Determinants of households expenditure on english language education, the case of ho chi minh city

ABSTRACT This study attempts to investigate household’s expenditure on English education for primary school children using the surveyed data from 267 randomly collected households in Ho

VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN

Academic Supervisor

Dr PHAM KHANH NAM

HO CHI MINH CITY, 2013

I thank so much for all lecturers because of their valuable supports as well as all my family, my friends because of their helps and encouragements during my study

One again, I am grateful to all of you Thank you very much!

LUU THI KIEU OANH

ABSTRACT

This study attempts to investigate household’s expenditure on English education for primary school children using the surveyed data from 267 randomly collected households

in Ho Chi Minh City As a result, the current payment for English education of household

in Ho Chi Minh City plays an important role in total expenditure for education It takes up 61.4% of basic education in average Compared with household’s total income and total expenditure, it makes up 4.7% and 6.7% of total income and total expenditure, respectively The study also found that total income, total expenditure, current payment for extra education, education level of wife and household own a house has a positive relationship with English education expenditure

Keywords: English education, expenditure, income

TABLE OF CONTENTS

CERTIFICATION i

ACKNOWLEGMENTS ii

ABSTRACT iii

LIST OF FIGURES vi

LIST OF TABLES vii

ABBREVIATIONS viii

CHAPTER 1: INTRODUCTION 1

1.1 Problem Statements 1

1.2 Research Objectives 2

1.3 Research Questions 3

1.4 Research methodology 3

1.5 The structure of the study 3

CHAPTER 2: LITERATURE REVIEW 5

2.1 Theoretical literature 5

Theory of Education expenditure: Education production function 5

Theory of household production function 6

Theory of income elasticity of educational expenditure: 6

2.2 Empirical literature 7

2.3 Conceptual framework 13

CHAPTER 3: RESEARCH METHODOLOGY 15

3.1 Sampling strategy and data collection 15

3.2 Variables’ measurement and explanation 15

3.3 Model specification 23

CHAPTER 4: EMPIRICAL ANALYSIS RESULTS 28

4.1 General information of the Household’s characteristic in HCM City 28

4.2 General information of English education in Vietnam 29

4.3 Importance of English expenditure in household’s decision 31

4.4 Empirical analysis results 32

4.4.1 Descriptive statistics 32

4.4.2 Regression results 40

CHAPTER 5: CONCLUSIONS, RECOMMENDATIONS AND LIMITATIONS 46

5.1 Conclusions and recommendations 46

5.1.1 Conclusion 46

5.1.2 Recommendations 46

5.2 Limitations 47

REFERENCES 49

APPENDIX 52

LIST OF FIGURES

Figure 2.1 Analytical framework for the impact explanatory variable on English education expenditure 14 Figure 4.1 The education level of the parents in Ho Chi Minh City – Author’s survey 2013 29 Figure 4.2 The education level of the parents in Ho Chi Minh City – VHLSS 2010 29 Figure 4.3 The distribution of current English education expenditure in basic education, total income and total expenditure 32 Figure 4.4 Histogram graphic of the current payment for English education in Ho Chi Minh City 33 Figure 4.5 Histogram graphic of the dummy variable of current payment for English

education in Ho Chi Minh City 34 Figure 4.6 Household’s total income distribution in Ho Chi Minh City 35

LIST OF TABLES

Table 3.1 Variable description and expected sign for Model 1 26

Table 4.1 The development of number of foreign language center through the years 31 Table 4.2 Summary statistic of variables 37 Table 4.3 The relationship of total income and current payment for English education 39 Table 4.4 The relationship of household’s total expenditure and current payment for

English education 39

Table 4.5 The results of the models 40

ABBREVIATIONS

Cpe Current payment for English education Cpec Current payment for extra class

Eduh Education level of husband

Eduw Education level of wife

Engh English level of husband

Engw English level of wife

Genderhh Gender of household’s head

MOET Ministry of Education and Training

Nrofchildren Number of children

Soeh Sector of employment of husband

Soew Sector of employment of wife

Totalexp Household’s total expenditure

Wifeworking Wife is working

VHLSS Vietnam Household Living Standard Survey

CHAPTER 1: INTRODUCTION

1.1 Problem Statements

Foreign language currently is a prominent issue in worldwide education and society With the rapid development of information and communication technologies, through computer systems and networks, people between nations are closer and closer together understand each other through intermediate language that commonly is the English Most

of the knowledge such as books, academic papers, documentation in economy, science and technology, culture and society are mainly written in English and widely circulated on the global network Bolton (2006) likened the English expansion as the most rapid spread language all over the world In addition, the communication between different languages people through this common language which can understand and move toward the agreed long-term good cooperation In reality, English is rated as the most used and common speaking language in the world with more than 1,226 million users in 2003 and spread on

75 countries on the world according to Crystal (2003) Graddol (2008) figures that there are more and more people learning English as a second language and it will nearly get big amount of two billions next ten years There was also a replacement of English as a school subject instead of Russian from the participation of some Eastern European countries (Modiano, 2006) Since educating English was identified as an important factor indispensable, it should be considered by government investment, according to Nergis, (2011)In some developing countries such as Turkey, they recognize that “the only way to bring Turkish up to the level of the communicational requirements of the modern world is through the interaction of knowledge of English with the derivational convenience provided by the Turkish language reform.”

In Vietnam, Thinh (2006) observed that in the past ten years before the joining WTO, the spread of English has facing with an unpredicted speed in Vietnam The statistics shows that among the foreign language learner, the English learner takes up more than 90% of that volume While other languages such as Chinese, Japanese, French,

Korean, and German shares the rest of volume with smaller classes There was the emergence of requiring English for job seeking and the overseas studies which impulse over 900,000 learners came to the language centers in Ho Chi Minh city Moreover, teaching English program was also started at the universities, high schools, government agencies at that period time After the accession to the WTO accompanied with open policy has attracted a large number of foreign investors to enter the domestic market with small and large projects and the huge investments To meet the sufficient qualified labor force and ability to work with foreign partners, to communicate and understand the partner

is an important issue emerged at the present time Therefore, learning language is an urgent issue arises in the context of the integration of the international economic today Recognizing the need for English language development, many parents are planning for their children a foundation of language, namely English, since they start school

For these reasons, this study is to investigate how important parents see the role of English education by identifying parents’ expenditure on educating English language of their children at starting school age (primary school) in Ho Chi Minh city Then, it helps to clarify the important role of English language education in general and English in particular for the development of the country which should be concerned and invested by both the individual investor and the government in Vietnam

1.3 Research Questions

In this study, we attempt to answer the question: Do parent’s income, parent working environment, parent’s English level and other demographic issue (age, gender, marital status, location, home ownership and employment status…) impact on the expenditure to English education for their children? If yes, how much does it contribute to?

1.4 Research methodology

This study bases on Pritchett and Filmer (1999)’s theory of education expenditure Empirically, it combines both descriptive statistics and regression analysis to analyze the importance of English education and its expenditure determinants First of all, descriptive statistic is used to analyse the overall view of English education expenditure of the households in Ho Chi Minh City based on their total income, total expenditure and some of the characteristics of the Ho Chi Minh’s household Second, the ordinary least square (OLS) is employed to figure out the influent factors of English language expenditure on the children at primary school age Third, the logit regression is applied to investigate whether

or not these determinants affect on the household’s decision to spend for their English language education Fourth, the ordered logit regression are applied to settle the problem when the dependent variable, in this case is the English education expenditure, is not a continuous but it is ordered In this study, the answer the expenditure on English is ranked from zero to sixth level which is not equal between these levels

1.5 The structure of the study

This study is basically divided into five chapters and here is the first chapter of introduction Following the first chapter, chapter 2 reviews the three theories of the education expenditure, the household production function and the income elasticity of education expenditure Some recent empirical researches concerning the expenditure on education in Vietnam and around the world are also employed as the reference in this thesis Next, chapter 3 is about the research methodology including the way of collecting data, the variables’ explanation and measurement as well as the model specification The

chapter 4 will firstly present the overall information of the education in general and the English education in particularly in Ho Chi Minh City Secondly, we will discuss about the influent factors such as parent’s total income, parent’s total expenditure, parent’s education level and English level, and some of the household’s demographic characteristics which will affect on the English education expense for their children at the primary school age or not?

CHAPTER 2: LITERATURE REVIEW 2.1 Theoretical literature

In order to comprehend the purpose of this research, we first discuss education expenditure theories

Theory of Education expenditure: Education production function

Pritchett and Filmer (1999) present a theory of education expenditure via the education production function Their theoretical literature shows that a “production function” is an explanation of how much the amount of each input contributes to the amount of output Then, they apply it into the function of education In particularly, the education production function will study the causal effect of school inputs and nonschool inputs on the children outcome The school inputs, for e.g the education level of teachers, the size of class, teacher experience, the tool supports education as textbooks would influence the children achievement score Whiles, the non-school inputs such as the family’s factors, environmental factors and children congenital ability would also affect the outcome of children

The specific function of education production is defined as below:

Where Cit presents for children’s output, Sit presents for school inputs, Fit denotes the school inputs (family inputs) and Ii is the children’s congenital ability In this case, they use

non-a genernon-al term to represent non-approximnon-ately the congenitnon-al vnon-arinon-able thnon-at is fixed student contribution Because there is no data sets which can measure the nonfigurative variable like congenital ability

Theory of household production function

Both Becker (1965) and Muth (1966) raise a household production function model

to explain that commodities which are purchased on market simultaneously are the input of the household production In particularly, they employ the production function of child health which can be shown as below:

Theory of income elasticity of educational expenditure:

According to Benson (1961), the level of household income has direct effect on the educational expenditure which is measured by the income elasticity of education There are three levels of income divided into low, middle and high income In detail, the households

at low and high income are expected to have the amount of income elasticity of education from zero to one In contrast, a middle income household has value of income elasticity

higher than one This explains clearly that the middle income families have the tendency to concern about their children’s education and they spend more money for them than the other class For the low income families, they less concern about the quality of schooling

so that their education expenditure are predicted increase less than their total income

Use the Lorenz curve to measure the distribution of income to the education expenditure The form of income elasticity of educational expenditure as follows:

Whereas E i is the elasticity of income v i (x) is the Engel function of the i expenditure items

2.2 Empirical literature

2.2.1 Dependent variable

In this research, we attempt to explain the dependent variable that is the expenditure

of English language education In the study of English-language proficiency among U.S immigrants, Espenshade (1997) consider English language education as a human capital investment It concerns the expenses for the investment and the relevant expected return which spenders would response differently Consequently, expenditure on English is a diversified and complex variable so that it could not be able to interpret by one independent variable There are not many studies regarding this field Mainly works did study the expenditure of education in general but not in the detail of educating English language According to (Kanellopoulos (1997), Tansel (2006) and Donkoh (2011), the disbursement of education is determined essentially by household income, household expenditure, parent’s schooling, sector of employment, employment status, age of household head, gender of household head, mother work or not, household size, local and own the house or not Learning from these previous studies of education expenditure, we employ the above independent variables and develop the new model to figure out the

determinants of spending on English language education We will talk about the specific model later in the methodology section

2.2.2 Independent variables

 Household ’s income

The impact of household’s income on education expenditure is precisely existed It

is clearly that once the households get wealthier they would like to enhance their living standards Education is one of the important standards which are considered to invest for their new generation in priority, Donkoh (2011) Both Glewwe (1999), Tansel and Bircan (2006) and Donkoh and Amikuzuno (2011) found that when the household’s income rise they are more willing to pay for the private tutoring

To evaluate the impact of income on education expenditure, the income educational expenditure elasticity of the household is employed as a proxy Most of the papers use the Tobit model to valuate this income elasticity value (we will talk about the model later in the methodology section) In particular, Hashimoto and Heath (1995) discovered the highest income elasticity of education expenditure relates to the middle income level of household and the value is higher than one; while it ranges from zero to one for lower and higher income class This means that the rise in education expenditure of the middle income household is higher than other class as long as their income increases a unit And the lower and higher income class seems regardless the quality of their children’s education Kanellopoulos and Psacharopoulos (1997) figured out the income educational expenditure elasticity in case of Greece is 3.18 Whereas, Hashimoto and Heath (1995) come up with different value 2.35 in Japan However, they both conclude that the education expenditure is a kind of luxury good But this is not true in the conclusion of Tansel and Bircan (2006) they found the unitary elasticity of income on education demand Hence, they conclude education is not a luxury good in the case of Turkey as it is discovery in some researches

 Household total expenditure

In this research, we analyze the issue regards the English language education, an informal educational activity or also named as a private tutoring According to Stevenson and Baker (1992), the household tend to expense more money for investing in the informal educational activities as long as they expect their children to get higher education and successful career For this reason, it is easily to find that the probability of expense for educational activities rises in accordance with the expenditure level of the households The tutor fee spending for extra educational activities is about 11.2% of total expenditure of a family Among them, the expenses for foreign languages education take more than sixty-three percents of their total expenditure, a dominated devote amount for education, Kanellopoulos and Psacharopoulos (1997) Their point of view is clearly that parents’ spending for their children’s education increase sharply along with their total expenses

In addition, there are some other outstanding issues concerning the way of choosing variable so that it can avoid the simultaneous problem Some authors argued the simultaneous between total expense and individual expense would result in the bias And they suggest use instrumental variable such as household’s income to settle this bias, Liviatan (1961) In order to do that, Tansel and Bircan (2006) attempt to test the exogeneity of total expense of a household This test is named as Tobit model, defined by Smith and Blundell (1986) and include two steps First is the regression of total spending

on household’s income Second is using the residuals from the regression one and put it in the Tobit model of extra education spending to test the significance As a result, they found the significance at one percent reject the null hypothesis Hence, they use the total expense

of a household as an explanatory variable instead of using household’s income variable

 Parent’s education

There are many studies employ this independent variable to explain its relationship

of education expenditure such as Knight and Shi (1996), Kanellopoulos and Psacharopoulos (1997), Tansel and Bircan (2006), Qian and Smyth (2011) and Donkoh and Amikuzuno (2011) Nearly all of them found the positive impact of parental education

to their spending for children’s education In there, Knight and Shi (1996) point out two

main findings: “(1) parent’s educational attainment is the most important factor influencing children’s education expenditure (2) Father’s education is more important factor than mother’s education” Kanellopoulos and Psacharopoulos (1997) found the significance of parental education on the household expenditure on their children’s education The households where the head has higher education range are more willing to devote more income for education Specifically, it is approximately 2.2% increase in probability of family agree to expense for education when a household head get one more year of education

Especially, some authors emphasize the important role of mother’s education in making the decision Qian and Smyth (2011) said “the household where mothers have senior secondary school or college education are likely to spend more on education” Donkoh and Amikuzuno (2011) claims it connects to the fact that the persons who are well educated would be able and please to give their son/daughter the same opportunity and expect them got higher education level

 Sector of employment:

Either Kanellopoulos and Psacharopoulos (1997) or Andreou (2012) come up with the similar conclusion of this concerned that the households which are working in the private sector have tendencies to finance more for their son/daughter’s education Recently, Qian and Smyth (2011) also found that “household where fathers are working in professional occupations, are likely to spend more on education” So, the private sector is the key determinants of in their research regarding the working environment of the household head

 Employment status

According to Kanellopoulos and Psacharopoulos (1997), Donkoh and Amikuzuno (2011), the occupation of the parents could also affect the willingness of a household to invest in children’s education There are two schools of thought about this issue One is the occupation has a positive effect since it is supposed the lower class of household tend to

compensate their lack of education by spending more for your children’s education Other

is the negative effect on their expense The upper class parents think that education brings

a relative low value in compared with their household value Hence, they disregard the role

of education Nevertheless, Andreou (2012) said the occupation of parents does not affect significantly on private spending for education

 Age of household ’s head

Most of the studies employ the age of household’s head as an essential variable to evaluate its effect on the household’s investment in education Generally, the age of head appeared to be negative efficiency to their choice of spending according to Kanellopoulos and Psacharopoulos (1997), Donkoh and Amikuzuno (2011) and Andreou (2012)

Donkoh and Amikuzuno (2011) employ the age and age square variable represent for the young and old household head As a result, the older the household’s head, the less willing to spend for education They argue that the household with younger head is considered more advanced and comprehend the long-term benefit and the role of education

in comparison with their older counterparts

Kanellopoulos and Psacharopoulos (1997) divide this variable into six age groups: under 25, from 25 to 34, from 35 to 44, from 45 to 54, from 55 to 64 and above 65 They discovered that the top of spender for education belong to the group with heads from 35-44 years old It is approximately 62.6 percents of the heads in this group are pleasure to pay And the probability of expenses decline after this group

Nevertheless, Tansel and Bircan (2006) stated differently about this context When analyzing the demand for private tutoring expenditure in Turkey, they found that the more accumulation of head’s age, the more willing to pay for tutoring their child Hence, we will test the affective trend of the household’s head age on the investment in their child’s education expenditure

 Gender of household head

Likewise, we will attempt to test whether the gender would affect the household decision on education expenditure Learning from the empirical studies, they apply the dummy for this variable Some studies found the differentiation between mother and father spending willingly on their son/daughter And that a female head tend to expense more than a male head By contrast, Andreou (2012) did not find the meaning of gender on the household’s head expenditure on education

 Household size

One of the research studies the related field in case of Vietnam, Dang (2007) worked on “The determinants and impact of private tutoring classes in Vietnam” He argues that the number of children in a household is possibly an endogenous variable since the family’s decision of their child may concern either the quantity or the quality simultaneously In more details, the parents would increase their investment in educating their children as their household size is smaller That would be a choice concerning the quality rather than the quantity In addition, Jelani and Tan (2012) stated the same with Dang (2007) He reasons that an additional child causes the rise of the family expenditures

so that it would result in the less investment of parent in the education However, in the paper of Donkoh and Amikuzuno (2011) they emphasize the quantity aspect As the larger size of the household, the more probability of the parents expense for education

 Local

The authors such as Tansel and Bircan (2006) and Donkoh and Amikuzuno (2011) found the precise link between the higher expenditure on education and the urban areas They all conclude that the household in urban areas tend to spend more money for educating their child in comparison with rural counterparts Only Andreou (2012) said the vary household’s region does not result in different of education’s disbursement

In particular, Donkoh and Amikuzuno (2011) states the differentiation of the number of household expense for private education (the quantity of spending) and the amount of money they spend (the quality of spending) Their study shows there is a larger

number of household agree to spend for education in the rural part than the city part By contrast, the concentrated city has the higher quality of education investment than the rural part Thus, we will have household’s location as an endogenous variable as same as the above mention about the household age

 Household own a house

Both Kim and Lee (2002) and Donkoh and Amikuzuno (2011) claim that the household own a house is more likely to expense for their children’s education It is well accepted that having a house is equivalent to the better living conditions as well as correlated to the higher income For these reason, they are easier to take the charges of their son/daughter’s education

2.3 Conceptual framework

Overall of the above empirical literature, the study is going to test the relationship between the private expenditure for English education and the explanation variables as the following:

Figure 2.1 Analytical framework for the impact explanatory variable on English education expenditure

Source: Author’s analysis

Age Gender (dummy V) Marital status Location (dummy V) Home ownership (dummy V) Employment status Number of children

…

PRIVATE EXPENDITURE for English language Education

CHAPTER 3: RESEARCH METHODOLOGY

3.1 Sampling strategy and data collection

The study used the primary data collection by face to face survey A questionnaire was composed to cover the main questions which includes household’s total income, parent’s education level and English level, career of the parent, and some other demographic issue (age, gender, marital status) The households surveyed include those with children at pre-primary and primary education level in Ho Chi Minh City The households were chosen randomly from each districts or sub districts Households were selected randomly by the number of house and in one random ward of the district About

20 households were interviewed in each districts/sub-districts The internal districts in Ho Chi Minh City comprise district one to twelve, Tan Phu, Tan Binh, Binh Tan, Binh Thanh,

Go Vap, Phu Nhuan The sub-districts include Binh Chanh, Hoc Mon, Cu Chi, Nha Be whereas we excluded the Can Gio’s observation because of its low population density

When a questioned household which has no children at the primary school age, we ask the next house for replacement If the next three houses have no children at our objective age, we will leave this sample empty Some chosen households leave the empty answer sheet due to their private issues Hence, some districts didn’t get enough 20 responses and finally the survey got 267 observations in total for this study The questionnaires table would be listed in the annex

3.2 Variables’ measurement and explanation

3.2.1 Current expenditure on English education

As mentioned above, the expenditure on English is a diversified and complicate variable It is considered as a human capital investment, Espenshade and Fu (1997) It concerns the expenses for the investment and the relevant expected return which spenders

would response differently Since there are not many studies related to the language education expenditure, we employ factor affect the spending for general education and then develop a new model for the case of English education expenditure Following the results from these researches of Kanellopoulos and Psacharopoulos (1997), Tansel and Bircan (2006) and Donkoh and Amikuzuno (2011), this study focus on researching some similar factors such as the household’s total income, household’s total expenditure, parent’s schooling, sector of employment, age of household head, gender of household head, mother work or not, household size, local and own the house or not Then pointing out whether or not these factors would impact on the spending manner for English education of these household’s

In this study, we use the current expenditure on English education to measure the interest level of their parents to the children’s language education It is divided into 6 levels: zero means no expenditure for English education at the present time, one means the current expenditure is ranked from 0 to 300,000 VND, two means the current expenditure

is ranked from above 300,000 to 500,000 VND, three means the current expenditure is ranked from above 500,000 to 1,000,000 VND, four means the current expenditure is ranked from above 1,000,000 to 2,000,000 VND, five means the current expenditure is ranked from above 2,000,000 These ranks are based on the average current fee of the extra English class at school and the English center The price is collected from ten school and eleven English center for the beginning and pre-intermediate class for primary children See the list of school and English center in the appendix

3.2.2 Household’s total income

As discussion in the previous chapter, the impact of household’s income on education expenditure is considerable Once the households get wealthier they tend to enhance their living standards in which education is one of the important standards to invest, Donkoh and Amikuzuno (2011) Both Glewwe and Desai (1999), Tansel and Bircan (2006) and Donkoh and Amikuzuno (2011) found that when the household’s income rise they are more willing to pay for the private tutoring In this study, we suppose

that English education is one of the types of private tutoring and hence, it is influenced by the household’s total income There are two methods to evaluate the impact of income on education expenditure: direct and indirect evaluation The direct evaluation measures the impact of all levels of income to the current payment for English education in the OLS and Probit/Logit model Whiles, in the indirect evaluation the income is divided into low, middle and high income and then applied to calculate the income elasticity of English education expenditure It is executed by the Tobit model

In this study, the household’s total income variable is also divided into five levels: (1) below 5,000,000 VND, (2) from 5,000,000 VND to 10,000,000 VND, (3) from 10,000,000 VND to 20,000,000 VND, (4) over 20,000,000 VND in direct evaluation method And it’s ranked as low, middle and high income in the indirect evaluation method Based on the classification of income group of the World Bank 2012, low income is the group which has GNI per capita equal or less than $1,035 per capita; lower middle income,

$1,036 - $4,085; upper middle income, $4,086 - $12,615; and high income, $12,616 or more As calculation to VND, it means that lower income, 1,811,250 VND/head/month or less; lower middle income, from 1,813,000 to 7,148,750 VND/head/month; upper middle income, from 7,148,750 to 22,076,250 VND/head/month; high income, 22,078,000 VND/head/month Relatively, we suppose that low income group is the level (1) income, below 5,000,000 VND; middle income group is comprised of level (2), (3) and (4); high income group is level (5)

In the previous empirical studies, the highest income elasticity of education expenditure belonged to the middle income level of household and the value is higher than one; while it ranges from zero and one for lower and higher income class This means that the rise in education expenditure of the middle income household is higher than other class

as long as their income increases a unit And the lower and higher income class seems regardless the quality of their children’s education according to Hashimoto and Heath (1995) However, some studies found no difference between these income groups which affect on education demand For these reasons, the paper is going to investigate whether

income increase would bring positive expenditure for English education or not and whether income elasticity of English expenditure between the low, middle and high income group are different? Overall, relied on theoretical studies and prior empirical researches, the study aims to test the following hypothesis:

H1: The household’s total income has positive relationship with the English education expenditure children

3.2.3 Household’s total expenditure

In this research, the household’s total expenditure is also employed to value the relationship with the spending on English education According to Stevenson and Stevenson and Baker (1992), the household tend to expense more money for investing in the informal educational activities as long as they expect their children to get higher education and successful career For this reason, it is easily to find that the probability of expense for educational activity rises in accordance with the expenditure level of the households Kanellopoulos and Psacharopoulos (1997) point out that the expense for foreign languages education takes more than sixty-three percents of their total expenditure

on general education Consequently, it is clearly that parents’ spending for their children’s education increase sharply along with their total expenses

In our case, we did use the multiple choices for the answer of total expenditure question Based on the result of VHLSS data set 2010, the total expenditure answer in this study is ranked into six levels: (1) below 2,000,000 VND, (2) from 2,000,000 VND to 3,500,000 VND, (3) from 3,500,000 VND to 5,000,000 VND, (4) from 5,000,000 VND to 10,000,000 VND, (5) over 10,000,000 VND The study will test two contents: first test is whether there is a positive relationship between the household’s total expenditure and English education expenditure Second is whether the simultaneous between total expense and individual expense would result in the bias by using the income as instrument variable according to Liviatan (1961) In order to do that, Tansel and Bircan (2006) attempt to test the exogeneity of total expense of a household This test is named as Tobit model, defined

by Smith and Blundell (1986) and include two steps Step one is the regression of total spending on household’s income Step two uses the residuals from the regression one and put it in the Tobit model of extra education spending to test the significance In case of the regression result find the significance to reject the null hypothesis, we can use the total expense of a household as an explanatory variable instead of using household’s total income variable

In summary, the study is going to test the proposed hypothesis as following:

H2: The level of household’s total expenditure has positive relationship with the English education expenditure

H3: The current expenditure on extra education relates to the English education

expenditure

3.2.4 Parents’ education level

Many studies found the positive impact of parental education to their spending for children’s education such as Knight and Shi (1996), Kanellopoulos and Psacharopoulos (1997), Tansel and Bircan (2006), Qian and Smyth (2011) and Donkoh and Amikuzuno (2011) More detail, Knight and Shi (1996) stated that “Father’s education is more important factor than mother’s education” Thus, the study investigates on the parent’s education level to find out the association of it to the English education expenditure Similarly to the VHLSS 2010, the education level is ranged from one to eight in which the mother or father had finished the Primary school, Junior high school, Senior high school, Primary vocational school, Colleges, University, Master and Physical of doctor respectively The study attempt to test the hypothesis as below:

H4 : Parents’ education level has a positive relationship with the expenditure on English education

H5: The level of English of Parents is positively related to their spending on English education

3.2.5 Parents’ sector of employment

As the above mention, we see the households which are working in the private sector have tendencies to finance more for their son/daughter’s education, Kanellopoulos and Psacharopoulos (1997) or Andreou (2012) In this part, we want to figure out whether the different sector of employment would impact on the manner of English education expenditure For example, the parents work in the private sector which concerned foreign factors like a foreign company or Non Government Organizations (NGOs) are likely to spend more on English education of their children or not? In the situation of Ho Chi Minh City, there are four universal sectors of employment: Public sectors, Private sector which includes the domestic company, Private sector which includes foreign company and the households which are working in the private sector have tendencies to finance more for their son/daughter’s education The rest would be classified as others This variable is modified as a dummy variable Each different sector would be considered as a dummy variable when running the regression Totally we got four dummy variables in compare with the control variable, public sector As a whole, the study would test the following hypothesis:

H6: P arents’ sector of employment could explain the current expenditure on English education

3.2.6 Gender of household’s head

Some studies found the differentiation between mother and father spending willingly on their son/daughter In order to clarify this point of view, we will attempt to test whether the gender of household head would affect the household decision on the English education expenditure Learning from the empirical studies, we apply the dummy for this variable Whereas, zero represent to female and one represent to male After that we run the test of below hypothesis:

H7: There is a difference between a mother household’s head and father household’s head

in their way of expense on their children’s English education

3.2.7 Parents’ age

Generally, the age of head appeared to be negative efficiency to their choice of spending according to Kanellopoulos and Psacharopoulos (1997), Donkoh and Amikuzuno (2011) and Andreou (2012) Donkoh and Amikuzuno (2011) explained the household with younger head is considered more advanced and comprehend the long-term benefit and the role of education in comparison with their older counterparts Nevertheless, Tansel and Bircan (2006) stated reversely When analyzing the demand for private tutoring expenditure in Turkey, they found that the more accumulation of head’s age, the more willing to pay for tutoring their child

Learning from Donkoh and Amikuzuno, we employ the age and age square variable represent for the young and old household’s head However, we separate it into four variables that are husband’s age, husband’s age square, wife’s age and wife’s age square in this study And then we will test the affective trend of the household’s head age on the investment in their child’s education expenditure according to the below hypothesis:

H8: Parents’ age has a negative impact on current payment for English education

H9: Household has wife are working is more likely to spend on their children’s English education

3.2.8 Location of household

The principal of this point of view is that the household in urban areas tend to spend more money for educating their child in comparison with rural counterparts according to Tansel and Bircan (2006) and Donkoh and Amikuzuno (2011) In this study, because we limit the research area to Ho Chi Minh City, so we could only test whether or not it has the difference from urban areas and the suburban areas on the expense for English education This variable is treated as a dummy variable in which one represent for the household in urban area and suburban is the comparison variable Accordingly, the proposed hypothesis is:

H10: Households in urban area are likely to spend more on English education for their children than the suburban area

3.2.9 Household owns a house

Some studies conclude that the household own a house is more likely to expense for their children’s education such as Kim and Lee (2002) and Donkoh and Amikuzuno (2011) It is also well accepted that having a house is equivalent to the better living conditions and more likely to spend for their children’s education For these reason, we’re going to test the relationship between household own a house to their spending for English education And attempt to test whether it has interaction with the household’s total expenditure Same as before, it is treated as a dummy variable with 1 present for household own a house and 0 for household does not own a house After all, the study would test the below specific hypothesis:

H11: Households which own a house are more willing to pay for their children’s English education

3.2.10 Number of children

According to Dang (2007), the number of children in a household is possibly an endogenous variable since the family’s decision of their child may concern either the quantity or the quality simultaneously In more details, Jelani and Tan (2012) support the quality aspect That is an additional child causes the rise of the family expenditures so that

it would result in the less investment of parent in the education In contrast, Donkoh and Amikuzuno (2011) emphasize the quantity aspect As the larger size of the household, the more probability of the parents expense for education Therefore, this study will work on the household size variable whether it had impact on English education expenditure If so, what is the trend of impact? Does it cause positive or negative effect on the English education expenditure?

H12: Number of children has a negative impact of the English education expenditure

Whereby, E represents for the current payment for English education, X denotes for

income variable, Y denotes for English level of parents, Z is the parents’ working environment; and T presents for the group of other demographic factors such as parent’s age, gender of household head, wife working or not, number of children, location of household and household ownership, etc…

Specifically, the different regression methods are applied to find out the above

research objectives Firstly, the Ordinary Least Square (OLS) regression is employed to

test whether there has relationship between dependent variables and independent variables

In this case, there are two specific models employed as following:

Model 1: OLS model 1

Model 2: OLS model 2

Model 3 and model 4: Logit models with income (model 3) and expenditure (model 4)

Secondly, the logit regression is applied to investigate whether or not these determinants affect on the household’s decision to spend for their English language education Because the spender is a 0-1 limited dependent variable, a logit model is fitted with this analysis Difference from the OLS regression, this model presents the probability (P) of a household expenditure on education as a function of various characteristics

Where P is probability of dependent variable of a household spending for education,

β is the logit coefficient and X is the bunch of independent variables use for evaluate the dependent variable such as household characteristic such as income, household’s head age, gender, location, according to Kanellopoulos and Psacharopoulos (1997) In this regression, it is separated into two model 3 and 4 whereas model 3 analyze bunch of income and other explanatory variables and model 4 employ the total expenditure and other explanatory variables to compare with model 3 The marginal effects of each explanatory variable after the logit model are calculated to evaluate the impact rate of them

on English education expenditure

Model 5: Ordered Logit model

As long as the English education expenditure in this study is not a numeric format type and it is categorized into six levels from zero to five which called ordinal dependent variable Hence, the ordered logit model is suitable for analyzing the relationship between this kind of ordered dependent variable and the set of independent variable In this model,

the ordered outcomes are estimated as a linear function of the set of explanation variables and the cutpoints According to Fu (1998), the probability of observing outcome I correspond to the probability of estimated independent variables plus the error term, within the level of the cutpoints It is treated as below function:

Pr (cpe j = i) = Pr (K i-1 < β 1 x 1j + β 2 x 2j + … + β k x kj + u j < K i )

Where,

cpe j is defined as current expenditure for English education at level 0 equal to no expenditure; 1 equal to the expense below 300,000 VND, 2 equal to the expense from 300,000 VND to 500,000 VND, 3 equal to the expense from 500,000 VND to 1,000,000 VND, 4 presents for the expense from 1,000,000 VND to 2,000,000 VND and 5 for the expense over 2,000,000 VND

x 1, x 2, … x kj denotes for the independent variables: the household’s total income, total expenditure, current payment for extra class, parents’ education level, parent’s English level and some others demographic characteristics such as parent’s age, gender of household head, wife working or not, number of children, location of household and household ownership

β 1 , β 2 , … β kdenote for the explanation coefficients

K 1 , K 2 , … K k-1 , are the cutpoints

u jis the logistically distributed of error term

Variable name Description Units Expected sign

dIncome2 Dummy household’s total income variable Equal to income rank from

5,000,000 to 10,000,000; otherwise equal to income below 5,000,000 VND Positive (+)

dIncome3 Dummy household’s total income variable Equal to income rank from

10,000,000 to 20,000,000; otherwise equal to income below 5,000,000 VND Positive (+)

dIncome4 Dummy household’s total income variable Equal to income rank over

dtotalexp2 Dummy total expenditure variable Equal to rank from 2,000,000 to

dtotalexp3 Dummy total expenditure variable Equal to rank from 3,500,000 to

dtotalexp4 Dummy total expenditure variable Equal to rank from 5,000,000 to

dtotalexp5 Dummy total expenditure variable Equal to rank over 10,000,000;

eduh Education level of husband Ranked from 1 to 8 level Positive (+)

englishh Dummy English level of husband variable Equal to 1 if he know English,

englishw Dummy English level of husband variable Equal to 1 if she know

dwifeworking Dummy employment status of the wife variable Equal to 1 if she stays at

dlocal Dummy location of the household variable Equal to 1 if it locates in

ownh Dummy variable of household own a house Equal to 1 if they own a

Source: Author analysis

CHAPTER 4: EMPIRICAL ANALYSIS RESULTS

This chapter mostly represents the empirical analysis result and discussion Firstly, the chapter presents the overview information of the education in Ho Chi Minh City In there,

we focus on the parent’s education and parent’s English level as well as the income structure in HCM City Secondly, the section summarizes the English education system in

Ho Chi Minh City Thirdly, the study comes out with the empirical analysis results of the collected data

4.1 General information of the Household’s characteristic in HCM City

From the survey, the study find that the parents’ education level of household which has children at primary school age in Ho Chi Minh City mainly belong to Senior high school and University level In there, the wife and the husband have nearly equivalent education distribution However, the husband seems having higher education level than the wife in general As we can see in the below figure 4.1, the percentage of husband complete university is 29.96% while it is only 25.09% for the wife Moreover, within the conducted households, the household has husband finish Physical of Doctor degree are 7.87% whereas there is no household with wife get this high education level

In comparison with the result of Vietnam household living standard data set (VHLSS 2010), the statistical overview of families shows: (1) the higher education level of husband than woman; (2) parents’ education level in HCM City substantially distribute in

no education, primary and junior high school and senior high school Besides, the parents get university also takes a considerable number at 11.69% in total

Figure 4.1 The education level of the parents in Ho Chi Minh City – Author’s survey 2013

Source: Author’s analysis

Figure 4.2 The education level of the parents in Ho Chi Minh City – VHLSS 2010

Source: VHLSS 2010

4.2 General information of English education in Vietnam

In Vietnam, the English in general education is assigned as a compulsory subject

two programs of learning and teaching English: the three year set for student from class 10 – 12 and the seven year program for those started learning from class 6 – 12 However, after 2002, English is a forced subject for junior and senior high school as well as being an elective subject at primary school In primary school, English is taught from class 3 – 5 with two periods week, whereas, it increases to three periods per week for junior and senior high school This new English language teaching program has been used for the whole general educational system of Vietnam to the present The propose of this program

is to equip the student after finished senior high school a basic knowledge of English be able to use proficiently English at four skills: listen, speak, read and write and the ability to understand the same level of textbook with the assist of dictionary

For the developing of English education, according to Ministry of Education and Training (MOET), the number of foreign language in Vietnam increases sharply from 843

to 1,935 centers in five years from 2008 to 2013 See below table for the yearly development of language center in last five years Moreover, there is a National Foreign Languages 2020 Project according to Decision No 1400/QD-TTg signed on 30/9/2008 by the Prime Minister to “Teach and learn foreign language in the national education system, period 2008 – 2020” It was a great and potential project if it had done successfully The project firstly develops and issues the unique, detailed qualifications framework of foreign language, including 6 levels, compatible with the international language proficiency level Secondly, develop and implement new foreign language training programs compulsory at the basic school level to meet the following qualifications after graduation: primary school level 1 qualifications according qualifications framework of foreign language; junior high school level 2, senior high school level 3 of qualifications framework of foreign language Thirdly, implement training according to new foreign language program for the university Establish and implement teaching by English in some basic subjects and major subjects for the last year student of university

However, the actual situation of teaching and learning English in Vietnam is still insufficient Along with learning English at school, some parents also equip their children

with English knowledge by sending them to the center of English Currently, there are many domestic and international English language centers which are operating efficiently and high profits The English international center such as the British Council, VUS, VASS, Apollo, ILA are available in Ho Chi Minh City and Hanoi; are planning for the large investments and expand on a national scale In addition, there are numerous other small centers, the type of English education at school and at home Hence, many schools and English center from either private or public sector established to meet the enormous demand of English education in Vietnam

Table 4.1 The development of number of foreign language center through the years

Source: Ministry of Education and Training

4.3 Importance of English expenditure in household’s decision

It can be seen from the figure 4.3 that the current payment for English education of household in Ho Chi Minh City plays an important role in total expenditure for education

It takes up 61.4% of basic education in average The basic education is the measure of total education fee at school On the other hand, compared with household’s total income and total expenditure, it makes up 4.7% and 6.7% of total income and total expenditure, respectively The study also performs that education English is considered as same as the private tutoring It shows 84.7% and 84.3% household agree to spend for English and for other extra class in turn Mainly the spending are below 300,000 VND per month and 300,000 VND to 500,000 VND per month See the table in appendix

Figure 4.3 The distribution of current English education expenditure in basic education,

total income and total expenditure