is
overestimating the actual value of education. With almost one third of its firms in public
sector, Hanoi has a bigger challenge in reforming ownership and management
incentives to improve the overall efficiency and productivity.
Table 4 goes through the same process but with interaction between Age (proxy for
experience) and City dummies. Again, it confirms the prediction at the beginning that
HCMC is more rewarding toward experience, though the number is quite small – at
0.66%. Note that with the Vietnamese currency, salaries almost always go to the
millions. Hence while these differences sound small in number, they do hold a certain
level of significance. As previous sections have argued, HCMC is much more dynamic
and diverse in terms of business environment, labor structure, industry structure.

37
Therefore it is understandable that it does not put as much emphasis on education and
pay more attention to other aspects, in this case - experience, than Hanoi.

Heckman correction bias model

Finally, regressions (4) in in Table 3, 4, 5 were run with the Heckman correction model
to address the sample selection bias issue, using four explanatory variables in the
Selection equation: Number of people in the household, Dummy for whether the person
is the household or not, Dummy for whether the person is married and Non-wage
household income. Of which, only marital status has no significant impact toward the
results.
In general, Heckman model does show that the Mincerian equation (even with all
controlling variables) constantly overestimates the return to education in the two cities,
however the size of the bias are not large. Figure 8 visualizes the differences in the
return to education in each educational level with and without correcting for selection
bias; while table 6 details the coefficients and their respective standard errors.

100 87.7
90
80 72.2
70 58.1
60 49.2
50
40 35.2
29.3
30 19.5 17.5 20.4
15.9 12.9 15.4
20
10
0
Elementery Middle 3 year college University Master PhD
vocational vocational -
Professional
school
OLS Heckman

Figure 8 – Return to schooling in Hanoi and HCMC by educational level, Mincer vs.
Heckman (in per cent)

38
 OLS Heckman
Elementary vocational 0.195*** 0.175***
(0.0462) (0.0438)
Middle vocational / 0.159*** 0.129***
Professional school (0.0452) (0.0428)
3-year college 0.204*** 0.154***
(0.0500) (0.0476)
University 0.352*** 0.293***
(0.0470) (0.0446)
Master 0.581*** 0.492***
(0.0610) (0.0594)
PhD 0.877*** 0.722***
(0.101) (0.103)
Table – Detail comparison of results from Mincer & Heckman models

In term of wage premium difference between the two cities for an additional year of
school and experience, Hanoi still pays 1.5% (OLS: 1.95%) for education and HCMC
still pays 0.632% more for experience (OLS: 0.656%). All in all, biases are less than 10
per cent in most cases, which is the result consistent to previous studies (Doan & Gibson,
2010; Phan & Coxhead, 2013).

39
 CONCLUSION
The main conclusions from this empirical analysis are that not only the labor market in
Hanoi and HCMC draw a meaningfully different picture from previous studies dealing
with country-level data; but there are differences existed between these two important
cities themselves as well.
As the two biggest economical hubs in Northern and Southern Vietnam, Hanoi and
HCMC are having a more sophisticated and competitive labor force compared to rest
of the country. General education no longer pays premium; instead workers need to
have at least a vocational degree to significantly increase their wages. This is
contradicted with the current problems of vocational schools not being able to draw
enough students (due to quality, due to the common perception that these schools are
only for kids who cannot pursue a more “proper” education), some were eventually
forced to close down. Lesson learned here is more and more proper career-orientation
activities need to be implemented at the secondary education levels, encouraging
students to at least earn a vocational education before entering the labor market – as the
markets in these two cities are in high need of workers with such vocational skills.

Regarding the differences between Hanoi and HCMC themselves, statistical results
have revealed that in general HCMC offers higher salaries; and also, the capital pays
more for education while the younger city is more rewarding toward experiences
(though not in a huge gap). These are after controlling multiple variables including
employment ownership type – one of the factors that can explain these particular
differences. We again come back to the argument that Hanoi has traditionally put greater
weight in education and status, which in this case might have overestimated the market
value of education. If this is somehow true, there is no doubt that it has contributed to
the existing social phenomenon (and corruption issues) of people “buying” credentials
and bribing in exchange for awards, titles to get jobs and higher salaries; along with
hindering the capital from hiring the “best” people – especially in areas when actual
hand-on experiences are more important than a degree, such as the creative sectors. Not

40
to mention with a higher number of (large) state-owned and equitisatized (Vietnamese
English term refers to SOEs who are converted into a public limited companies or a
corporation) firms, the reality of degrees are used as tools to move up the career ladder
is even more prominent. This is not to say that this an exclusive problem of the capital
– these are actually happening everywhere in the country, HCMC included. But within
this thesis, statistical results indicate that this seems to be more of a problem for Hanoi
than the other city.
In fact, the old capital can look at HCMC to find solutions for their issues. There are
many reasons that make HCMC become the economic leader of country, one of which
is how it successfully built and developed an environment that foster entrepreneurs and
business-minded people. It is those new start-ups, private firms that makes the city a
diverse, flexible, open place for most skillful (and not necessarily most well-educated)
workers to come and prosper. As of 2016, HCMC accounts for one third (200,000 over
600,000) of the number of operating firms in the country. There is also a noticeable
trend of well educated, capable and enterprising workers moving from Hanoi to base in
HCMC as the later can offer more job opportunities and overall better business and
working environment. With a population of 90 million people, Vietnam in general and
Hanoi in particular has a lot of room to expand itself. The capital possesses the
infrastructures, the educated workforce and the political advantages; what it needs now
is a more open, tolerant and pro-business mindset. And even though it might sound like
a natural way to go, it is actually quite a challenge for the capital. Because if there is
one thing the Northerners are afraid of, it is change.

With that being said, Hanoi can make use of its large pool of highly-educated workers
to develop more knowledge-intensive activities; whereas HCMC can achieve further
(say, at the robust growth rate of leading cities in China) by improving the quality of
general and especially non-formal education to meet the labor’s diverse learning
demand. The younger city has been receiving increasing pressure from the needs to
invest more in infrastructures and education. On the other hand, we can really see that
in some ways, these two cities are very compatible and have the equipment to make up
41
for the other’s limitations. How Hanoi and HCMC can learn to complement and
compensate for each other will contribute largely to development of not only the two
cities themselves but certainly the country as well.

In conclusion, this thesis once again revisited the linkage between educational
attainment and labor market outcome in Vietnam; however this time in a more
specifically geographical scale and take a relatively different direction. There are much
more to discuss and explore further, though this is a good starting point to showcase the
need of studying return to education in a smaller setting – either city/province, or region
and economic zones. As Vietnam is a developing economy with highly economical,
developmental and educational heterogeneity within the country, studies target certain
geographical locations can offer more specific implications and solution courses for
policy makers.

42
 LIMITATION
The thesis, unfortunately, could not address the ability bias issue, as discussed earlier.
In the topic of ability measurement, the study could not also take into account cognitive
skills (critical thinking, problem-solving) and behavioral skills (team work,
communication) into the analysis. These are the skills that the labor force is still
severely lack, but have been becoming increasingly important to employers in a now
modern market economy setting of Vietnam. So important that the World Bank has
conducted a detail research in 2014 called “Skilling up Vietnam” to tackle this issue.
Again, such complex variables are not included in the VHLSS dataset.
In addition, Heckman correction model can use even more appropriate instruments such
as number of children (especially under 6); and the author could have gone further by
testing the chosen instrumental variables for invalid / weak instruments. And lastly, the
thesis did not discuss about the education quality in Vietnam, which is a subject that
can shed light to many problems, but is in a very different realm from the restriction of
this thesis.

43
 REFERENCES
[1] Altonji, J., Thomas, A. D. (1996). The effects of family characteristics on the return
to schooling. Review of Economics and Statistics, 78 (4), 692-704.
[2] Altonji, J., Thomas, A. D. (1996). Using siblings to estimate the effect of school
quality on Wages. Review of Economics and Statistics, 78 (4), 665-671.
[3] Angrist, J. D., Krurger, A. B. (1991). Does Compulsory School Attendance Affect
Schooling and Earnings?. Quarterly Journal of Economics, 106 (4), 979-1014.
[4] Arbex, M., Galao, A. F., Gomes, F. (2007). Heterogeneity in the Returns to
Education and Informal Activities. Insper Working Paper (Insper Instituto de
Ensino e Pesquisa, Sao Paulo).
[5] Arias, O., Hallock, K. F. & Sosa-Escudero, W (2001). Individual Heterogeneity in
the Returns to Schooling: Instrumental Variables Quantile Regression Using Twins
Data. Empirical Economics, 26 (1), 7-40.
[6] Ashenfelter, O., Rouse, C. E. (1998). Income, schooling and ability: evidence from
a new sample of identical twins. Quarterly Journal of Economics, 113 (1), 253-
284.
[7] Ashenfelter, O., Zimmerman, D. (1997). Estimates of the return to schooling from
sibling data: fathers, sons and brothers. Review of Economics and Statistics 79, (1),
1-9.
[8] Backman, M. & Bjerke, L. (2009). "Returns to Higher Education - a regional
perspective," Working Paper Series in Economics and Institutions of Innovation
171, Royal Institute of Technology, CESIS - Centre of Excellence for Science and
Innovation Studies.
[9] Becker, G. (1962). Investment in Human Capital: A Theoretical Analysis. Journal
of Political Economy, 70 (9).
[10] Becker, G. S. (1964). Human capital: a theoretical and empirical analysis, with
special reference to education. New York, NY: Columbia University Press.
[11] Becker, G. S., Tomes, N. (1986). Human Capital and the Rise and Fall of Families.
Journal of Labor Economics, 4 (3), S1-S39.
[12] Belman, D., Heywood, J. (1991). Sheepskin Effects in the Returns to Education:
An Examination on Women and Minorities. The Review of Economics and
Statistics, 73 (4), 720-24.
[13] Belzil, C., Hansen, J. (2002). Unobserved Ability and the Return to Schooling.
Econometrica, 70 (5), 2075 - 2091.
[14] Bronars, S., Oettinger, G. (2006) Estimates of the return to schooling and ability:
evidence from sibling data. Labour Economics, 13 (1), 19-34.
[15] Card, D. (1993). Using Geographic Variation in College Proximity to Estimate the
Return to Schooling. Working Paper No. 4483 (NBER, Cambridge, MA).
[16] Card, D. (1999). The Causal Effect of Education on Earnings. Handbook of Labor
Economics, 3 (1), 1801 - 1863.
[17] Card, D. (2001). Estimating the return to schooling: Progress on some persistent
econometric problems. Econometrica, 69 (5), 1127-1160
44
[18] Chatelain, J., Ralf, K. (2010). Inference on Time-Invariant Variables using Panel
Data: A Pre-Test Estimator with an Application to the Returns to Schooling. PSE
Working Papers hal-00492039, HAL.
[19] Cieslik, A. & Rokicki, B. (2016). Individual Wages and Regional Market Potential.
Economics of Transition, 24 (4), 661-682.
[20] Conneely, K., Uusitalo, R. (1997). Estimating Heterogeneous Treatment Effects in
the Becker Schooling Model. Unpublished Discussion Paper, Princeton University
Industrial Relations Section.
[21] Cooper, G. (2016, September 28). A Tale Of Two Cities In Vietnam. Retrieved from:
https://www.duanemorris.com/articles/static/cooper_two_cities_law360_1016.pd
f
[22] Dapice, D. (2003). Vietnam’s Economy: Success Story or Weird Dualism? A
SWOT Analysis. Journal of Political Economy, 70 (9).
[23] Dickson, M., Smith, S. (2011). What determines the return to education: an extra
year or hurdle cleared?. Working paper No. 11/256 (University of Bristol, Bristol,
UK).
[24] Doan, T. (2011). Labour Market Returns to Higher Education in Vietnam. SSRN
Journal.
[25] Doan, T., & Tran, Q. T. & Le, Q. (2016). Lost in Transition? Declining Returns to
Education in Vietnam. Working Papers in Economics No.16/01 (University of
Waikato).
[26] Fally, T., Paillacar, R., & Terra, C. (2010). Economic geography and wages in
Brazil: Evidence from micro-data. Journal of Development Economics, 91 (1).
[27] Florida, R. L. (2012). The rise of the creative class, revisited. New York: Basic
Books.
[28] Friedman, M. (1962). Capitalism and Freedom (ISBN 0-226-26421-1 ed.).
Chicago, IL: University of Chicago Press
[29] Gallup, J. L. (2002). The wage labor market and inequality in Viet Nam in the
1990s (English). World Bank Policy Research Working Paper No. WPS 2896. (The
World Bank, Washington DC).
[30] General Statistics Office, Ministry of Planning and Investment. (Quarter 3, 2017).
Report on Labour Force Survey: Quarter 3, 2017. Retrieved from
http://www.gso.gov.vn/default_en.aspx?tabid=515&idmid=5&ItemID=18724
[31] Gensowski, M. (2018). Personality, IQ, and lifetime earnings. Labour Economics,
51, 170-183.
[32] Glewwe, P., Patrinos, H. A. (1998). The role of the private sector in education in
Vietnam : evidence from the Vietnam Living Standards Survey (English). Living
standards measurement study (LSMS) working paper ; no. LSM 132. (The World
Bank, Washington DC).
[33] Griliches, Z. (1977). Estimating the Returns to Schooling: Some Econometric
Problems. Econometrica, 45 (1), 1 - 22
[34] Hanushek, E. A., & Kim, D. (1995). Schooling, Labor Force Quality, and
Economic Growth (Working Paper No. 5399). Retrieved from National Bureau of
Economic Research website: http://www.nber.org/papers/w5399
45
[35] Harmon, C., Walker, I. (1995). Estimates of the Economic Return to Schooling for
the United Kingdom. American Economic Review, 85 (5), 1278-86.
[36] Hausman, J., Taylor, W. (1981). Panel Data and Unobservable Individual Effects.
Econometrica, 49 (6), 1377-98.
[37] Heckman, J. J. (1979). Sample Selection Bias as a Specification Error.
Econometrica, 47 (1), 153-161.
[38] Heckman, J. J., Lochner, L. J & Todd, P. E. (2005). Earnings Functions, Rates of
Return and Treatment Effects: The Mincer Equation and Beyond. Working paper
No. 11544 (NBER, Cambridge, MA).
[39] Hering, L. & Poncet, S. (2010). Market Access and Individual Wages: Evidence
from China. Review of Economics and Statistics, 92 (1).
[40] Hungerford, T., Solon, G. (1987). Sheepskin Effects in the Returns to Education.
The Review of Economics and Statistics, 69 (1), 175-77.
[41] Ichino, A., Winter-Ebmer, R. (1998). Lower and Upper Bounds of Returns to
Schooling: An exercise in IV estimation with different instruments. Retrieved from
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=141407
[42] Kane, T. J, Rouse, C. E. (1993). Labor market returns to two- and four-year
colleges: is a credit a credit and do degrees matter?. Working paper no. 4268
(NBER, Cambridge, MA).
[43] Kazuhiro, Y. (2007). Urbanization, informal sector, and development, Journal of
Development Economics, 84 (1), 76-103.
[44] Lall, A., & Sakellariou, C. (2010). Evolution of Education Premiums in Cambodia:
1997–2007. Asian Economic Journal, 24 (4), 333–354.
[45] Lang, K. (1993). Ability Bias, Discount Rate Bias and the Return to Education.
MPRA Paper 24651, University Library of Munich, Germany.
[46] Liu, A. Y. C. (2004). Sectoral gender wage gap in Vietnam. Oxford Development
Studies, 32 (2), 225-239.
[47] Liu, A. Y. C. (2006). Changing wage structure and education in Vietnam, 1992–98.
Economics of Transition, 14 (4), 681-706.
[48] McKinsey Global Institute. (February 2012). Sustaining Vietnam’s growth: The
productivity challenge. Washington, DC.
[49] Mann, P., & Kapoor, B. (1988). Earnings differentials between public, private and
joint sector in Punjab (India). The Journal of Development Studies, 25 (1), 97-111.
[50] Mincer, J. (1974) Schooling, Experience, and Earnings. National Bureau of
Economic Research. New York: Columbia University Press.
[51] Montenegro, C. , E. & Patrinos, H. , A. (2014). Comparable estimates of returns
to schooling around the world (English). Policy Research working paper ; no. WPS
7020. Washington, DC: World Bank Group. Retrieved from:
http://documents.worldbank.org/curated/en/830831468147839247/Comparable-
estimates-of-returns-to-schooling-around-the-world
[52] Moock, P. R., Patrinos, H. A. & Venkataraman, M. (1998). Education and Earnings
in a Transition Economy (Vietnam). World Bank Policy Research Working Paper
No. 1920. (The World Bank, Washington DC).
[53] Murphy, K., Welch, F. (1990). Empirical Age-Earnings Profiles. Journal of Labor
46
 Economics, 8 (2), 202-29.
[54] Nguyen, K. Q. & Nguyen, C. C (2008). Education in Vietnam: development history,
challenges and solutions. An African exploration of the East Asian education
experience. Retrieved from
http://documents.worldbank.org/curated/en/565731468314994155/An-African-
exploration-of-the-East-Asian-education-experience
[55] The Nielson Company. (June 2009). HCMC v Hanoi: Understanding consumer
differences. Retrieved from
http://www.nielsen.com/content/dam/corporate/t3/vietnam/reports/Nielsen%20Vi
etnam_%20HCMC%20v%20Hanoi%20consumer%20differences_June%202009
_slide7edit.pdf.
[56] Peet, E. D., Fink, G., Fawzi, W. (2015). Returns to education in developing
countries: Evidence from the living standards and measurement study surveys.
Economics of Education Review, 22 (C), 69-90.
[57] Pham, H. T., & Reilly, B. (2007). The Gender Pay Gap In Vietnam, 1993-2002: A
Quantile Regression Approach. Journal of Asian Economics, 185 (5), 775-806.
[58] Pham, H. T., & Reilly, B. (2009). Ethnic wage inequality in Vietnam, International
Journal of Manpower, 30 (3), 192-219.
[59] Phan, D., & Coxhead, I. (2013). Long-run costs of piecemeal reform: Wage
inequality and returns to education in Vietnam, Journal of Comparative Economics,
41 (4), 1106-1122.
[60] Psacharopoulos, G. (1985). Returns to education a further international update and
implications. The Journal of Human Resources, 20 (4), 583-604.
[61] Psacharopoulos, G. (1994). Returns to investment in education: A global update.
World Development, 22 (9), 1325-1343.
[62] Psacharopoulos, G., & Patrinos, H. , A. (2002). Returns to investment in
education : A Further Update (English). Policy, Research working paper series ; no.
WPS 2881. Washington, DC: World Bank. Retrieved from:
http://documents.worldbank.org/curated/en/512891468739485757/Returns-to-
investment-in-education-a-further-update
[63] Puhani, P. (2000). The Heckman Correction for Sample Selection and Its Critique.
Journal of Economic Surveys, 14 (1), 53-68.
[64] Quek, C. (2017, May 21). Vietnam through the eyes of local startups. Retrieved
from: https://www.techinasia.com/talk/vietnam-eyes-local-startups
[65] Roback, J. (1982). Wages, Rents, and the Quality of Life. Journal of Political
Economy, 90 (6), 1257-78.
[66] Roback, J. (1988). Wages, Rents, and Amenities: Differences among Workers and
Regions. Economic Inquiry, 26 (1), 23-41.
[67] Spence, M. (1973). Job Market Signaling. Quarterly Journal of Economics, 87 (3),
355-374.
[68] Patrinos, H. A. (2016). Estimating the return to schooling using the Mincer
equation, IZA World of Labor, Institute for the Study of Labor (IZA), (278), 278-
278.
[69] Phan, D., & Coxhead, I. (2013). Long-run costs of piecemeal reform: Wage
47
 inequality and returns to education in Vietnam, Journal of Comparative Economics,
41 (4), 1106-1122.
[70] Psacharopoulos, G. (1985). Returns to education a further international update
and implications. The Journal of Human Resources, 20 (4), 583-604.
[71] Psacharopoulos, G. (1994). Returns to investment in education: A global update.
World Development, 22 (9), 1325-1343.
[72] Psacharopoulos, G., & Patrinos, H. , A. (2002). Returns to investment in
education : A Further Update (English). Policy, Research working paper series ; no.
WPS 2881. Washington, DC: World Bank. Retrieved from:
http://documents.worldbank.org/curated/en/512891468739485757/Returns-to-
investment-in-education-a-further-update
[73] Puhani, P. (2000). The Heckman Correction for Sample Selection and Its Critique.
Journal of Economic Surveys, 14 (1), 53-68.
[74] Roback, J. (1982). Wages, Rents, and the Quality of Life. Journal of Political
Economy, 90 (6), 1257-78.
[75] Roback, J. (1988). Wages, Rents, and Amenities: Differences among Workers and
Regions. Economic Inquiry, 26 (1), 23-41.
[76] Spence, M. (1973). Job Market Signaling. Quarterly Journal of Economics, 87 (3),
355-374.

48
 FIGURES
Figure 3 – Average monthly income in Hanoi, by education and gender (in thousand VND)

15,951
13,936

10,512 10,771 11,167

8,841
6,542 7,098
5,1865,356 5,201 5,873
4,870
4,489
3,496 3,733 4,0724,694

No Primary Lower Upper Vocational College University Master PhD

education seconday secondary education

Female Male Expon. (Female) Expon. (Male)

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

Figure 4 – Average monthly income in HCMC, by education and gender (in thousand VND)

17,298

13,682
11,693
10,833
9,154 9,833
7,582 7,8928,106
6,696 6,791 6,104
5,623 5,897 5,390
4,546 5,148
4,094

No Primary Lower Upper Vocational College University Master PhD

education seconday secondary education

Female Male Expon. (Female) Expon. (Male)

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

49
 Figure 5 – Average monthly income in Hanoi, by education and firm ownership (in
thousand VND)

Individual household Private Public FDI 16,535 16,583

16,121
14,715
13,834

11,718
10,539
10,000
8,658 8,583
8,226
7,181 7,105 7,334
6,730
6,163 6,035 6,110 6,339 6,366 6,265
5,397
5,356 5,511 5,446 5,571 5,632 5,491 5,350
4,903 4,715
4,456 4,528 4,505
4,077

No Primary Lower Upper Vocational College University Master PhD

qualification seconday secondary education

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

Figure 6 – Average monthly income in HCMC, by education and firm ownership (in
thousand VND)
18,962
Individual household Private Public FDI
17,044

13,798
12,884

10,879
10,339 10,000 10,333

8,759
8,134
7,374 7,351
6,552 6,648 6,707
6,491 6,800 6,536 6,453
6,315 5,937 5,676 6,290 5,948
5,383
5,307 5,356 5,549 5,327 5,673
4,805 5,032 5,020

No Primary Lower Upper Vocational College University Master PhD

qualification seconday secondary education

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

50
 Figure 7 – Average monthly income in Hanoi & HCMC by Job sector (in thousand VND)

Water Whole Activit House

Activit
supply sale, Servic ies of hold
Proces Admin ies of
; retail, es of Financ Profes the Arts, emplo
sing Infor istrati Health intern
mana and accom e, sionali comm Educa recrea yment
Minin and Trans matio on care ationa
geme repair modat banki sm, unist tion tion Other gener
g and manuf Constr port, n and and and l
nt and of ion, ng, scienc party and and servic ated
quarry acturi uction wareh comm suppo social organi
treat autom food and e and and trainin entert es by
ing ng ouse unicat rting assista zation
ment obiles, and insura techn socio- g ainme house
indust ion servic nce s and
of motor bever nce ology politic nt holds;
ries es agenci
sewer bikes, ages al house
es
age… scoo… orga… hold…
Hanoi 10,5 6,01 5,84 5,85 7,37 8,83 5,76 10,2 11,6 10,4 6,63 6,55 6,64 7,52 5,79 4,71 3,52 19,2
HCM 9,46 6,65 7,75 8,35 7,17 7,60 5,79 11,3 10,2 10,2 8,10 6,31 6,53 7,85 5,74 5,55 3,86

Source: Vietnam Household Living Standard Survey 2016 - Author’s own calculations

51
 APPENDIX A
Vietnam Education System

Adapted from Source: https://wenr.wes.org/2017/11/education-in-vietnam

52
 APPENDIX B
Dummy variables description

Variable Description
Education Level Equivalent year of
schooling
1 = No qualification (Reference group) 0
2 = Primary education 5
3 = Lower secondary 4
4 = Upper secondary 3
5 = Elementary vocational 0.5
6 = Middle vocational /Professional 1–2
school
7 = 3-year college 3
8 = University 4
9 = MA/MSc 2
10 = PhD 3
Gender 1 = Female (Reference category)
2 = Male
Marital status 1 = Married (Reference category)
2 = Not married
Ethnicity 1 = Kinh (Reference category)
2 = Minor ethnicity
Area 1 = Rural (Reference category)
2 = Urban
Firm type 1 = Public (Reference category)
2 = Private
3 = Individual household business
4 = Foreign Direct Investment (FDI)
Job sector 1 = Industry and Trade (Reference category)
2 = Construction
3 = Service
Job 1 = Low-skilled workers (Reference category)
level/occupation 2 = Manual labor
3 = Assembly workers
4 = Skilled sales & service staff
5 = Office staff
6 = Middle level expert
7 = High level expert
8 = Leaders/Manager

53
 APPENDIX C
Table 3 – Returns to education in Hanoi & HCM
(1) (2) (3) (4)
With personal All controlling Heckman
VARIABLES Basic Mincer
characteristics variables selection model

Age 0.0802*** 0.0796*** 0.0788*** 0.0724***

(0.00390) (0.00380) (0.00367) (0.00348)
Age2 -0.00101*** -0.00102*** -0.000994*** -0.000907***
(4.89e-05) (4.75e-05) (4.60e-05) (4.33e-05)
Hanoi -0.113*** -0.0751*** -0.0445*** -0.0462***
(0.0145) (0.0154) (0.0150) (0.0143)
Primary 0.0237 0.0257 0.00878 -0.00204
(0.0424) (0.0414) (0.0402) (0.0383)
Lower secondary 0.0506 0.0447 0.00454 0.000746
(0.0416) (0.0406) (0.0398) (0.0375)
Upper secondary 0.165*** 0.154*** 0.0932 0.0665
(0.0417) (0.0409) (0.0406) (0.0384)
Elementary vocational 0.322*** 0.263*** 0.195*** 0.175***
(0.0472) (0.0464) (0.0462) (0.0438)
Higher vocational /
0.259*** 0.242*** 0.159*** 0.129***
Professional school
(0.0447) (0.0439) (0.0452) (0.0428)
3 year college 0.364*** 0.352*** 0.204*** 0.154***
(0.0483) (0.0474) (0.0500) (0.0476)
University 0.605*** 0.576*** 0.352*** 0.293***
(0.0403) (0.0403) (0.0470) (0.0446)
MA/MSc 0.840*** 0.788*** 0.581*** 0.492***
(0.0569) (0.0566) (0.0610) (0.0594)
PhD 1.167*** 1.062*** 0.877*** 0.722***
(0.104) (0.102) (0.101) (0.103)
Male 0.217*** 0.214*** 0.210***
(0.0138) (0.0134) (0.0129)
Minor ethnicity 0.0336 0.0272 0.0410
(0.0333) (0.0322) (0.0301)
Migrant 0.0868*** 0.0633*** 0.0565***
(0.0165) (0.0157) (0.0155)
Urban 0.0635*** 0.0638*** 0.0434***
(0.0169) (0.0165) (0.0158)
Private 0.231*** 0.205***
(0.0186) (0.0177)
Individual household 0.112*** 0.108***
54
 (0.0240) (0.0224)
FDI 0.354*** 0.336***
(0.0273) (0.0265)
Manual labor 0.120*** 0.112***
(0.0272) (0.0251)
Assembly workers 0.142*** 0.148***
(0.0292) (0.0276)
Skilled sales & service
staff 0.0837*** 0.0747***
(0.0294) (0.0274)
Office staff 0.0608 0.135***
(0.0376) (0.0349)
Middle level expert 0.256*** 0.242***
(0.0366) (0.0345)
High level expert 0.337*** 0.343***
(0.0371) (0.0351)
Leader/Manager 0.687*** 0.629***
(0.0523) (0.0500)
Construction 0.0559*** 0.0533**
(0.0215) (0.0208)
Service 0.0267 0.0155
(0.0178) (0.0169)
Household size -0.0833***
(0.0116)
Head household
-0.149***
dummy
(0.0402)
Marriage -0.0261
(0.0388)
Non-wage income 2.18e-06***
(1.32e-07)
Constant 6.906*** 6.759*** 6.479*** 0.868***
(0.0828) (0.0812) (0.0839) (0.0564)

Observations 4,706 4,706 4,606 5,551

R-squared 0.280 0.321 0.393
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

55
 Table 4 – Returns to education in Hanoi & HCM – Wage premium for schooling year
(1) (2) (3) (3)* (4)
With personal All controlling All controlling variables (Uni
Basic Mincer Heckman selection model
VARIABLES characteristics variables & below)

Age 0.0377*** 0.0340*** 0.0367*** 0.0367*** 0.0335***

(0.00201) (0.00192) (0.00184) (0.00185) (0.00173)
Age2 -0.000808*** -0.000758*** -0.000792*** -0.000792*** -0.000710***
(4.21e-05) (4.03e-05) (3.87e-05) (3.87e-05) (3.57e-05)
Hanoi -0.475*** -0.339*** -0.269*** -0.241*** -0.212***
(0.0457) (0.0448) (0.0433) (0.0446) (0.0389)
yearofschooling 0.0463*** 0.0375*** 0.0137*** 0.0120*** 0.0106***
(0.00235) (0.00229) (0.00263) (0.00264) (0.00254)
Hanoi#c.yearofschooling 0.0276*** 0.0226*** 0.0195*** 0.0175*** 0.0150***
(0.00368) (0.00341) (0.00327) (0.00343) (0.00298)
Male 0.202*** 0.213*** 0.213*** 0.209***
(0.0143) (0.0135) (0.0136) (0.0130)
Minor ethnicity 0.0679** 0.0536* 0.0585* 0.0603**
(0.0339) (0.0321) (0.0320) (0.0301)
Migrant 0.0942*** 0.0618*** 0.0621*** 0.0571***
(0.0170) (0.0159) (0.0161) (0.0157)
Urban 0.0908*** 0.0692*** 0.0676*** 0.0495***
(0.0177) (0.0169) (0.0168) (0.0161)
Private 0.244*** 0.255*** 0.218***
(0.0189) (0.0192) (0.0180)
Individual household 0.143*** 0.146*** 0.130***

56
 (0.0244) (0.0245) (0.0228)
FDI 0.371*** 0.386*** 0.347***
(0.0277) (0.0281) (0.0269)
Manual labor 0.109*** 0.114*** 0.104***
(0.0274) (0.0272) (0.0253)
Assembly workers 0.124*** 0.129*** 0.140***
(0.0288) (0.0286) (0.0273)
Skilled sales & service
staff 0.0617** 0.0647** 0.0565**
(0.0295) (0.0293) (0.0277)
Office staff 0.0898** 0.0807** 0.172***
(0.0376) (0.0377) (0.0351)
Middle level expert 0.301*** 0.319*** 0.281***
(0.0349) (0.0348) (0.0332)
High level expert 0.513*** 0.516*** 0.499***
(0.0323) (0.0322) (0.0307)
Leader/Manager 0.763*** 0.745*** 0.692***
(0.0507) (0.0531) (0.0487)
Construction 0.0455** 0.0548** 0.0418**
(0.0217) (0.0217) (0.0210)
Service 0.0117 0.00606 -0.00216
(0.0179) (0.0181) (0.0170)
Household size -0.0833*** -0.0768***
(0.0116) (0.0115)
Head household dummy -0.149*** -0.175***
(0.0402) (0.0396)
Marriage -0.0261 0.0598264

57
 (0.0388) (0.0385)
Non-wage income 2.18e-06*** 2.21e-06***
(1.32e-07) (1.31e-07)
Constant 7.868*** 7.736*** 7.569*** 7.572*** 0.842***
(0.0348) (0.0353) (0.0449) (0.0451) (0.0534)

Observations 4,743 4,743 4,640 4,482 5,585

R-squared 0.226 0.267 0.375 0.361
Standard errors in
parentheses
*** p<0.01, ** p<0.05, *
p<0.1

58
 Table 5 – Returns to education in Hanoi & HCM – Wage premium for experience
(1) (2) (3) (4)
With personal All controlling Heckman selection
Basic Mincer
VARIABLES characteristics variables model

Age 0.0405*** 0.0379*** 0.0397*** 0.0364***

(0.00207) (0.00202) (0.00193) (0.00185)
Age2 -0.000779*** -0.000759*** -0.000789*** 0.0364***
(4.13e-05) (4.02e-05) (3.87e-05) (0.00185)
Hanoi#c.age -0.00746*** -0.00774*** -0.00656*** -0.00632***
(0.00116) (0.00113) (0.00106) (0.000984)
Year of schooling 0.0530*** 0.0467*** 0.0208*** 0.0172***
(0.00188) (0.00201) (0.00250) (0.00235)
Hanoi -0.0192 -0.0806 -0.0937 -0.0859
(0.0264) (0.0263) (0.0246) (0.0235)
Male 0.204*** 0.215*** 0.210***
(0.0143) (0.0135) (0.0129)
Minor ethnicity 0.0622* 0.0492 0.0551*
(0.0339) (0.0321) (0.0301)
Migrant 0.105*** 0.0716*** 0.0619***
(0.0170) (0.0158) (0.0157)
Urban 0.104*** 0.0796*** 0.0585***
(0.0174) (0.0167) (0.0159)
Private 0.234*** 0.209***
(0.0188) (0.0179)
Individual household 0.129*** 0.121***
(0.0243) (0.0226)
FDI 0.363*** 0.339***
(0.0277) (0.0268)
Manual labor 0.102*** 0.0992***
(0.0273) (0.0253)
Assembly workers 0.129*** 0.144***
(0.0288) (0.0272)
Skilled sales & service
staff 0.0599** 0.0557**
(0.0295) (0.0276)
Office staff 0.0893** 0.170***
(0.0376) (0.0351)
Middle level expert 0.298*** 0.279***
(0.0349) (0.0332)
High level expert 0.513*** 0.497***
(0.0323) (0.0306)
Leader/Manager 0.781*** 0.699***

59
 (0.0506) (0.0488)
Construction 0.0477** 0.0443**
(0.0217) (0.0209)
Service 0.0103 -0.00349
(0.0179) (0.0170)
Household size -0.0813***
(0.0116)
Head household
-0.168***
dummy
(0.0402)
Marriage -0.0531
(0.0385)
Non-wage income 2.22e-06***
(1.31e-07)
Constant 7.672*** 7.548*** 7.429*** 0.869***
(0.0353) (0.0353) (0.0464) (0.0565)

Observations 4,743 4,743 4,640 5,585

R-squared 0.223 0.267 0.375
Standard errors in
parentheses
*** p<0.01, ** p<0.05,
* p<0.1

60
 Statement of Originality

The thesis is independently written by the author under the direction of the advisor. In
addition to what is specially labeled and contained in the acknowledgement, the thesis does not
include anything published or written by other persons or organizations. Inspirations and
contributions by other people have been clearly stated and appreciated in the thesis.

Author’s Signature Advisor’s Signature: Date: March 24, 2018

Ha Truong Li Huailu

Authorization Statement for Thesis Use

The author fully understands the provisions of Fudan University on keeping and using degree
theses, namely: the school reserves the right to retain a copy of the thesis and allow it to be
searched and read by others. The school may publish all or part of the thesis and retain it by
photocopying, micro printing or other reproduction means. Confidential theses Comply with these
provisions after the decryption.

Author’s Signature Advisor’s Signature: Date: March 24, 2018

Ha Truong Li Huailu

61
 论文独创性声明

本论文是我个人在导师指导下进行的研究工作及取得的研究成果。论文中除

了特别加以标注和致谢的地方外，不包含其他人或其它机构已经发表或撰写过的

研究成果。其他同志对本研究的启发和所做的贡献均已在论文中作了明确的声明

并表示了谢意。

作者签名： Ha Truong 导师签名：李怀露 日期： 2018 年 3 月 26 日

论文使用授权声明

本人完全了解复旦大学有关保留、使用学位论文的规定，即：学校有权保留

送交论文的复印件，允许论文被查阅和借阅；学校可以公布论文的全部或部分内

容，可以采用影印、缩印或其它复制手段保存论文。保密的论文在解密后遵守此

规定。

作者签名： Ha Truong 导师签名：李怀露 日期： 2018 年 3 月 26 日

62