Internationnal remittances and the education of yuong generations the case of Vietnam

Employment status of young generations Households Incomes and Expenditures National Survey of Institute of Statistics, Geography and Informatics Enrollment status ofyoung generations For

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM- THE NETHERLANDS

PROJECT OF M.A ON DEVELOPMENT ECONOMICS

INTERNATIONAL REMITTANCES AND THE EDUCATION OF YOUNG GENERATIONS: THE

CASE OF VIETNAM

A thesis submitted in partial fulfillment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

ACKNOWLEDGEMENT

I would like to express my sincere gratitude to my supervisor Prof Peter Calkins for his kind support of my Master study and research from initial to final level, for his patience, enthusiasm, encouragement and immense knowledge His assistance has helped me a lot during the time of designing and writing this research

I am heartily thankful to Dr Nguyen Trang Hoai, Dr Nguyen Minh Due and Dr Tran Tien Khai for their comments and evaluation of my initial research proposal

I also send my gratefulness to Mr Truong Thanh Vu and Mr Luong Vinh Quae Duy who help me to deal with VHLSS 2006 dataset and Stata software

Special thanks go to my friends for their support and motivation during my study at the Vietnam -Netherlands Program for M.A in Development Economics

Last but not the least, I would like to thank my family: my parents for giving birth to me, my parents-in-law, my husband and my older brother for supporting me spiritually throughout my life

Finally, I offer my regards and blessings to all of those who supported me

in any respect during the completion of the thesis

ABSTRACT

Over the last decades, Vietnam has experienced a sharp increase in international remittances inflows from overseas migrants The total volume of remittances accounts for approximately 6-8% of GDP Many Vietnamese people have migrated not only with the hope of getting more income and sending more remittances to support their poor families in hometown, but also with the desire to pay school fee for their young generations

By lifting liquidity constraints, remittances raise the educational attainment

of young people left behind in receipt countries The impact of remittances on the education of young generations has been studied in many papers However, this relationship has not been tested empirically in Vietnam Therefore, the contribution of this paper is to investigate empirical evidences on the link between international remittances and probability of being enrolled in schools of young individuals in Vietnamese households through exploring data of the Vietnam Household Living Standard Survey 2006 by applying Logit econometric model

Our results show that remittance receipts statistically significantly increase the probability of school enrollment, particular for girls and in rural areas In addition, this research also finds that the young who have to work tend to gain lower chances to go to schools than their friends without any job Based on the findings, the author suggests many ways to improve the school enrollment rate of young individuals Methods aiming at stimulation of remittance income are appropriate such as encouragement of labor exportation programs since Viet Kieu has been getting older and less altruistic as well as lowering cross border money transfer fees Moreover, when reserved financial budget for young people increases via the effect of attracting more remittance receipts, number of school aged individuals participating in labor market at early ages tends to diminish further

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION 1

1.1 PROBLEM STATEMENT 1

1.2 RESEARCH OBJECTIVES 3

1.3 RESEARCH QUESTIONS 4

1.4 RESEARCH HYPOTHESES 4

1.5 METHODOLOGY 4

1.6 RESEARCH SCOPE 5

1.7 THESIS STRUCTURE 6

CHAPTER 2: LITERATURE REVIEW 7

2.1 INTRODUCTION 7

2.2 THE CONCEPT OF INTERNATIONAL MIGRANT REMITTANCES 7

2.3 THEORETICAL FRAMEWORK AND EMPIRICAL STUDIES RELATED TO THE IMPACT OF INTERNATIONAL REMITTANCES ON THE EDUCATION 8

2.3 1 Theoretical literature 8

2.3 2 Empirical literature 13

2.4 THE ANALYTICAL FRAMEWORK 18

2.4.1 Empirical model 18

2.4.2 Variables introduction 20

2.5 SUMMARY 21

CHAPTER 3: OVERVIEW OF INTERNATIONAL REMITTANCES AND EDUCATIONAL ATTAINMENT OF THE YOUNG IN VIETNAM 22

3 1 INTRODUCTION 22

3.2 OVERVIEW OF INTERNATIONAL REMITTANCES IN VIETNAM 22

3.3 OVERVIEW OF EDUCATION OF THE YOUNG IN VIETNAM 29

3.4 SUMMARY 32

CHAPTER 4: METHODOLOGY 33

4.1 INTRODUCTION 33

4.2 DATA DESCRIPTION 33

4.2.1 Sampling method and sample size 33

4.2.2 Description of variables 34

4.2.3 Descriptive statistics of variables 36

4.3 STRENGTH AND WEAKNESS OF DATASET 39

4.4 MODEL SPECIFICATION 40

4.5 ESTIMATION STRATEGY 43

CHAPTER 5: EMPIRICAL ANALYSIS 45

5.1 ESTIMATION RESULTS 45

5.2 INTERPRETATION OF THE RESULTS 45

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS 51

6.1 CONCLUSIONS 51

6.2 RECOMMENDATIONS 53

REFERENCES 55

APPENDICES 61

APPENDIX 1 61

APPENDIX 2 65

APPENDIX 3 70

LIST OF FIGURES

Figure 2.1 Home Investments in Young generations 9

Figure 3.1 Number of exported workers during 1992-2007 23

Figure 3.2 Percentage of Vietnamese migrant workers by countries 1992- 2006 24

Figure 3.3 Inward remittances from Vietnamese migrants during 2001-2009 25

Figure 3.4 Remittances as percentage of GDP during 1999-2008 26

Figure 3.5 Uses of international remittances(%) 28

Figure 4.1 Distribution of independent variable "remittances per capita" 65

Figure 4.2 Distribution of independent variable "logarithm of remittances per capita" 65

Figure 4.3 Distribution of independent variable "income excluding remittances per capita" 66

Figure 4.4 Distribution of independent variable "logarithm of income excluding remittances per capita" 66

Figure 4.5 Distribution of independent variable "age of young generations" 67

Figure 4.6 Distribution of independent variable "percentage of school-aged children in the household" 67

Figure 4.7 Distribution of independent variable "age of household head" 68

Figure 4.8 Scatter plot of predicted value and standardized Peason residuals 74

Figure 4.9 Scatter plot of predicted value and deviance residuals 74

Figure 4.10 Scatter plot of predicted value and leverage 75

Figure 5.1 Predicted probability of school enrollment by remittances per capita 46

Figure 5.2 Predicted probabilities of school enrollment by remittances per capita and urban-rural areas 48

Figure 5.3 Predicted probabilities of school enrollment by remittances per capita and gender 49

Figure 5.4 Predicted probabilities of school enrollment by remittances per capita and employment status of young generations 50

LIST OF TABLES

Table 3.1 Flow of international remittances by origin(%) 26

Table 3.2 Percentage of households receiving international remittances 27

Table 3.3 Share of total remittances in urban and rural area 27

Table 3.4 Net enrollment rate in year 2006 by urban rural areas and sex 29

Table 3.5 Average expense on education and training per person in the past 12 months in year 2006 by expense item, urban rural, sex, age group and type of school 31

Table 4.1 Investigated objects 34

Table 4.2 Descriptive statistics of independent variables 37

Table 4.3 Descriptive statistics of dependent variable "Enrollment status" 68

Table 4.4 School enrollment status of young generations by amount of international remittances 38

Table 4.5 School enrollment status of young generations by age group 38

Table 4.6 School enrollment status of young generations by employment 39

Table 4 7 Description and measurement of variables 40

Table 4.8 Logistic regression results of Model I (full) 70

Table 4.9 Logistic regression results of Model 2 (restricted) 70

Table 4.10 Diagnostic test to compare Model 1 and Model 2 70

Table 4.11 Hosmer and Lemeshow's goodness-of-fit test 70

Table 4.12 Diagnostic test for model specification error 71

Table 4.13 Diagnostic test for Multicollinearity 72

Table 4.14 Logistic regression results of Model 3 (after excluding influential observations) 76

Table 5.1 Estimation results of Logit Model 45

Table 5.2 Marginal effects of explanatory variables on the probabilities of enrollment 76

Employment status of young generations Households Incomes and Expenditures National Survey of Institute of Statistics, Geography and Informatics

Enrollment status ofyoung generations Foreign Direct Investment

Gross Domestic Products General Statistics Office Human Development Index Age of household head Employment status of household head Gender of household head

Number of completed school years of household head International Monetary Fund

Income excluding remittances per capita Latin American Migration Project-Dominican Republic Survey

South Africa Labor Force Survey Logarithm of income excluding remittances per capita Logarithm of remittances per capita

ML Maximum Likelihood

PER SA CHILD Percentage of school-aged children in household

PSLSD Project for Statistics on Living Standards and Development

CHAPTER 1: INTRODUCTION

This chapter will present the thesis topic that this study focuses on The problem statement will be discussed in section 1.1 Additional contents including research objectives, questions, hypotheses and methodology are presented further

in section 1.2, 1.3, 1.4 and 1.5 respectively Finally, research scope is mentioned

in section 1.6 and thesis structure will concludes this chapter

International migrant remittances play a crucial role in developing countries The huge and growing volume of international migrant remittances is even larger than the ODA amount flowing into developing countries Most of the remittance inflows come from overseas migrants with the purpose of family support and investment in business or human capital Remittances may g1ve significant impacts on households especially low-income families Remittances may help receipt households to establish or expand their small business (Woodruff

& Zenteno, 2007; Amuedo-Dorantes & Pozo, 2006) and to pay for medical expenses (Amuedo-Dorantes et al., 2007) In addition, remittance receipts are used

as an investment in the education of young generations in remittance receiving families

Vietnam was one of the top recipients of overseas migrant remittances among developing countries in 2008 (WB, 2009) In 2007, average remittances per person were 63 USD, compared with the average for OECD of 108 USD Remittance amount accounted for approximately 8% of total GDP and 220% of net ODA receipts The ratio of remittances to FDI was 0.8 (UNDP, 2009) International remittance inflows reached 7.2 billion USD, accounting for about 6.4% of total GDP in 2008 (Nguyen, 2008) International remittance inflows come from two main sources including temporary exported workers and permanent migrants (Viet Kieu) While Viet Kieu has become older and less altruistic, number of exported labors will be the main sources of international remittances in

future Remittances are considered as person-to-person flows that are stable and well targeted to the specific need of each individual recipient In addition, international remittances are not necessary to be paid back since they seem to be altruistic money transfers For that reason, policymakers will need much more empirical studies on the use of international remittance inflows since remittances,

in fact, positively contribute to communities and hometown nations of abroad migrants

A large number of Vietnamese households are currently living below the poverty line mainly because of unemployment, low wage and low educational level One of the ways for poor people gaining higher income is to migrate abroad with the desire to send remittances back home for supporting their families and offering more opportunities for their young children to go to school During 1985 and 2007, the HDI of VietNam increased by 1.16% yearly from 0.561 to 0.725 Vietnam has a rank of 116111 out of 182 countries with data in 2007 (UNDP, 2009) One of Vietnam's development goals is to push up its recent low ranking on HDI

by improving educational factor in international HDI scale In Vietnam, education programs are still costly and of low quality Moreover, national public budget is not enough to support free or low cost education programs for young individuals Thus, remittance receipts seem to be one of substitute capital sources for public educational investment rather than a complement

Currently, there are a series of studies examining the impact of remittances

on the schooling of young generations in El Salvador (Edwards & Ureta, 2003; Acosta, 2006), in Mexico (Hanson & Woodruff, 2003; Malone, 2007), in Philippines (Yang, 2008), in Dominican Republic (Amueto-Dorantes et al., 2008) and in Ecuador (Calero et al., 2008) and so on Those studies have found that remittances significantly reduce the likelihood of quitting schools, increasing children's years of schooling used to invest in their education path as well as pushing up the propmiion of school-aged children attending school Through the mechanism of diminishing liquidity constraints, amount of international

remittances probably encourages education investment and to some extent, increasing financial budget spent on academic training for young people

Until now, there is still no research on the impact of remittances on the likelihood of school enrollment of young generations in Vietnam Therefore, this paper will explore whether remittances statistically significantly influence young generations's schooling in VietNam This paper will apply the Logit model using Maximum Likelihood (ML) estimation method with data mainly taken from the Vietnam Household Living Standard Survey 2006 (VHLSS) The result of this study is crucial for policymakers to produce policies to encourage human capital investment while experiencing large out-migration and remittance inflows

The principal objective of this thesis is to determine whether, how and to what extent international remittances significantly affect the school enrollment rate of young generations in specific types of Vietnamese households

To meet this overall goal, the study will meet the following specific objectives:

(i) To investigate whether international remittances significantly

influence the school enrollment status of young generations in households in VietNam

(ii) To consider whether there is differential impact of remittances on

educational outcomes by gender, urban-rural areas and employment status ofyoung generations

(iii) To make recommendations to government on the policies and program that might best promote the effective use of remittances in human capital investment

school enrollment status by sex, urban-rural areas and employment status ofyoung generations in remittance receiving households?

government better channel international remittance inflows into human capital formation for the sustainable long term growth of the Vietnamese economy in the age of Knowledge Economy?

The research uses both qualitative and quantitative methods to evaluate the impact of international remittance income on the enrollment status of young generations in households Qualitative analysis suppmis evidences on the right selection of determinants affecting the education attainment of young generations Moreover, the Logit model is applied to investigate the significance and magnitude of the effect of each input factor on the improvement of young

individuals' school enrollment rates Upon the analytical results, suitable policies will be suggested and discussed further

Cross section dataset from Vietnam Household Living Standard Survey

2006 (VHLSS) will be used for this research All information of the VHLSS was collected by two times in the year of 2006 and by methods of direct interview with the heads households & key commune officials Main data used to run the regression model are extracted from the Section 1 "List of household members", Section 2 "Education" and Section 4 "Income" stated in the VHLSS dataset of 9,189 households surveyed

1.6 RESEARCH SCOPE

In order to identify the impact of migrant international remittances on the probability of attending schools of young generations, we only focus on individuals ranging from the age of 6 to 20 Each individual is treated as one unit

of analysis instead of one household This research will not choose a household as one unit of analysis because of two reasons First, this method will not allow the author to examine effects of specific characteristics of an individual including age, gender, job of each person Second, due to limitation of VHLSS 2006 dataset, exploration of the whole data at individual level will offer advantages of large sample size enough to run Logit model by using ML estimation method

This paper aims to study the impact of international remittances from migrants on education status of young generations left behind in all provinces in Vietnam It is difficult to focus on only one region in Vietnam because of limited data on remittance receiving families in VHLSS 2006

1 7 THESIS STRUCTURE

The study consists of 6 chapters Beside the chapter 1, the rest of this paper includes 4 chapters as follows:

Chapter 2: Literature review will be discussed in this chapter One section

is devoted for introducing the definition of international remittances In addition, theoretical frameworks relating to the motivation to remit, the microeconomic effects of international migrant's remittances on the schooling of young generations in destination households, determinants of young generations's education attainment are also reviewed And then the analytical framework will be introduced with detailed variable descriptions

Chapter 3: An overview of international remittances and education status of young generations in Vietnam will be illustrated in this chapter

Chapter 4: Methodology applied in this paper is mentioned This chapter will firstly mention the background of VHLSS 2006 data, sampling method and sample size In addition, it is necessary to evaluate the advantages and disadvantages of VHLSS data in order to argue its representative and reliability Finally, the econometric model, estimation strategy are orderly presented

Chapter 5: Estimation analysis is explained in details in this chapter

Chapter 6: Conclusions and Recommendations are mentioned in this chapter It summarizes the finding results and raises some useful policy recommendations

CHAPTER 2: LITERATURE REVIEW

The main objective of this paper IS to investigate the influence of international remittance inflows on the status of school emollment of young generations Therefore, the concept of international remittances will be clearly illustrated in section 2.2 Then, theoretical and empirical literature related to the link between international remittances and education will be discussed in section 2.3 Moreover, section 2.4 will be devoted to introduce the suggested analytical framework Finally, section 2.5 will briefly summarize some main ideas found in this chapter

Appendix 5 on remittances to the "Balance of Payments and International Investment Position Manual", International Monetary Fund introduced the economic concept of remittances as: "Remittances represent household income from foreign economies arising mainly fi'om the temporary or permanent movement of people to those economies Remittances include cash and non-cash items that flow through formal channels, such as across electronic wire, or through informal channels, such as money or goods carried across borders They largely consist of funds and non-cash items sent or given by individuals who have migrated to a new economy and become residents there, and the net compensation

of border, seasonal, or other short-term workers who are temporarily employed in

an economy in which they are not resident." (IMF, 2008)

That Appendix also offered 3 different definitions of remittances as follows:

(i) Personal remittance is considered as one kind of the

household-to-household transfer and the net earnings of nomesident workers

(ii) Total remittances include personal remittances, social security and

pension remittance amounts

(iii) Total remittances and transfers to nonprofit institutions serving

households include total remittances and donation amount from the government or enterprises to a charitable organization abroad And then this charitable institution will use such remittance receipts to support families

For the purpose of this research, remittances are "limited to personal international remittances but not include social security, pension transfers or transfers to nonprofit institutions serving households

2.3 THEORETICAL FRAMEWORK AND EMPIRICAL STUDIES RELATED TO THE

IMPACT OF INTERNATIONAL REMITTANCES ON THE EDUCATION OF

YOUNG GENERATIONS

2.3.1 Theoreticalliterature

The theory of family behavior

The process of children's education attainment has been viewed as one of the aspects of the theory of family behavior (Haveman & Wolfe, 1995) This theory considers one household as a production unit in which parents (or head of household) will make important decisions on many fields such as generation of family economic resources and uses of those resources (investment or consumption) Actions of parents will influence the attainment of young people in that family depending on the amount, the characteristics and the timing of the distribution of above mentioned resources

Figure 2.1 Home Investments in Young generations

Source: Leibowitz (1974)

Ability (lQ)

Final Schooling Level

Post-school investment

Figure 2.1 shows that final schooling level of children is generated from genetic characteristics transmitted by their parents, family income and home investment in time and goods devoted to young generations The framework above can be illustrated in three equation follows

(1) Children's IQ or ability= F (Heredity, Home Investment)

(2) Children's Schooling= F (Children's Ability, Home Investment, Family Income)

(3) Children's Income= F (Home Investment, Children's Schooling, school Investment, Children's ability, Family Income)

Post-In this case, international remittances constitute a crucial part of total family income Inward remittances are considered as additional financial sources

to total family income that help to release liquidity constraints and allow extra investment in education of young generations

The theories related to motivation to remit

A series of studies have focused on exploring two aspects of remittances including the motivation of foreign migrants to remit and the use of remittances in destination countries Over the past two decades, several researches have done to explain and provide more evidences on the reason why international remittances exist

Firstly, one of the motives of international migrants to send money back home is altruism or solidarity, the willingness of a migrant employee living abroad

to remit money in order to support the family members in the same household in home country (Lucas & Stark, 1985) Thus, remittance flows create a close link between international migrants and family member left behind in hometown Rapoport & Docquier (2005) presented that the amount transferred by a migrant increases when his income goes up and the degree of altruism is higher In contrast, he will send back home less if his relatives in the home country earn more income or his degree of altruism is lower

Second, another motive to remit is exchanging (Cox, 1987) Abroad migrants transfer money to their relatives with the purpose of repayment of loans utilized to finance investment in human capital or expenses arising in their household

The theories on how remittances are spent, especially used in educational investment

In addition to researches on remitting motivation, an increasing number of current studies have concentrated on the use of remittances at the micro level Remittance receiving households may use the money sent by their relatives living outside to run or develop their small companies (Woodruff & Zenteno, 2007; Amuedo Dorantes & Pozo, 2006) or to cover medical expenditures (Amuedo Dorantes et al., 2007) Remittances are not only spent on daily consumption, but also are used for investment in human capital (Edwards & Ureta, 2003) This

paper only focus on discussing a series of papers exploring the mechanisms by which the migration and remittances influence the educational results of young generations in family left behind

First, international remittances will rmse the educational attainment of children in receipt households through the mechanism of lifting liquidity constraints (Ilahi, 2001; Edwards & Ureta, 2003; Hanson & Woodruff, 2003; Borraz, 2005; Yang, 2005; Rapoport & Docquier, 2005; Calero et al., 2008) In developing countries, the standard living in households is still very low Parents do not have enough money to support the education fees of their children even supported by small public budget on education Attending school will impose an additional cost and thus, lead to the substantial decrease in total income or household consumption In this case, internationally transferred money is a considerable fraction of the total family's budget and may help loosen the liquidity constraints Rapoport and Docquier (2005) mentioned the model of liquidity constraint and human capital investment The model consumed that in the first period, people left behind earn little income and regularly receive a fraction of money saved by relatives living outside Therefore, the next generation in remittance receipt households obtains more opportunities to attend schools with the condition that savings are not negative at the end of the first period In the second period, it assumed that migrants wished to return and living in their original country, then send all savings back home

Second, by studying liquidity constraint and human capital investment theory, one would expect that remittances significantly have positive effects on education attainment of young generations in the households having overseas migrating members However, the above-mentioned positive impact seems to be ambiguous because remittances may produce the so-called disruptive effect of household migration on the educational results of young generations in the event that the household has one or more family members abroad (Hanson & Woodruff, 2003; Rapopmi & McKenzie, 2006; Amuedo-Dorantes & Pozo, 2006; Milligan &

Bohara, 2007) If households have a family member going overseas, for example father, mother or older brother and so on, that may cause a negative effect on total household production, as a result, the inflow of income may also decrease Thus, members left behind especially young boys may need to pick up the slack and work harder than before in order to gain more wages Therefore, the probability of higher educational level of those young is lower than that of children in no-migration households

Finally, negative effects of international remittances on the educational attainment may arise if remittances from family members focus on facilitating the migration of other people left behind Kandel and Kao (200 1) argued that migrants' transfer money might produce negative or positive impacts on the attainment of education of family members in the home country depending on the benefits of the education in the host country For instance, if education is actually not valued in the host nation, the probability of migrants in the home country using remitting resources to facilitate out-migration instead of investing in education in the home country becomes higher

In order to describe the conceptual framework of relationship between remittances and education investment, it is important to investigate the interrelationship between migration, remittances and human capital decisions Migration, remittances and human capital decisions are one part of the overall household strategy, thus rarely made by an individual but within the whole family (Brown, 2005) The link of three factors above is so complicated and sophisticated that one may get easily confused At first, the correlation between remittances and migration will be reviewed It is reasonable to argue that migration causes remittances since remittances will never flow into host countries if migration does not exist On another hand, the decision to migrate overseas mainly comes from the desire to earn more income, and is influenced by remit motivation (remittances probably cause migration) Similarly, the relationship of migration and human capital investment also occurs like the former While the schooling attainment of

the young generation partly drives parents to migrate (human capital causes migration), so too can migration opportunities influence the decision of investment

in human capital (migration causes human capital investment) Finally, there is also a tight relationship between remittances and investment in human capital Any investment including human capital investment need to be funded by financial sources such as savings or credit of the households People can not implement investment in education if savings or credit sources are not available at the time of making investment decisions In this event, remittances can be a considerable fraction of families' budget to finance human capital investment strategy The inflow of remittances will help eliminate the financial constraint faced by households and make the decision of human capital investment become real Upon the relationship between remittances and human capital investment, it

is clear that the existence of remittances is a necessary condition but not a sufficient condition for investment in human capital (Brown, 2005) This argument

is rational in the case that remittance receiving households may not use the money sent by their overseas relatives to cover education related expenditures but may prefer to finance in other fields such as business establishment or new house building

2.3.2 Empirical literature

The previous empirical studies have found some important determinants of schooling attainment of young generations including families' characteristics (household income, place of birth, parental schoolings, parents' occupation, household size, gender of household head, and so on) and young individuals' characteristics (age, gender, birth order, and so on)

Parents' education is one of the most important factors that determine the schooling results of children in households · Educational rates of parents and grandparents significantly influence young generations's schooling attainment (Lillard & Willis, 1994 ) The result was found by using data extracted from the

Second Malaysia Family Life Survey, which provided enough information on the education of four generations within a family Especially, the schooling of mother significantly have larger impacts on schooling of children than that of the father in Nicaragua (Behrman & Wolfe, 1984)

In addition, parents' income and occupation also affect the final education outcomes of children (Parish & Willis, 1993) This empirical study was conducted

in order to explain the completed education level of each child in one Taiwanese household The regression results showed that the background of an individual is very important to their schooling level Children, who obtained higher education level than others, born and growing up in urban areas and had mother and father originating from mainland, and living in families with parents of higher status in income, education and work

Gender of household head also significantly influence the education attainment of young generations Previous studies have reported that female headship in a family would lead to higher educational level of young generations since mothers were more likely than fathers to spend money on education expenditure of their young generations (Buchmann, 2000)

Young individuals' characteristics including birth order, age and sex also affected their education attainment People born early would get less education than those born late in Taiwan The academic process is also different between boys and girls Boys usually get higher schooling level than girls in Taiwanese urban areas (Parish & Willis, 1993)

There are still very few empirical studies analyzing the impact of international remittances on schooling The complicated relationship between the international remittance receipts and academic outcome of young generations in households is still unclear and inconsistent among related studies in recent years While many studies found out that remittance inflows did have significantly positive effects on educational outcomes of children (Edwards & Ureta, 2003; Acosta, 2006; Yao & Donald, 2007; Malone, 2007; Amueto-Dorantes et al., 2008;

Calero et al., 2008), others reported negative or unambiguous results (Borraz, 2005; Amueto-Dorantes & Pozo, 2006; Milligan & Bohara, 2007) That inconsistence partly comes from different methodologies, dataset and research places selected by each author

Edwards and Ureta (2003) conducted a nationally representative households survey from El Salvador (1997 EHPM) in order to study the impact of remittances on school retention rates of young generations between 6 and 24 years

of age The study applied the Cox proportional hazard model with the purpose to investigate how two types of income, international remittances and other financial sources in total family budget, affect the schooling of the young This paper presented the results that remittances have a much larger influence on the hazard

of leaving school in both rural and urban areas than other kinds of income The effect of remittances on school attendance rates in poor areas was much larger than this in rich provinces although parents obtained low education level In urban areas, the effect of remittances is 10 times larger than the size of the impact of other financial sources In rural areas, the effect of remittances is about 2.6 times larger than the size of the impact of other kinds of income Estimated results also insisted that parental schooling is one important determinant that significantly influences the probability of leaving school Moreover, household's budget constraint is proven being a key factor that determining the education level of young generations However, one of the large disadvantages of Edwards and Ureta (2003) is the methodological problem of failing to consider the potential sample selection and endogeneity of remittance receipts

By using a different model compared with Edwards and Ureta (2003 ), Borraz (2005) tried to explore relationships between international remittances and education ofyoung generations by multivariate regression model

Y = + tR + + + e

Where Y denotes for the highest grade of the child And R stands for remittances receipt, A denotes for absent headed household Xi includes control variables such as child gender, age, mother education and location size This research is different from other previous studies on the impact of remittances on school attendance of children because it adds one more crucial variable of absent headed household The results show a positive and small impact of remittances on human capital only in cities with fewer than 2500 residents and within households

in which mothers obtain low level of education However the magnitude of this effect is not substantial It is important to note that in the above specification, the coefficient estimate for remittance receipt will be biased in the presence of any correlation between household remittance receipt and the error term Remittances may also be related to household income and wealth This correlation may result

in inconsistent and biased estimates of the educational impact of household remittances

While Borraz (2005) decided to estimate the impact of international remittances on young generations's schooling by multivariate regression model, Amuedo-Dorantes and Pozo (2006), on the other hands, applied the the IV Probit model to examine the likelihood that a school-aged child has an age-appropriate education:

P (Age appropriate education ir= liRir, Xit) = <PWRit> o'Xir) + eir

Where €jf I Rif' xif N(O,l) for i=l, ,n individuals in the sample The vector Rir is a dummy variable denoting whether the household receives international remittances The vector Xir includes the information on households' and children's characteristics as Appendix 1 illustrates Amuedo-Dorantes and Pozo (2006) estimated the equation above as a Probit model with endogenous regressors via maximum likelihood because they want to account for the endogeneity of remittance receipt which refers to the potential correlation between remittances and the error term One interesting thing of this study is to compare

the impact of remittances on education attendance in out-migration households and those in families without migration It is noted that the coefficient on remittance receipt factor may be partially decreased due to disruptive impact of out-migration households Furthermore, by doing this, they can investigate which

of the two disruptive effects and income effects dominate The estimated results showed that remittances have not significantly affected on age appropriate schooling level of young generations living in out-migration households Those outcomes proven that in Dominican Republic nation, the disruptive effect of out migration dominated that of income In addition, girls get more benefit from the receipt of international remittances than boys

Acosta (2006) investigated the schooling of children by considering two groups of ages 11-14 and 15-17 because the group with more than 15 years old may have to enter into the labor market to gain more income for their families This paper concluded that girls and boys aged 11-14 tend to obtain higher enrolment rates in case of receiving remittances Surprisingly, the positive impact did not apply to older boys of 15-17 ages because they may have to work and pick

up the slack instead of going to school Acosta (2006) also applied the IV Probit model and use Robust estimates to control both selection and endogeneity issues

Milligan and Bohara (2007) compared the influence of remittance income and non-remittance income on education of young generations by robust OLS regression method In this study, the impact of remittance receipts on schoolings

of young generations is unambiguously positive The author insists that income is better divided by regional price indices to generate real aggregate income since prices are different across regions in Nepal Malone (2007) suggested the new way

of measuring education attainment by including both years of schooling and ages

in the formula By using OLS regression method, the author found that remittances did effect the education of young generations and mother preferred to invest remittance income in the education of their young generations

Amueto-Dorantes et al (2008) also used IV Linear Probability Model and

IV Probit model to estimate the impact of international remittances on schooling

of children in Dominican Republic However, this paper focused on separating the remittance receipt and migration impacts They re-estimate the schooling model

by using database in which households do not have out-migration family members Then, they compare the results using non-migration households with ones using all households in the sample and conclude about the differential impact

of remittance receipt and migration on the likelihood of the education of young people Remittances significantly lead to higher school attendance for all young generations in the community which is full of private and public schools in both non-migration and out-migration households

Calero et al (2008) explored the link between migrant's remittances and school enrollment of young generations in Ecuador Like other researches, this paper used the IV Probit model to investigate the probability of school enrollment

of young generations In addition to the remittance receipt variable, this research also added more independent variables including child's characteristics, household head's characteristics and regional welfare and demographic characteristics as illustrated in Appendix 1 This research found out that one dollar increase in remittances per month leads to 0.09 percent point increases in enrollment rate Moreover, girls likely obtain more benefits than boys, especially in rural areas Based on previous related empirical researches, the next section will introduce the analytical framework of this paper

2.4.1 Empirical model

From theories and empirical studies mentioned above, it is true that school enrollment status of young generations is influenced by international remittance receipts, and other alternative variables including families' and young

generations's characteristics In this paper, the suggested research model is Logit model as follows:

X: characteristics ofthe family

Z: characteristics of the young generations

This research chooses to measure the impact of international remittances on the probability of young generations to attend schools The outcome variable is young generations's current enrollment status, coded 1 if the child is recently enrolled, or coded 0 if the child is not enrolled now The choice of a binary dependent variable rather than cardinal years of schooling fit the objective of this study that aims to access the variation of probabilities of being enrolled in schools

of young generations in case that yearly per capita remittance receipts as well as other crucial determinants of school enrollment change It is, therefore, appropriate to apply the Logit model to estimate the above mentioned relationship

We treat annual per capita remittance receipts as the main explanatory variable of interest in this research Moreover, this research will not separate the impact remittance receipts impacts and migration effects on the education of young generations because the information of migrants in households was not obtained in the VHLSS 2006 This paper can only use the data of annual remittance receipts in households provided in VHLSS 2006

This study is expected to find that international remittances receipts will significantly increase the probability of emolling schools of young generations that is consistent with the results obtained by Amueto-Dorantes et al (2008) and

Calero et al (2008) Because parents in Vietnam who migrate abroad always wish

to send back home more income to facilitate their young generations's school fee and related expenses with the desire for their young generations to get brighter future We also expect that there are differential impacts of international remittances on the probability of school emollment among boys and girls in both urban and rural areas

2.4.2 Variables introduction

a) Dependent variables:

The research will exploit the information about the school emollment status

of young generations in households as variable of interest Dependent variable is binary one It is coded 1 if young generations are currently emolling in schools In contrast, it is coded 0 if young generations have not yet attended academic program This research treats a child as a unit of analysis instead of one household The observed sample will be restricted on young generations aged 6 to 20

b) Independent variables:

Remittance receipts (R) is the mam explanatory variable of interest Remittance receipts are measured by annual amount of international remittances per capita that a household receives The remittance amount is expressed in local currency (VND)

Family's characteristics (X) reflect the role of the family intervention in improving the education outcomes of young generations They include abilities and occupation status of household head, income as well as household size and regions where the family currently locates

Family income, measured in local currency, is current permanent income per capita in household (excluding remittance receipt income)

Percentage of school aged young generations in the household It is measured by the number of young generations aged between 6 and 1 7 divided by the total household size

Age of household head is measured by calendar age, in year

Gender of household head is a binary variable It coded 1 as males and 0 as females The female headship is expected to lead to higher education process of young generations since women tend to invest more money in their young generations's education than men

Education of household head is measured by the highest educational grade obtained by the head of family

Occupation status of household head is a dummy variable The variable is coded 1 if the household head is currently working In contrast, the variable is coded 0 if the household head is unemployed at the time of survey

Area is a dummy variable It includes urban or rural area

employment status

Gender is a binary variable It includes male and female

Age of young generations is measured by the calendar age

Employment status is a dummy variable, recoded 1 if the child is currently working, vice versus recoded 0 if the child is not currently employed

of being enrolled in schools of young generations The next chapter will present in-depth overview of international remittances and education status of young generations in Vietnam

CHAPTER 3: OVERVIEW OF INTERNATIONAL REMITTANCES AND EDUCATIONAL ATTAINMENT OF YOUNG GENERATIONS

IN VIETNAM

This chapter commences with section 3.2 in which will demonstrate a general overview of international remittance flows during recent years in Vietnam

as well as related government's policies in attracting more overseas remittances It

then discusses about the education status of Vietnamese young generations in section 3.3 Finally, a summary of international remittances and education background of young generations in Vietnam will be illustrated in section 3.4

(a) Vietnamese migrants and exported workers

Remittances flowing into Vietnam come from two kinds of people including abroad migrants and exported workers In 2007, it was reported that there are about 3.2 million of Vietnamese migrants permanently residing in foreign countries In addition, 80 percent of those are currently living in developed nations (Nguyen, 2008) While number of permanent migrants has become unchanged during recent years, it has seen a very sharp increase in exported labors The Vietnamese government has started to exporting labors since 1980s The migrant workers moved to socialist countries in Eastern Europe such as the former Soviet Union, Germany and Bulgaria Since 1990s, the exportation of labors has been conducted by individual enterprises instead of government via the labor cooperation program In 2007, there were approximately 79,000 workers sent overseas and this number is expected to reach about more than 100,000 employees per year in 2010 (Dang, 2007) The year of 2008 has faced a large number of approximately total 500,000 Vietnamese temporary migrant labors now working in more than 40 different nations and territories (Nguyen, 2008) It is the

prospect of better life with good wealth that stimulates people to migrate abroad Figure 3.1 below clearly captures the increasing trend of number of exported workers during 1992 and 2007

Figure 3.2 Percentage of Vietnamese migrant workers by countries 1992- 2006

OTHERS, 24.01%

L _

MALAYSIA, 27.48%

(b) Impacts of foreign remittances in Vietnam

Remittances are obviously considered as one of the most important sources

of external financial funds for development of any economy In association with increasing number of migrants working or living abroad, inward remittances in Vietnam have also quickly reached very high volumes since recent years From

1999 to 2008, the volume of internationally remittances always goes up very quickly Figure 3.3 captures sufficient data of remittance receipts within 11 years from 1999 to 2009

Figure 3.3 Inward remittances from Vietnamese migrants during 2001-2009

Source: WB (2009) and SBV's estimation in 2009 Ol

From 2001 to 2007, the average growth rate of international remittance inflows was approximately 17% The year of 2008 has seen a huge volume of remittances coming into Vietnam with the growth rate of about 44% However, because of financial crisis in 2009, remittance receipts are expected to be lower than those of last years but higher than those of year 2007 It is serious that slowing remittance inflows will probably have negative influences on Vietnam's balance of payment position, particularly in the event of smaller foreign direct investment Remittances can be considered as one of the most important external financial sources since they account for nearly 6-8% of total GDP as illustrated in figure 3.4

(1) Data of inward remittances of in 2009 is estimated by The State Bank of Vietnam This figure can be obtained from http://www.vietnewsonline.vn/News/Business/6668/Vietnam-ups-2009- remittance-feast- to-6-bln.htm

Source: Hernandez-Cess (2005), IMF (2009), WB (February 2009), Economist

Intelligence Unit Report (2009) cited in Nguyen (2008)

Inward remittance volume is reducing mainly due to crisis existing in United States since more than 50% of remittance receipts come from migrants living in North America Table 3.1 illustrates some main sources of international remittances by origin

Source: Pfau & Long (2008b}'s calculation from VHLSS 1992/93 and 1997/98

The proportion of total volume of international remittances flowing from North America, in fact, goes up from 47.3% to 63.8% during 1992 and 1998 In contrast, the percentage of remittance receipts coming from European nations lightly decreases to 15.6% in 1997/98 from 22% in 1992/93

Total number of households rece1vmg international remittances from overseas migrants is quite small in VHLSS 1992/93, 1997/98, 2002, 2004 and

2006 as expressed in Table 3.2

Table 3.2 Percentage of households receiving international remittances

and author's calculation from VHLSS 2006

The proportion of families getting remittances only reaches less than 6% in VHLSS 1992/93, 1997/98, 2002 The percentage of remittance receipt households lifts up to 7.3% in VHLSS 2004, while the figure in VHLSS 2006 goes down to 6.36% The share of total remittances is quite different by years as the table 3.3 demonstrates

Table 3.3 Share of total remittances in urban and rural area

Rural 80.0 20.9 0.3 77.6 25.2 0.3 76.8 49.0 0.6 74.1 49.9 0.7 Urban 20.0 79.1 4.0 22.4 74.8 3.3 23.2 51.0 2.2 25.9 50.1 1.9

*Note:

(1 ): Share of total population

(2): Share oftotal remittances

(3): Ratio of remittances to population

and 2004

During 1990s, the proportion of inward remittances in urban areas, which is around 79-75%, is apparently larger than that in rural areas Conversely, after the year of 2002, every thing changes quite obviously The share of remittances distributes to urban and rural areas roughly equivalent

Moreover, the figure 3.4 will explain how the source of international remittances is used in Vietnamese households

Figure 3.5 Uses of international remittances (%)

Consumption, 73%

Source: Pfau & Long (2006) cited in Nguyen (2008) Vietnamese families use most of their inward remittances to facilitate daily consumptions of households In reality, remittance receipts spent on consumptions account for about 73% of total amount Other fractions of remittance budget are allocated for household construction (14.4%) or investment in non-farm activities (6%) On another hand, there is only 6.6% of total remittance income left for other spending purposes including education investment for young young generations Those figures make clear the idea that most of international remittances were used for daily consumption and just a little fraction was allocated for investment purpose such as human capital investment Therefore, the larger volume of international remittances the households receive, the lower the liquidity constraints are and as a result, remittance receivers are able to spend more financial sources

on education investment

During past decades, increasing international remittance inflows have actually played an important role in the development of Vietnamese economy since they account for around 6-7% of total GDP This kind of external financial sources, in fact, goes directly to households that have migrant family members and not needed to be refunded as FDI or ODA inflows Remittance resources contribute to families' total income budget and allow individuals in households consume or invest more, especially spending on academic programs The next section will review some figures of education in Vietnam

Table 3.4 demonstrates in details the net enrollment rate in year 2006 by urban rural areas and gender It is obvious that people in rural areas usually have low enrollment rate in both primary and secondary schools because of poor life, lacking of infrastructure and low education level In addition, the higher education level, the lower net enrollment rate Moreover, the net school enrollment rate in both rural and urban areas is nearly the same at primary level The difference, however, becomes visibly larger at lower and upper secondary level of schoolings Table 3.4 Net enrollment rate in year 2006 by urban rural areas and sex

The rate of primary enrollment is about 89% in both rural and urban areas

At lower secondary level, the rate in urban areas reaches 83% while the number of rural young generations going to lower secondary schools only accounts for around 78% The difference in this case is only 5% between 2 areas At upper secondary level, the difference in the enrollment rate of rural and urban areas becomes larger and larger, which is approximately 16%

Moreover, the net school attendance rates are also different between boys and girls Girls in many less developed countries usually gain lower school enrollment rates than boys Nevertheless, this conclusion is not suitable with the case of Vietnam since table 3.4 shows that girls tend to obtain higher enroll rates

in primary and secondary schools rather than boys The difference between the enrollment rates of boys and girls is about 1-5% at lower and supper secondary level This phenomenon is easy to understand because boys are likely to pick up the slack at very early ages to earn more income with the purpose of helping their families As a result, they have a tendency to spend all their time on working instead of going to schools like their female friends

Table 3.5 will indicate the average expenditure on education by many social demographic factors of young generations The average expense on education per person per annual is around 1 million VND This value is extremely high in urban areas (2 million VND) and a little bit low in rural areas (894 thousand VND) A large amount of education expenditure is spent on school fees, textbook and extra classes Average school expense is nearly similar between boys and girls Older young generations need more financial resources since they need

to enter into higher academic programs that are associated with higher tuition fees and other related expenses People aged 15 or older will need double amount of money than others younger This evidence also supports the explanation that older young generations tend to quit going to schools earlier than others having smaller ages