Returns to education a case study in the mekong delta vietnam

In addition, RORE of workers working in public sector is higher compared to private sector at all below education levels; however, RORE of individuals with university and working in priv

HO CHI MINH CITY, DECEMBER 2009

HO CHI MINH CITY

INSTITUTE OF SOCIAL

STUDY THE HAGUE

THE VIETNAM-NETHERLANDS PROJECT FOR MA PROGRAM IN

DEVELOPMENT ECONOMICS

By TRAN NAM QUOC

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER 0F ARTS IN DEVELOPMENT ECONOMICS

SUPERVISOR: Prof NGUYEN TRONG HOAI

RETURNS TO EDUCATION: A CASE STUDY IN THE

MEKONG DELTA- VIETNAM

By

TRAN NAM QUOC

SUPERVISOR: Prof NGUYEN TRONG HOAI

HO CHI MINH CITY, DECEMBER 2009

Ho Chi Minh City, December 2009

TRAN NAM QUOC

I would like to thank Prof Karel Jansen, for worthy comments on Thesis Research Design (TRD) formation of this thesis

I would like to express sincere thanks to Dr Nguyen Huu Dung who has provided valuable comments for completion of this thesis

I would like to thank sincerely Mr Truong Thanh Vu, the visiting lecturer of VNP, for valuable comments and instructions in filtering the appropriate data set

I am indebted to all teachers and staffs of the Vietnam - Netherlands Program at University of Economics HCM

Finally, I am indebted to my family and others who gtve me great encouragement and support for my study

ABSTRACT

The purpose of this research is to examine the return to education in Mekong Delta The author has applied descriptive statistics, econometric model to examine return to education classified by return to schooling (RTS) and rate of return to education (RORE) within each different level of education The return to education

is also investigated separately for each of controlling variables in sub-groups The data sources are selected from VHLSS 2004 and VHLSS 2006 ofGSO

This study found out meaningful findings Firstly, on average, RTS is increased overtime Thus, it increased from 3.46 percent in 2004 to 4.13 percent in

2006 RTS is higher for female than for male and higher for persons living in urban than those living in rural, at 7.45 percent and 3.26 percent, respectively It is also higher for individuals working in service sector and industry-construction compared

to whose working in agriculture-forestry sector Secondly, RORE at high - school level is highest, at 11.49 percent" in 2004 and 11.89 percent in 2006 In addition, RORE of workers working in public sector is higher compared to private sector at all below education levels; however, RORE of individuals with university and working in private sector is 14.19 percent, higher than those in public sector (8.29 percent) Finally, in particular, RORE of workers with university and working in industry- construction sector obtained highest rate, at 24.81 percent in 2006

This study also argued mainly significant policy implications Firstly, the policies attracting high educated labor force to work in rural areas are worth to be considered Secondly, the process of rural urbanization should be strengthened more quickly to develop private sector and create more employment in rural areas Thirdly, high school, college and university education should be prioritized and more improved to increase more high-skilled labor force serving for economic development strategy in this region Favorable conditions and subsidies to make more opportunities for women increasing enrollment should be more considered Finally, advocacy programs should be strengthened for all organizations, families and individuals to change their perceptions and share roles and responsibilities for education and human development

CONTENTS

Certification i

Acknowledgements ii

Abstract iii

Content table iv

List of Tables viii

List of Figures ix

List of Abbreviations x

CHAPTER 1: INTRODUCTION 1.1 Introduction !

1.2 Research objectives 3

1.3 Research hypotheses 4

1.4 Research methodology 4

1.5 Data 5

1.6 Thesis structure 5

CHAPTER 2: LITERATURE REVIEW 6

2.1 Theoretical background 6

2.1.1 Economics of education 6

2.1.2 Human capital theory 7

-2.1.3 Education and economic development 10

2.2 Theoretical methods and conceptual models 12

2.2.1 Theoretical methods 12

2.2.1.1 Cost Benefit Analysis methods 12

2.2.1.2 The Short-cut method 13

2.2.1.3 The Reverse Cost- Benefit method 13

2.2.1.4 The Earning function methods 14

2.2.2 Conceptual models 14

2.2.2.1 Utility function 14

2.2.2.2 Mincerian model ; · 15

2.2.2.3 Other conceptual models 16

2.3 Empirical approaches with determinants of rate of return to education 17

2.3 1 Empirical models 18

2.3 2 Empirical evidences 21

2.3.2.1 Schooling year and education level 21

2.3.2.2 Experience 24

2.3.2.3 Gender 24

2.3 2.4 Household characteristics 27

2.3.2.5 Private and public sector 27

2.3 2.6 Rural and urban area 28

2.3.2.7 Agricultural and Non- Agricultural sector 29

2.3.2.8 Ethnic group 29

2.4 Chapter summary : 30

CHAPTER 3: RESEARCH METHODOLOGY 31

3 1 Key concepts 31

3 2 Model specification 31

3.3 Identification bias ability of missing an innate variable 35

3.4 Data set and variable measurement 36

3.5 Steps to estimate parameters in regression models 38

3.6 Chapter summary 39

CHAPTER 4: RESEARCH CONTEXT: THE MEKONG DELTA OVERVIEW 40

4.1 Geographic 40

4.2 Demographics 41

4.3 Education and Earning in Mekong Delta 41

4.3.1 Enrolment in general education 42

4.3.2 The important role of primary 42

4.3.3 Vocational and tertiary education 43

4.4 Regional and Sector 44

4.5 Human resource development of the region 45

4.6 Distribution of education attainment as a policy tools 46

CHAPTER 5: EDUCATION AND RETURNS TO EDUCATION IN THE MEKONG DELTA 47

5.1 Education - human capital and earnings of workers in the Mekong Delta 4 7 5 1.1 Education and human capital 4 7 5 1.1.1 Educational level structure 4 7 5 1.1.2 Labor structure classified by sectors 48

5 1.1.3 Schooling year classified by gender and region characteristics 49

5 1.1.4 Schooling year classified by sectors 50

5 1.2 Earnings of workers 51

5.1.2.1 Per capita income classified by gender and regions 51

5.1.2.2 Per capital income classified by sectors 52

5.1.2.3 Per capita income classified by education levels 53

5.2 Empirical findings 54

5.2.1 Examine the effect of parental education on RTS of workers 54

5.2.2 Estimation ofRTS in 2004 and 2006, Modell 55

5.2.3 Estimation ofRTS in 2004 and 2006 associated with Sub-group variables, Modell 56

5.2.4 Estimation ofRORE within each education level in 2004 and 2006, Model2 57

5.2.5 Calculation RORE within each education level for sub- group variables in 2006, Model 2 59

5.3 Charter remarks 61

CHAPTER 6: CONCLUSION AND POLICY IMPLICATION 63

6.1 Conclusion ~ 63

6.2 Policy implications 65

6.2.1 Economic policy 65

6.2.2 Education policy 66

LIMITATION OF THE RESEARCH 67

SUGGESSION FOR FURTHER RESEARCH 67

REFERENCES

APPENDICES

LIST OF TABLES

Table 2.1: Return to education in the world ; 22

Table 3.1: Definitions and notations of model variables 33

Table 4.1: Gini coefficients in the Mekong Delta···'··· 46

Table 5.1: Mean of schooling year classified by gender and region characteristics 49

Table 5.2: Mean of schooling year classified by sectors 50

Table 5.3: RTS for workers having parental education (2006) 54

Table 5.4: Estimation results of Modell ~ 55

Table 5.5: RTS associated with Sub-group variables, Modell 56

Table 5.6: Estimation results ofModel2 58

Table 5.7: RORE for each education level 58

Table 5.8: Estimation results for sub- group variables, Model2 59

Table 5.9: RORE within each education level 60

LIST OF FIGURES

Figure 2.1: Human capital: conceptual schema of an individual 9

Figure 2.2: Intellectual capital: conceptual schema of a company 10

Figure 2.3: Research schema of earnings functions and growth models from the human capital point of view 11

Figure 2.4: Stylized Age-Earnings Profile 13

Figure 4.1: The geographic of the Mekong Delta 40

Figure 5.1: The structure of education level in the MKD in 2004 and 2006 48

Figure 5.2: The labor structure classified by sectors 48

Figure 5.3: Means of per capita income classified by gender and regions 51

Figure 5.4: Means of per capital income classified by sectors 52

Figure 5.5: Means of per capita income classified by education levels 53

:Instrument Variable : On-the-Job Training : Ordinary Least Square : Participatory Poverty Assessment : Rate of return to Education

: Return to Education :Return to Schooling :United Nations Development Program :VietNam Household Living Standards Survey : Vietnam Dong

: Two- Stage Least Square

CHAPTER 1: INTRODUCTION

1.1 Introduction

Nowadays, the developing of knowledge - economic as well as emerging technologically - centered economic system calls for the acquisition and mastery of new knowledge, or for the upgrading of skills Therefore, the need for human capital development and accumulation is a priority condition for modem economic growth Human capital, which is normally known as the skills and knowledge intensity of the labor force in an economy, is essentially acquired through schooling and training According to Olaniyan and Okemakinde (2008), the improvement in education has been a significant explanatory variable for East Asian economic growth In fact, East Asian countries are not abundant and rich in physical capital as well as natural resources However, they have gained great achievement in economic growth due to appropriate policies, particularly, the rational investment strategy for education to develop human capital

On the national view, that education increases productivity and income of workers is a good argument to persuade the governments investing more in education Moreover, education investment by public and private are driven by the rate of return considered as the benefit of education in the labor market The nature

of labor market may be understood by investigating return to education in association with gender, by region at a time, and its changes overtime This will help to develop the strategies and policies for investment education more appropriately (Duraisamy, 2000)

It is becoming clearly to policy makers that education is a key factor in the poverty reduction Schooling is integral to this approach as it enhances the adaptability and efficiency of workers The return to education plays a crucial role

in determination of educational attainment, participation and ultigmtely on earning received by workers in the labor market It also influences the regionally human resource allocation (Sackey, 2008; Psacharopolous and Patrinos, 2002) Therefore, studying on returns to education is crucial to guide macro-policy decisions relating

to the organization and financing of education reforms

For the case of Vietnam, the Constitution of the Socialist Republic of Vietnam 1992, the Education Law 1998, the Political Report at IXth Congress of the Vietnam Communist Party 2001 and the socio-economic strategic plan for 2001 -

20 10 have identified guiding view-points for the educational development that

"Education is the foremost national policy" Vietnamese Communist Party stated that "In 21st century, the revolution of science and technology will continue to be developed, changing the society from era of industry to the era of information and knowledge based economy, at the same time affecting all fields of human activities, changing rapidly the physical and spirituals life of society The distance between scientific-technological inventions and their applications in practice is gradually reduced The knowledge treasure of mankind is becoming diversified, abundant and increasing by geometric progression"

Furthermore, rapid progress of science, technology as well as dynamic development of economies in quickly integration and globalization have been creating opportunities that may help to reduce the gaps among countries Science and technology are becoming the main driving force for socio-economic development In this context, education is the foundation of science-technology development and human capital development to satisfy the requirements of a modem society It plays the key role in enhancing national awareness for the capability and responsibilities of the present as well as future generations Then, investment in education is considered as investment for development rather than for social welfare as before (the National Assembly of Vietnam 1992, 1998; the Vietnamese Communist Party 2001; and the MOET 2001 )

Moreover, education is also one of the important driving forces to accelerate the industrialization and modernization process Thus, the key for long-term sustainable growth in the new emerging economy is the contribution of knowledge

to output growth Therefore, firstly, the creation of indigenous knowledge through human capital development is the key component to maintain competitiveness Secondly, participation in WTO requires that Vietnam has to enhance education to develop human capital serving for global integration process Finally, education

development is a key factor for long-term strategy of economic growth and social equality

In addition, according to PPA in Mekong Delta 2003 (UNDP and AUSAID,

2003 ), Mekong Delta is among the poorest regions in Vietnam Enrolment rates (VHLSS 2001- 02) indicated that the Mekong Delta had continued to have lower enrollment rates in comparison to the nation as a whole The enrolment rates· in primary, lower secondary and upper secondary are 88.8, 59.5 and 29.3 percent compared to 93.1, 71.2 and 42.1 percent as a whole nation respectively Furthermore, it accounts for the highest ratio of poor people in Vietnam and its majority of population are living in rural areas and working in agriculture

After The Doimoi since 1986, major economic reforms in Vietnam, as well

as in the Mekong Delta, have posed positive impacts on the labor market, education and the returns to education as well Because of lower education level and higher drop-out of school, the quality of human capital resource in this region is still weak and inadequate capacity for application and taking advantages of technological progressiveness It is clearly that human capital development is a key driver for economic growth, and has become even more crucial and important in this region However, the evidence for this issue is relatively limited; and in recently, study on return to education in Mekong Delta has not yet been done Therefore, research on education to enhance capacity of human capital is necessary and significant for the Mekong Delta Hence, the Mekong Delta area is chosen for study in this research It

is expected to examine the overall context and effect of education, and draw some sound implications for policy makers This research concentrates on the topic

1.2 Research objectives

The aim of the paper Is to examine the situation of human capital and education development to have meaningful policy implications in the Mekong Delta In this research, the basic premise is that education enhances the productivity

of a worker, which is, in tum reflected in his earnings Mincer - type earning function is adopted to estimate the rates of return to education The specific

objectives of the study are tailored towards addressing the key research issues raised above The specific objectives of this research are to:

(1) Investigate the difference in returns and rates of return to education among different levels of education, with take into account of effects of gender, sectors, and regions

(2) Examine the effects of parental education on return to education

1.3 Research hypotheses:

The following hypotheses (H) are of interest in this research:

H 1 : There is a positive impact of schooling years on yearly income per capita in the Mekong Delta

H2: Return to education is influenced by parental education

H3: Return to education has changed overtime

H4: Return to education of female is higher than that of male

H5: Return to education of workers in urban is higher than that in rural areas

H6: Return to education for workers in public sector is higher than that of those working in private sector

H7: Return to education for individuals working non-agricultural sector is higher than that in agricultural sector

1.4 Research Methodology

To fulfill answer the above objectives and research questions, descriptive statistics and quantitative methods are applied in the study Descriptive statistics will accommodate readers with a general view about educational and human resource characteristics in the Mekong Delta The econometric analysis technique, particularly Multiple Regression Model estimated by OLS and 2SLS is employed to estimate the parameters and returns to schooling The theory applied in this research

is based on Human capital theory and the earning function of Mincer, namely, extended Mincerian model

1.5 Data

The main data for analyses is drawn from VHLSS 2004, and VHLSS 2006 conducted by GSO supported by UNDP These data sets are used for many research goals Details of the surveys and data information will be described in the methodology chapter

1.6 Thesis Structure

The structure of the thesis includes 6 chapters organized as follow: the first is

an introduction chapter, presenting research problem, the need for doing an investigation of returns to education, research questions, objectives and hypotheses The second chapter is for reviewing of literature on theories of human capital and return to education, theoretical background of return to education studies and empirical evidences The third chapter describes research methodology In chapter

3, the variables, models and data will be specified in detail Chapter 4 is to provide

an overview of the research context, and synthesis of main information relating to human capital and education in the Mekong Delta Chapter 5 presents analytical results on returns and rates of return to education Finally, Chapter 6 is for conclusion, policy implications, limitations of the research, and suggestions for future research

CHAPTER 2: LITERATURE REVIEW

This chapter will be presented 3 parts The first part concentrates on general theories including economic theories of human capital and education, and aspects

on return to education The second part investigates theoretical background and empirical models relating to the return and rates of return to education The final one presents empirical evidences on the studies of return to education

2.1 Theoretical background

2.1.1 Economics of education

Economics of education is a significant issue documented in literature The crucial role of education for raising the productive capacity of society was raised by Adam Smith in The wealth of Nations in 1776 It has been developed continuously

by famous economists like Gary Becker, Jacob Mincer and Theodore Schultz who conducted major, innovated researches applying economics to education issues

Education can be defined as all studying forms and specific means In the other words, education is referred to activities trained in the place that is named school Education is the most important institution to develop human capital in association to many of the following aspects: "(i) people are demanding education highly and believe that education benefits them and their children; (ii) education and earning have a relationship individually as well as socially; (iii) budget for education accounts for a high ratio in family and national budget" (Torado, 1994) Torado also said that human capital investment is useful and conducted through investment in formal education, OJT, and health It will increase quality of labor force as enhancing in skills, capacity, ideas and health Furthermore, it is profitable

if discounted profits are higher than discounted expenditures

Education makes people knowledge abundant, creates comfortable life, and generates benefit for individuals and society as a whole Education also increases working productivity, management capacity as well as scientific and cultural progressiveness The person with higher level of education would gain a higher-starting salary and the steeper the rise in earnings during the early working life

Return to education is crucial information to address the relationship between labor market earning and education They are significant not only for the government and policy makers but also for all people investing in education (World Bank, 1991,

Mock et al., 1998, and Arcandy et al., 2004)

In addition, education was regarded as both consumer and capital good because it offers utilities to consumers and also serves as an input into the production of other goods and services In term of capital good, education can be used to develop the human resources for socio - economic transformation by creating qualified citizens and upgrading the general standard of living in a society Therefore, positive social change is likely to be · associated with appropriate education (Olaniyan and Okemakinde, 2008) Furthermore, to clarify the relationship between human capital and education, on human capital aspect, Machin (2007) indicated that education should be seen primary as an investment good in stead of consumption good Individuals invest in human capital, such as schooling, because human capital makes a person more productive reflected in higher wages in the future Therefore, individuals primarily make investments in schooling and other forms of human capital to earn returns, i.e to increase their income in the future

In term of study approach, according to Fleischhauer (2007), economics of education is divided into two different branches for studying the returns to education The micro labor approach investigates returns to education by measuring the extra earning of a worker for an additional year of schooling and training Meanwhile, the macro approach investigates whether the level of education in a cross-section of countries related to the GDP growth rate In this study, the aim is to examine private rate of return to education, so only the micro literature will be focused

2.1.2 Human capital theory

In the 1950s and 1960s, developed by American economists such as Gary Becker and Theodore Schultz, human capital refers to knowledge and acquires skills of a person accumulated It helps to increase his or her ability to conduct activities with economic value Human capital theory has also emphasized that

individuals invest in human capital as investors invest in education to earn higher income in the future

The fundamental of the human capital theory is the analogy between physical and human capital that means individuals can invest in human capital by education and training with an expectation of higher returns in the future It emphasizes how education increases the productivity and efficiency of workers by increasing the level of cognitive stock of human capacity This is a product of integration between innate abilities and investment in human beings All of these things lead to belief that education is a driven of change in many developing countries in economic development process (Olaniyan and Okemakinde, 2008)

Human capital is a factor of production just like labor and physical capital

In general, it is defined as the stock of knowledge, skills, attitudes, aptitudes, and other abilities contributing to productivity and work performance and a better life There are two main components of human capital which are strongly complementary: early ability (whether acquired or· innate) and skills acquired through formal education or training on the job It can be accumulated by different kinds of investment such as formal schooling (i.e the individual devotes his whole time to learning), on-the-job training (i.e post school training provided by the current employer), and off-the job training (i.e post school training provided by for-profit proprietary institution (Becker, 1975; Hietala, 2005 and Fleischhauer, 2007)

Therefore, human capital can be a-target for individuals' well-being and a mean to achieve fundamental and long-term goals in work or in the lifetime-career,

it is also a mean of achieving higher productivity thereby higher wages is gained These arguments may be summarized as follow:

Figure 2.1: Human capital: conceptual schema of an individual

Flow of human capital

- Learning (from education and training

The model of human capital was developed firstly by Becker (1975) as a model of individual investment, in which human capital has been considered similar

to physical means of production and all activities influencing real income in the future In addition, according to Olaniyan and Okemakinde (2008), "The basic implication of the human capital model is that allocation of resources on education should be expanded to the point where the present value of the flow of returns to marginal investment is equal or higher than the marginal costs"1• Many ofthe developing countries have realized that the fundamental mechanism for developing human knowledge is through the education system By this way, education has been invested considerably not only as an attempt on enhancing knowledge and skills to individuals but also imparting values, ideas, attitudes, aspirations as well as

1

Olaniyan and Okemakinde, 2008

enhancing the brand name of companies, communities, regions, organizations, or nations In term of household view, parents strongly believe that in context of scarce skilled human capital, the better the education their children can get, the better opportunities their children can approach to get well-paid jobs as a result of the poor often look at their children's education as the best mean of escaping poverty (Olaniyan and Okemakinde, 2008)

2.1.3 Education and economic development

According to Hietala (2005), there are several ways to examine how to the educational expansion accelerate economic growth and development The first is to view education as an investment in human capital The next is to view of the role of education in creating positive externalities Another way is considered education as

a critical input for innovations, research and development activities (R&D) This view shows a close correlation between new product development and educational levels, which may be figured as below:

Flow of intellectual capital

Automatic increase of intellectual capital capital

more schooling is associated with higher individual's earnings Schooling years play a crucial role in economic development because education can enhance the human capital in the labor force; thereby leading to an increase in labor productivity and higher equilibrium level of output The important role of education as well as the relationship among education, human capital, and economic growth may be schemed as following:

Figure 2.3: Research schema of earnings functions and growth models from the

human capital point of view

Education and Training

2.2 Theoretical methods and conceptual models

2.2.1 Theoretical methods

Relating to investigation return to education, the neoclassical assumptions of profit and utility maximizing rational agents and perfect competition in the labor market has being made in the standard human capital theory The higher productivity level obtained from the education investments is reflected in higher earning However, training is also associated with costs covering both direct costs, e.g tuition fee and indirect costs, e.g foregone earnings Therefore, the decision to invest in training is dependent on a higher present value of the future productivity increase compared to the present value of all the investment costs as a result the question is raised that 'what is the profitability of this investment in order to compare it to alternatives?' (Sorensen,2000 and Pshacharopoulos,1995) For investigation these views in detail, Pshacharopoulos clarified the fundamental methods to estimate return to education more crucial as below

2.2.1~1 Cost Benefit Analysis methods

According to Psacharopoulos (1995) and Patrinos et al (2004), "The private return to an investment at a given level of education can be estimated by finding the rate of discount (r) At this rate the stream of discounted benefits is equalizes to the stream of costs at a given point in time In which, the costs are foregone earnings while studying, plus any education fees or incidental expenses the individual incurs during schooling The private benefits amount to what a more educated individual earns above a control group of individuals with less education"

For example, at university level, the formula to estimate rate of return is:

Figure 2.4: Stylized Age-Earnings Profile

Source: Psacharopoulos (1995) and Patrinos et al (2004)

2.2.1.2 The Short-cut method

Given the shape of the age-earnings profiles, an individual can approximate them as flat curves and the RORE can be estimated by following simple formula:

where W refers to the mean earnmgs of an individual with the subscripted educational level, and 5 is the length of the university cycle The weakness of this method is that, in reality, age-earnings profiles are concaved, not flatted and the discounting process (in estimating the true rate of return) is very sensitive to the values of the early working ages (Psacharopoulos, 1995)

2.2.1.3 The Reverse Cost-Benefit method

This method is based on the short-cut rate of return formula The question needed to solve that given the cost of the investment, what level of annual benefits would produce a given rate of return (I 0 percent, for example) on the investment?

[2.3]

This calculation can be made easily and can further analyses on how to reduce the costs or increase the benefits to possibly justify the investment (Psacharopoulos, 1995)

2.2.1.4 The earning function method

This method is known as the Mincerian method and involves the fitting function of log wages (Ln W), using years of schooling (S), years of labor market experience and its square as independent variables (Mincer 1974, 1989) In this

semi- log specification the coefficient on years of schooling ( f3) can be interpreted

as the average private rate of return to one additional year of schooling, regardless

of the educational level this year of schooling refers to

In fact, f3 coefficient in the above semi - log basic earnmgs function

corresponds to the rate of return estimated by the short - cut method This can be seen in the following approximation,

fJ = o In W = relative earnings differential = [ Ws - W 0 ] _1_ = Ws - W 0 = r [2.4]

oS education differential W 0 118 118.W 0

respectively, and 118 is the difference in years of educational attainment between

the two groups

2.2.2 Conceptual models

2.2.2.1 Utility function

The human capital model of household or individual decision-making has its

individual i would receive if he or she acquires schooling level Si Assuming that the individual's utility function U(Sb Yi) is a function of level of schooling Si and average earnings, Yb and individuals maximize their utility functions by choosing their level of schooling si so the utility function takes a simple form:

[2.5]

where g(Si) is an increasing convex function representing the costs from schooling

In the case of assuming only the private benefits be considered, individuals could

earn nothing while in school and y afterwards and individuals discount their stream

of future earnings at rate, r

2.2.2.2 Mincerian model

Based on propose of Mincer (1974), Becker (1975), develop the empirical model to investigate the correlation between education and wage The early Beeker-Mincer model could be presented as follow:

other direct costs of schooling, earnings while in school, and public subsidization to schooling In this framework, the relationship between wages and schooling is understand as · a compensating wage differential emerging from individual differences in discount rates

In other words, Mincerian model could be derived in the simpler way Assuming that the individual invests an amount of time in education and then the return shows up in terms of enhanced future earnings If opportunity costs are the only important schooling costs, then the rate of return on first year of schooling is the incremental benefits relative to incremental costs

[2.7]

After two years of schooling

After s years of schooling

And if the rate of return is the same on all years of schooling

Appending a disturbance term

[2.13]

2.2.2.3 Other conceptual models

Belzil (2006) assumed that individuals are paid their marginal product according to their individual specific level of skills as following model:

[2.14]

where Wt is labor market wage, Pt is the rental price of skill and Kt represents the human capital This model reveals the inherent identification problem that both human capital (K) and skill prices (P) are basically unobservable This problem may

be solved by Mincerian approach because of assumption that human capital accumulation (or skill acquisition) is rendered possible by combining inputs such a time spent in school, time spent in the labor market, and innate ability

According to Heckman et al (2003, 2008), the self - selection model

literature to investigate the decision to attend college is started with a model of life

time earnings (Yu) basing on that individual is endowed with the following wage equations:

[2.16]

[2.17]

where i is the individual subscript, j is the education level achieved by the individual (either high school or college), Wi,hs is the wages if the individual stops after high-school graduation, Wi,c is the potential wage as a college graduate, Ei,hs is represent unobserved ability in job requiring high-school training and Xi is a vector

of individual characteristics So the optimal decision is represented by the following latent structural equation:

[2.18]

where Si = 1 when Si* > 0 and 0 if not In most related applications, it is assumed that Ei hs' Ei c and Ei s follow a multivariate normal distribution and the vector Zi , , , contains all elements of Xi plus other regressors that would affect the discount rate

The self-selection model is appropriate for examining the schooling decision process With [2.16] and [2.17] equations, it could evaluate the return to schooling more than one parameter In other words, the return to schooling in the between [2.16] and [2.17] are different Vice versa, out-put of [2.16] is a single parameter However, in recently, most studies used Mincerian model to estimate returns to education Thus, in this study, Mincerian approach is followed that helps our findings becomes comparable with previous studies

In conclusion, basing on above arguments, both aspects of calculation method and econometric models all support for application the extended Mincerian model and earning function method for estimation and calculation the R TS and RORE

2.3 Empirical studies on the determinants of return to education

What economists debated when discussion the role of education associated with human capital and economic growth? While Becker (1975) suggested that education or training raise the productivity of workers by imparting useful

knowledge and skills, there are different explanations for how education is related

to worker productivity For example, Spence (1973) and Schultz (1975) argued that education is used as a market signal to indicate the potential productivity of workers and education enhances an individual's ability to successfully deal with disequilibria in changing economic conditions; Thurow (1975) maintained that productivity is affected by largely characteristic of jobs rather than of workers, employers use education credentials to select workers Recently, Hall and Jones (1999) indicated that social infrastructure including the institutions and government policies affect to productivity by determining the economic environment, in which individuals accumulate skills and firms accumulate capital and produce output Basing on above arguments, productivity can be affected by ability of workers and other conditions such as social infrastructure, institution, organization of production, etc However, this study is just focus on micro aspects relating to earnings corresponding with individual decision in education investment

In term of micro aspect, there are two methods to estimate the rate of return

m education which are the earning flow comparison method and the earning function method The second is that regress the function including dependent variable - logarithm of income and the independent variables which are the number

of years in school, working experience, working experience square, and other control variables It can be extended by assumption that the rate of return may change through educational levels (Chiswick, 2003) This method is used widely through the Mincerian model in a lot of researches on return to education

2.3.1 Empirical Models

As mentioned above, Mincerian model is a benchmark model to study return

to education Recently, most of studies on rates of return to education have applied Mincerian model Starting with simple form ofMincerian model (Mincer, 1974)

[2.19]

where In Yi is the natural logarithm of annually earnings, Si denotes each individual's schooling year, Expi is working experience, Expi2 is square of working experience, and Ei is error term

Brunello, Comi and Lucifora (2000) studied the returns to education in Italy based on a simple human capital, where optimal level of schooling is given by equating the marginal return to the marginal cost of education The estimates based

on instrumental variables model:

[2.20]

where ln(Yi) is net hourly wages of individual i, Si is years of schooling (in years)

of individual i, Expi is the years of experience, Exp? is the square of experience, D

is the vector of variables including years of schooling, that are interacted with working experience This study extends Mincerian model by investigation the interaction between the schooling year and experience Its meaning is help to explain that skill of individuals by formal education can be supported and integrated

by OJT

Other studies investigated returns to education in term of highest education level obtained instead of schooling years Tsaklogkou and Cholezas (2000) investigated private returns to education in Greece Basing on standard Mincer equation, they applied the model as follow:

[2.21]

where Yi is net hourly earning, Lk is highest education level obtained by individual

i, Ex is potential experience, Part-time is dummy variable for female working less · than full-time, Li is dummy variables for different levels of education In this model, estimating rate of return in term of highest education level instead of schooling years, it helps to examine more clearly the rate of return for each education level that might reflect the quality of education more clearly

Campos and Jolliffe (2002) estimated returns to years of schooling in the Hungarian transition in 1986-1998 from centrally planned to market economy The model suggested as below:

where Xi contains a set of variables to control for institutional and demographic characteristics (rural or urban, what sector individual is working in, gender)

Control variables could help to explain that wage can be affected by others variables not only education and experience This approach will be considered to our suggested model

Yang (2005) studied returns to education in big cities in China and its changes from 1988-1995 He applied the extended Mincerian model written as:

3

j=l

earnings namely gender, communist party membership, and minority These variables have much contribution to examine more what other factors affect to the return to education Gender variable helps to examine the difference between male and female; ethnicity variable shows difference in term of race Both of variables may be significant in policy issues in Mekong delta

Similar to the above studies, Oyelere (2005) studied returns to education in Nigeria, the benchmark Mincerian earning function was used to estimate the returns

to education by OLS method, as follow:

[2.24] where S is the numbers of years of schooling of individual i, Exp is years of

possible exogenous/control variables including dummies for regions, gender, cohort dummies and so on

In Vietnam, Liu (2005) developed the model including a vector of explanatory variables and vector of dummy variables He mentioned the relationship between earning and education He estimated separately for government sector, SOEs, and the private sector, and for males and females for each survey period In this study, Mincerian earnings equation was specified as:

dummy variables, (e.g migrant and marital status, seven regional dummies, an urban-rural dummy, a major-minority dummy, and a set of occupational dummies such as professionals, clerical and trade-related jobs, laborers and agricultural occupations)

In another study relating to the returns to education in Vietnam, also basing

on Mincer's benchmark model, Moock, Patrinos and Venkataraman (1998), using data from VHLSS 1992 - 1993 examined the effects of market liberalization and education refor:n1s The Mincerian model was specified as follow:

1n Yi = ~o + ~~ Primi + ~2 Seci + ~3 Hischoi+ ~4Univi + ~5 Expi + ~6 Expi2

+ aiZi+Ei [2.26]

where In Yi is the natural logarithm of per capital income, S is the numbers of years

of schooling of individual i, Exp is the year of experience, Exp2 is the square of experience, Z is the vector of control variables, Prim, Sec, Hischo and Univ stand for primary, secondary, high school, and university level respectively These models of this research are relatively specific and may be appropriate to estimate return to education in Vietnam

2.3.2 Empirical evidences

In the following sections, the impact of key independent variables on the return to education will be investigated and analyzed basing on the general findings from these researches and typical others relating to return to education such as Palme and Wright (1998) in Sweden, Andrew (2000) for Russia, Aromolaran (2002) in ducation in Nigeria; Patrinos and Sakellariou (2002) in Venezuela;

Vernon (2002) in Russia; Psacharopoulos et a! (2002) in the global; Amin and

A wung (2005) in Cameroon; Barrow et a! (2006) in United States; Sackey (2008)

in Ghana; and Moock et a! (1998), Thanh (2006), Imai et a! (2007), and Van et

a! (2008) in Vietnam, etc

2.3.2.1 Schooling year and educational level

Palme and Wright (1998) studied the returns to education in Sweden in the period 1968 - 1991 The rate of return to education is measured in terms of the differences in wage rates associated with differences in educational levels Two

measures of education are the number of years of schooling completed (S) and the highest educational qualification obtained Year of schooling is intended to capture the 'quantity' dimension of education; the highest qualification obtained is intended

to capture the 'quality' dimension of education The schooling is considered as the 'input' to the educational process and educational level obtained as the 'output' Palme and Wright (1998) used three concepts of wage differential associated with education defined as follow: (i) the return to an additional year of formal schooling, (ii) the return associated with possessing a given level of education and (iii) the return to an additional year of schooling within a given level of education

In term of schooling year, Psacharopoulos (1994) summarized that returns to schooling in developed countries is 6.8 percent In Asia and developing in Latin American, return to schooling are 9.6 and 12.4 percent respectively (see Table 2.1) Table 2.1: Return to schooling in the world

* OECD is not included

Source: World Development (1994)

5.9 8.4

8.5

7.9 10.9

3.4 9.6 8.2 12.4 6.8

Aromolaran (2002) estimated rate of returns to education (RORE) associated with levels of educational attainment across gender and age categories in Nigeria (1996-1999) Thereby, RORE for both men and women with primary and secondary have been at low rate, about 2 to 4 percent versus higher RORE for people with post-secondary education level, about 10 to 15 percent In fact, returns to education

in primary, secondary and higher in higher education level as tertiary This may be explained by that in Nigeria, the cost of primary may be high included opportunity cost of spending for education as well as wage for people with primary level may low and the people graduated from university have high salary Therefore, the gap

indicates that the larger gap in education leads to the larger gap m earnmg, especially in developing countries

In addition, Campos and Jolliffe (2002, 2007) found that returns to education

in Hungary increased in during transition period from 1986 to 2004 according to educational levels Primary and vocational education show the smallest return, and general secondary education and university show the largest changes in returns from 1986- 1998 These results are not consistent to many studies that indicated return to education in primary was highest For example, Psacharopoulos and Patrinos (2002) pointed out that the return to education is highest in primary This may be due to difference in methodological approach because they usually stand on the views of World Bank that investing in education in primary is play crucial role in poverty reduction and make equality However, this is not always appropriate in other empirical studies

In other studies, basing on the survey data collected from five provinces of Cameroon in 1994, Amin and A wung (2005) pointed out that percentage change of earnings associated with change in educational levels, the higher educational level

an individual obtains the higher rates of return he/she gains Yang (2005) studies returns to education in big cities in China overtime (1988-1995) The results indicated that return has increased from 1988 The average return to education increased substantially over the seven-year period (1988-1995); it increased from a range of3.3 - 3.9 percent to 5.9- 7.3 percent

In Vietnam, Moock, Patrinos and Venkataraman (1998) pointed out that on average the rate of return are still relatively low Private RORE to primary and university education is averaged 13 and 15 percent, but only 4 and 5 percent at secondary and vocational levels which are low compared with that in other developing countries and average private return in worldwide, at 10 percent (Psacharopoulos 1994) They also indicated that, in general, returns to schooling are low in centrally planned and transition economies

In order to compare with previous empirical findings, in this study, return to education will be mentioned including return to schooling and rate of return to

education Therefore, schooling and educational levels will be examined in the model as independent variables

2.3.2.2 Experience

Investigation the relationship between education and earning in Ghana, Teal (200 1) pointed out that OJT as well as experience give effects on returns to education and experience strongly associates non-linear with rate of returns Tsaklogkou and Cholezas (200 1) found out the relationship between log-earning and potential experience is bell - shaped for both sexes In addition, for the cases of transitional countries, Vernon (2002) applied Mincerian model for dataset of Russia

in the period 1992-2000 as a result in the returns to experience increased over the years and it is higher for women than men However, Campos and Joliffe (2007) found out that returns to experience decline in the transitional period 1986-2004 in Hungary and found no significant evidences supporting for state that returns to education changes over time Yang (2005) studied returns to education in big cities

in China overtime He indicated that the returns to experience increased from 4.2 percent to 7.6 percent (in the period of 1988-1995)

In Vietnam, research findings of Moock et al's (1998) pointed out that return to experience is 6.4 percent, for male is 5.7 percent, and for female is 6.6 percent Return to schooling for workers in public sector is 4.6 percent vs 7.2 percent in private sector In addition, according to Liu (2005), work experience has been found be only significant for private sector employees Experience and its squared term indicated the usual inverted-U shaped relationship between wage rates and labor market experience A usual inverted-U shaped relationship has been observed between wage rates and labor market experience The return to experience has declined for all wage earners (although the decline is larger for males) indicating recent labor market experience is more valuable than that acquired under central planning

2.3.2.3 Gender

According to Desai (200 1 ), education is the most important dimension for gender equity Firstly, the efficiency achieving from increasing the level of

education of women is enormous because women's education creates positive effects on children's health and education and increasing women's schooling also benefits households through higher incomes from agricultural production and non-farm enterprises Secondly, focusing on women's education makes positive effect

on wage earning capacity of them because, in market economies, wages tend to be highly correlated with human capital The persons who have less schooling are more likely to have low-wage jobs and also be paid less in the same type of job Finally, increasing women's education in not only qualify them for higher wages in existing jobs but also permit them to be for a wider set of, potentially higher wage, employment options

In addition, for the case of Nigeria, Aromolaran (2002) pointed out that return to education for both men and women are small at primary and secondary level, about 2 to 4 percent versus higher rate at post-secondary education level, say,

10 to 15 percent In detail, female wage earners receive 18.5% percent less than males Wage rate varies by gender, education level within age groups, and age within education levels Private returns to schooling are 4.6 percent and 5.3 percent for male and female wage earners; meanwhile, for self-employed earners, these rates are 3.6 percent and 2.8 percent for male and female respectively Returns to primary level are lowest, at 2.5 percent for men, and 2.4 percept for women Returns to secondary are also small at 3.9 and 4.4 percent while returns to schooling

of post secondary education are 10.4 and 12.2 percent for men and women respectively Among the self-employed, returns to schooling at the primary school level are 3.2 percent for men and 1.9 percent for women; returns to schooling at secondary education are 3.7 percent for men and 3.8 percent for women; meanwhile returns to an additional year of post-secondary education are 13.7 percent for men and 15.4 percent for women Thus, in general, the return to education of women is considerably higher than that of men Similarly, Patrinos and Sakellariou (2002) found that return to education for men are quite lower than those for women, at 8.4 percent in 1992 and 9.9 percent in 2002 for men versus 11.9 percent and 12.7 percent for women in V ~nezuela Vernon (2002) also indicated the same finding in Russian in the period 1992-2000

Furthermore, Sackey (2008) examined the private returns to education in Ghana over a seven year period (1992-1999) He pointed out that the private returns

to schooling at higher levels of education have increased for both female and male workers For female, the return to an additional year of secondary schooling increased from 7.3 percent in 1992 to 12.3 percent and form 11.4 percent to 18.4

· percent for tertiary education level in 1999 For male, this return decreased from about 7.0 percent to 6.0 percent for secondary level while increased from 13.0 percent to 19.0 percent for tertiary education level In particularly, the difference in the returns tertiary level associated with gender has been existed overtime and this return for women is considerably lower than that for men) This reflects gender inequity issue Similarly, Yang (2005) studied returns to education in big cities in China overtime He found out that the gender earnings gap was nearly doubles in percentage terms, rising from around 9.7 percent to between 15.5 and 16.7 percent

in the period of 1988-1995

For the case of Vietnam, Moock et al's (1998) indicated that, in general, males have received a 3.0 percent increase in earnings for every year of schooling meanwhile females have received a 7.0 percent increase; returns to higher education for females gained 12 percent, higher than for males, at 1 0 percent In addition, According to Scervini (2005), gender inequality has involved two different fields The first is the difference in the wage earned by women and men with the same characteristics of education, age, productivity, experience The second one regards exactly the differences in those characteristics, whenever a woman can not obtain the same characteristics of men, or is not allowed to be employed in certain positions

Furthermore, Liu (2005) applied Mincerian earnings equation for examination the wage structure m Vietnam The finding showed that the government sector exhibits the least gender earnings disparity However, male employees in SOEs and the private sector were paid more than the female employees For example, male workers in the private sector received an hourly wage rate of about 26 percent more than their female counterparts

Results of these studies found that return to education for female is often higher than male Gender variable is very important to investigate gender equality issues in a developing country like Vietnam, especially in rural areas as Mekong Delta Therefore, it will be included into models to examine the impact of gender on the returns to education

2.3.2.4 Household characteristics

Innate ability is critical to returns to education Individual with higher innate ability is expected to get higher return to education However, this factor is unobservable; in other words, it is very hard to measure So, some authors used family factors as proxy variables, including family income, and parental education Typically, parental education is often used as proxy for innate ability that implies more educated parents are associated with a higher return to education for their children (Fleischhauer 2007)

2.3.2.5 Private and public sector

Tsaklogkou and Cholezas (2000) investigated private returns to education in Greece and found out insignificant difference between returns to education in public and private sector Meanwhile, other studies such as Brunnello (2000) for Italy and Andrew (2000) for Russia indicated that returns in public is higher than in private sector Teal (2001) for Ghana pointed out that persons who work in public sector have highest level in schooling and higher return compared to private sector In addition, according Kazianga (2004 ), returns to education for men at secondary are

10 percent in public sector and 14 percent in private sector in for Burkiaso In contract, Campos and Jolliffe (2002) found that returns to education higher for private sector workers in Hungary

Differences between public and private sector may be due to difference in wage policy among countries For example, in Russia, after transition to market economy, the enterprises in private sector have developed and applied dynamic wage mechanism while wage policy in public sector is not effective longer So that return to education for workers in private may be higher th~n those in public sector

In Vietnam, Moock et al's (1998) indicated that workers in the public sector gain higher private returns to schooling than private sector workers The return to education for public workers was 6.2 percent and for private sector was 3.9 percent

In addition, Thanh (2006) pointed out that workers in public sector (including the government and state-owned enterprises) earned 19.43 percent more than that of workers in the private sector

Liu (2005) found out that private sector workers, in average, were less educated and most of them are laborers The return to schooling was 7.5 percent for

an additional year of education in the government sector, 4.0 percent and 4.2 percent for SOEs and the private sector, respectively The returns to schooling were 3.3 percent for male and 5.5 percent for female working in private sector

These results may suggest that in developing countries, private sector should

be supported to develop because they can help increasing job opportunities and wage as a result of increase in returns to education This plays important role in economic development for reducing gap in income ·distribution and poverty Therefore, in our study, this sector variable will be included into the model to examine effects of these sectors on returns to education

2.3.2.6 Rural and urban area

Thanh (2006) applied Mincerian model to estimate rate of returns to education in Vietnam and investigate the relationship between schooling and earning He found out that a male worker earned 16.79 percent more than that of a female worker Workers in Hanoi and Ho Chi Minh City earned 17.34 percent and

69 percent respectively higher than workers working in the rest of regions This revealed the massive attractiveness ofHo Chi Minh City for job seekers

In addition, Van et al (2008) indicated that an individual in urban areas

earned and consumed twice as much as an individual in rural areas did In 2004, urban income was as much as 1.94 times higher than rural income; whereas, in

2002, the income and expenditure ratio was 1.93 and 1.98 percent respectively So regarding to income, urban households earn more than rural households <fo This