Các mô hình phân tích một số chỉ tiêu giáo dục việt nam tt tiếng anh

With the purpose of analyzing factors affecting expenditure on education of Vietnamese households at all educational levels and its impact on enrollment rate, thereby providing relevant

1

INTRODUCTION

1 Rationale

Investment in education and human capital is always considered as one of the

most crucial factors in economic development of each country, especially for

developing countries, including Vietnam On a macro level, education helps individuals

in society acquire better knowledge and skills and remains a basic way to accumulate

human capital, thereby having a huge impact on economic growth of each nation On a

micro level, investment in education is considered as the main road to increase income,

eliminate hunger and reduce poverty for households Another reason that should be

mentioned is social status; highly educated people in general are always respected and

appreciated by people in a society; therefore, education has been developing in both

quantity and quality In Vietnam, the education right is stipulated in the Constitution,

the Education Law and other legal documents of the State The 12th National

Congress of the Communist Party of Viet Nam emphasized the significance of

education, and considered education to be the top national policy and proposed to

"complete the national education system in the direction of open education, lifelong

learning and building a learning society” (Political report of the 11th Central

Committee of the Communist Party of Vietnam presented at the 12th National

Congress in 2016)

In the past few years, Vietnam has achieved significant achievements in

socio-economic development, witnessed more than two decades of impressive socio-economic

growth and created practically unprecedented change with GDP ratio, increasing by 7%

per year on an average Although the world has been facing financial crises, the general

economic development trend in Vietnam is still very positive In recent years, the poverty

reduction in Vietnam has reached its ideal stage: the poverty rate decreases from 57% in

1990 to about 13.5% in 2014 (Ministry of Labor, Invalids and Social Affairs, 2016)

However, the achievements in growth and poverty reduction in general do not

necessarily reflect the essence of all issues in the society Attention paid to equality and

integration is commonly observed in developing countries Therefore, it is not unusual

that one of its consequences puts more pressure on the education system For many

Vietnamese, the safest path to achieve higher social status and salary is education Aside

from high demand for education and training of the society, the knowledge economy

currently developing under the impact of globalization and specifically the impact of

accession to the World Trade Organization (WTO) also shows thirst for that

Vietnam has made great efforts to address some of these growing pressures The

government has shown strong commitments and considerable efforts to expand access

to education opportunities for all walks of life, especially children This commitment

has been proven through the remarkable progress in terms of educational background

2 since the early 1990s According to Viet Nam Household Living Standards Survey (VHLSS) in the period 2010 - 2016, the percentage of the population aged 25 - 55 years old who have no educational background reduced to less than 1% These achievements

at mainly at the primary and secondary levels of education Enrollment rate at primary schools is nearly at the rate of compulsory education while the enrollment rates at lower and upper secondary levels are generally increasing However, in consideration of enrollment rates at all levels, the higher the level of education is, the lower the rate of enrollment becomes (mostly poor students); ability to pay for educational services of the poor is low, school fees and contributions become a great burden to poor families; therefore, poor students have little opportunity to take extra classes and must spend time helping their families to earn a living, etc., having greatly affected the quality of learning

It can be said that the economic situation of families has a substantial impact on children's access to educational services, especially high-quality education Education

of children born in low-income families will be impeded as they cannot afford to pay for education or access to high-quality education Moreover, low income and difficult finance drive children to join labor force early with their families and fail to have sufficient conditions to go to school In addition, extra classes have been known as a common phenomenon at all levels and in different forms, and methods of teaching and learning, etc., which also causes a decrease in access to educational services for children

With the purpose of analyzing factors affecting expenditure on education of Vietnamese households at all educational levels and its impact on enrollment rate, thereby providing relevant information and proposing recommendations to education policymakers, I select the research topic "Analytical models for educational indicators

in Vietnam" as my doctoral thesis

2 Thesis objectives

The overall objective of the thesis is to study analytical models for educational indicators in Vietnam, namely: household expenditure on education and its impact on enrollment rates at all levels

In addition, the thesis aims at the following specific objectives:

1 Analyze the current situation of Vietnamese household expenditure on education on the period of 2004 -2016 according to influencing factors having been identified in the theoretical model and the reality of enrollment rate

2 Research the theoretical model and propose an empirical research model to analyze the indicator of household expenditure on education and provincial-level enrollment rate

3

3 Research factors impacting the total expenditure on education of households;

factors impacting expenditure on education at higher education and secondary education

levels of households In addition, the thesis also takes into consideration different

influences of the factors on expenditure on education of households in urban and rural

areas

4 Research impacts of expenditure on education on provincial-level enrollment

rate

5 Based on the estimating results of empirical models, recommendations and

proposals will be made in the thesis

To accomplish the above research objectives, the thesis focuses on answering the

following research questions:

What is the current situation of household expenditure on education in Vietnam

in recent years?

What factors impact household expenditure on education in general and

expenditure on secondary and higher education, and how do they impact?

How do the factors impacting education spending differ between urban and rural

areas?

What factors impact the enrollment rate and how does expenditure on education

affect the enrollment rate?

3 Research subject and scope

3.1 Research subject

The research subject of the thesis is to study analytical models for educational

indicators in Vietnam, specifically indicators of household expenditure on education

and enrollment rate in provinces and cities

3.2 Research scope

The research scope of the thesis only focuses on studying two basic educational

indicators in Vietnam namely: household expenditure on education and enrollment rate

Factors impacting the total expenditure on education of households will be clarified;

household expenditure on secondary and higher education via different econometric

models; impact of expenditure on education (such as extra classes, tuition fees, etc.)

and the state budget for education to enrollment rates at all levels of provinces and

cities will also be specified

4 Research method

The thesis uses quantitative methods of research and econometric models in which

data are collected from the Household Living Standard Survey and the Statistical

Yearbook of the General Statistics Office of Vietnam (GSO) Techniques of descriptive

statistics are also used in the thesis

4

5 Thesis structure

In addition to the introduction, conclusion and recommendation, list of references, the thesis is structured into four chapters as follows:

Chapter 1: Rationale and literature review

Chapter 2: Analytical models

Chapter 3: Current situation of household expenditure on education

Chapter 4: Factors impacting household expenditure on education and impacts of

expenditure on education on enrollment rate

6 New contributions of the thesis

The thesis mainly uses quantitative methods combined with modern econometric models of high reliability to solve certain of scientific issues of great significance in both theory and practice New ideas of the thesis are as follows:

(1) The thesis clarifies the theoretical basis and provide a general review of previous studies with regard to household expenditure on education as well as points out characteristics of householders, households and household size showing impacts on household expenditure on education In addition, the thesis also assesses how education spending impacts on enrollment rates at all levels The thesis thereby has built a theoretical framework to serve as the basis for empirical models specified in subsequent parts

(2) The thesis has also used the method of descriptive statistics to clearly analyze the current situation of expenditure on education of Vietnamese households according

to statistical layers via VHLSS in the period of 2004-2016

(3) The thesis simultaneously uses a number of modern quantitative research models rarely used in previous studies, especially the use of panel data for these models On the other hand, the thesis also takes advantage of the panel data structure stratified in accordance with the selected models to analyze three forms of household expenditure on education in various aspects and levels and analyzes impacts of expenditure on education such as spending on extra classes, tuition and other expenses (clothing, books, contributions, etc.) and impacts of the state budget for education on enrollment rates It is expected that this thesis will serve as a prerequisite for subsequent studies with the approach of modern econometrics including panel data models, multi-grade panel data model, Tobit model with panel data, etc., to allow further analysis

(4) Thesis findings collected from the empirical research model will serve as a basis for proposing certain of policy recommendations related to expenditure on education in the private sector, especially in households

1

CHAPTER 1

RATIONALE AND LITERATURE REVIEW

1.1 Theories laying groundwork for the thesis

1.1.1 The theory of human capital and the role of education in human capital

development

1.1.1.1 Concept of human capital

In the dictionary of economics, capital is defined as the value of capital or

investment goods used in business to generate benefits In this sense, capital means

tangible capital According to Mincer Jacob et al (1974), just like tangible capital,

people must invest for accumulation through education and training from which each

person can have earnings

1.1.1.2 Human capital investment

a Models of parental investment in children's human capital

According to Yueh (2001), a life goes through three stages: the first stage is

when parents will invest in their child's education, the next stage is when each

individual has become an adult and is able to work to earn income a portion of which is

distributed back to their parents In the third stage, they are retirees whose income

comes from profits of the assets accumulated in stage two and a portion of their income

is from their children Thus, investment in education gives parents two benefits one of

which is directly from the distribution of future income given by their children as adults

and the other is the indirect benefit gained when their well-educated children have high

possibility of marrying high-income individual, thereby creating better income for the

family in the future

b Human capital investment via expenditure on education

Education is considered as a tool to accumulate human capital, so initial

human capital investments for education in include direct costs in the course of

learning and income lost during that process (Schultz, TW, 1961; Becker, GS,

1964) Ehrenberg and Smith (2011) argue that human capital investment requires

short-term costs and is expected to generate more profits in the future, whereby the

additional cost of human capital is divided into three types: (1) direct costs including

costs of admission, books and other costs; (2) lost income; (3) mental loss in the

course of learning

1.1.1.3 Role of human capital

Human capital plays a pivotal role in the process of economic development: (1)

they are skills created by education and training, and human capital is an element of the

production process combined with tangible capital and "raw" (unskilled) labor to create

products; (2) it is the knowledge to create creativity which is a basic element of

economic development ” (Mincer, 1981) In addition, human capital has been included

2

as an input for economic growth analysis and has shown its positive effects as tangible capital but with an increasing level

The importance and positive role of human capital has always been expressively affirmed, so most households invest in human capital through education and training for their children as education is considered as public goods as well as goods recommended for people regardless of their income

1.1.1.4 Role of education for human capital development

Becker, G.S (1964) states that education and training are the most important investment in strategies of human capital development Borjas (2005) in the study of labor economics affirms that the capacity, expertise, skills and experiences are formed and accumulated through formal training, living and working process The amount of human capital accumulated is in proportion to the capacity, amount of knowledge, skills and experience that each individual gains from the process of learning, training and working

Bui Quang Binh (2009) states that education and training equip people with knowledge, skills and experience that have already been accumulated, further supplemented and equipped people with new knowledge to meet life demands

1.1.2 Theory of household production function

The household production function is collective models of the household behaviour developed by Behrman, Pollack and Taubman (1982) According to Kutty (2008), in the household production function model, households are considered as a place to create educational, cognitive and social-emotional results for their children by applying specific inputs These inputs include schools, learning materials, after-school tutoring, housing, living environment, time and parental supervision as well as other cognitive stimuli The model assumes that household decisions are best analyzed using the household utility function model The combined household utility function maximizes benefits, and decisions on resource allocation are made by householders (Becker, 1995) Households maximize the utility from their own education and from consuming other goods due to budget constraints

1.2 Literature review of household expenditure on education

Although previous studies are conducted in different countries whose socio-economic characteristics differ greatly, conclusions on the factors affecting household expenditure on education are of many similarities, including the group of factors as follows:

1) Characteristics of householders such as: gender, age, marital status, education and ethnic group

2) Characteristics of households such as: place of residence (urban, rural), living area, household size, number of household members at school age

3 3) Economic characteristics of households such as: total income and total

expenditure of the household during the year

Group of factors regarding householder characteristics

Gender of householders: Research carried out by Patrinos & Psacharopoulos

(1997) shows that female-headed households have the increasing likelihood of children

working in Peru, others like Lloyd & Gage-Brandon (1994) and Canagarajah &

Coulombe (1998) find that female-headed households in sub-Saharan Africa and Ghana

have higher enrollment rate Studies by Aslam and Kingdon (2005), Huy Vu Quang

(2012), Donkoh and Amgiuzuno (2011) show impacts of householder gender on

expenditure on education

Educational background of householders: Studies by Psacharopoulos & Arriagada

(1989) and Kingdon (2001) show that educational background of parents will affect the

posibility of their children attending school or not However, others, such as Handa

(1996), Rosenzweig & Wolngn (1994), Lillard & Willis (1994) and Unni (1998), also

point out that impacts of parents' educational background on children vary by gender

Studies by Tilak (2002), Xiaolei Qian, and Russell Smith (2010) suggest that

educational background of householders influences expenditure on education

Age of householders: A study by Andreou (2012) shows that householders who

are over 30 years old show more concerns for education than the ones under the age of

30 Donkoh and Amgiuzuno (2011) prove that householders of old age tend to have

higher expenditure than young householders

Group of factors regarding household characteristics

Studies by Aslam and Kingdon (2005) and Andreou (2012) suggest that urban

households have higher spending on education than the ones living in rural areas

Studies by Xiaolei Qian, and Russell Smith (2010) and Andreou (2012) prove that the

number of members attending school is correlated with household expenditure on

education

Economic characteristics of households: Studies of Tilak (2002) and Andreou

(2012) also show that household income impacts household expenditure on education

1.3 Literature review of enrollment rate

Studies on a global scale show that economic conditions affect the possibility of

school attendance and education demands, typically studies carried out by Tansel

(2002), MengZhao and Paul Glewwe (2007), Mariara and Kirii (2006), Al-Samarra and

Tessa (1992) and Owen and Nerman (2011)

There has been certain evidence of factors affecting education levels, enrollment

rate and school completion in Vietnam However, most of the available evidence is

related to the effect of income on education levels in which a deep understanding of

impacts of long-term factors is presented For example, income is found to be

4 correlated with the age at which children start to go to school, the number of years they attends school, their education level and test scores

Up to now, the most complete analysis of the factors impacting school development is Glewwe's study in 2004 In his study, it is pointed out that the most important factors affecting the completion of primary education are children's age (inversely), educational attainment of parents (proportional), ethnic group (inversely) and teacher's degree (proportional)

5

CHAPTER 2

ANALYSIS MODEL 2.1 Mathematical model

Supposed that in each household, there is a ‘decision maker’ who decides how

much to spend and what is basically being done in the household Household decision

maker h have an utility function:

In which: xih (i = 1, , I): goods and services, including goods such as education,

health, consumer goods and services of the household h; ĥjh (j = 1, , J): education

conducted by the number of members j of the household h during the review period

According to Rodriguez-Gutierrez (1992), students can be considered as a production

agent, who transforms this cost into qualification, through the following educational

production function:

ĥ jh = A h k

jh 0 < k <1 The purpose of the decision maker is to maximize household benefits in which:

(i) expenditure constraint on consumer goods and services, including education, does

not exceed family income; and

(ii) constraint is imposed by the student's educational production functions

Demand for education is shown in the following maximization problem:

max U h (x ih ; ĥ jh ; E h)

∑   is the exogenous (unearned) income of the household, and is a

measure of the income received for labor market-related activities by the household

The household's spending time for income L is inversely proportional to the members'

spending time for on education

2.2 Econometrics model

Marshall function for education:

2.2.1 Tobit model

2.2.1.1 Theory of Tobit model

2.2.1.2 Conditional expectations E Y Y( | >0,X)

2.2.1.3 Effect of independent variables on E Y X( | )

2.2.2 Regression model of Panel data and multilevel panel data

) , ˆ ,

h x h E U

k

jh

jh

M

j jh M

j jh jh J

j

jh i

i

ih

h

A

h

z L w h p

x

.

ˆ

1 1

1

=

+

=

∑

=

=

=

∑

=

M

j

jh

z

1

∑

=

M

j jh

jh L w

1

( , , )

jh jh h h h

h =h y z E

6

2.2.2.1 Panel data and multilevel panel data 2.2.2.2 Regression model of multilevel panel data Comments: Some advantages of regression model of multilevel panel data:

- Consider both fixed and random effects (mixed model);

- Consider the differences in groups; reduce the unequal variance;

- Consider the influences within groups and among groups

- Not require balanced panel data both in space and time;

- Accept the case of missing observations

2.2.2.3 Panel data model Fixed effects model

The fixed effects regression model is an extended form of the classical linear regression model given by:

1 1 2 2

Y = βX + β X +v + ε

Estimation method:

There are two estimation methods

i) Least Squares Dummy Variable (LSDV) estimator, in which each dummy variable represents for each observed object of the sample

ii) Fixed effects estimator

Random effects model

The random effects model is written as

1 1 2 2

Y = βX + β X +v + ε (2.35) with i=1 N and t= 1, 2, T

Estimation method: In order that the estimation results are not biased and effective,

we can use the Feasible Generalized Least Squares (FGLS) estimator to overcome the autocorrelation noise error

2.3 Analytical framework of the thesis

2.3.1 Analytical framework for the factors affecting expenditure on education

From the research review, the impact of the independent variables on the dependent variable according to the analytical framework is as follows:

7

Figure 2.1 Analytical framework for the factors affecting expenditure on

education

2.3.2 Analytical framework for the factors affecting enrollment rates

Figure 2.2 Analytical framework for the factors affecting enrollment rates

2.4 Empirical analysis model

2.4.1 Analytical models for factors affecting Vietnamese household expenditure on

education

Due to the characteristics of the survey data which is not possible to link the

2016 data with the 2010-2014 data array, so the study is divided into two different

samples, in which sample 1 is panel data from 2010 to 2014 and sample 2 is cross data

Characteristics of householder:

Ethnic group

Occupation

Gender

Age

Marital status

Education

Household characteristics:

Household size

Place of residence

Number of members attending

schools of different types

Expenditure on education:

Household expenditure on education Expenditure on general education Expenditure on higher education

Household Economic

Characteristics:

Economic factor:

Province-level GDP; per capita income

School

Number of primary schools

Number of secondary

schools

Number of high schools

Number of classes in

primary education

Teacher

Number of teachers in

primary education

Number of teachers in

secondary education

Enrollment rate of children

Expenditure on education:

Public expenditure on education

Provincial population

8

in 2016 Accordingly, for the purpose of analyzing the factors affecting household expenditure on education in general, the thesis uses two models for two research samples

Model 1: Panel data model for analyzing the factors affecting Vietnamese

household expenditure on education from 2010 to 2014

0

p

Y =β +∑ β X +ε

Where Y is the dependent variable, calculated by the natural logarithm of the total household expenditure on education

p X

is independent variable

p

β

is the coefficient of the independent variables

Model 2: Tobit model for analyzing the factors affecting Vietnamese household

expenditure on education in 2016

Because observations of household expenditure on education show that households do not have such expenditure, the Tobit model with the following censored sample is used:

*

0 (**)

p i

Y





∑

Where: (*) in case of Y i*> and (**) in case of (**) if 0 Y i*≤ 0

Y is the dependent variable, calculated by the natural logarithm of the total household expenditure on education

X is independent variable p

βp is the coefficient of the independent variables

2.4.2 Multilevel models for analyzing the factors affecting household

expenditure on general education

In general, Vietnamese households, except for households without children of school age, have such expenditure on education for this educational level With the above reasons and the characteristics of data used in this study, we select the multi-dimensional panel data model This model gives a better solution of correlation problems in the groups included in data characteristics The specific model in this study

is defined as follows:

0

ijkl ijkl pijkl pijkl ijkl

p

∑

Where:

9

- Y is the dependent variable at the first level (household level), calculated by ijkl

the natural logarithm of expenditure on general education of the i th household (i = 1, 2

nj), in the jth commune (j = 1, 2, m k ), the k th district k (k = 1, 2, , r l ), the l th province (l

= 1, 2, …, 63)

- β0ijkl is Intercept (for the i th household of j th commune, the k th district, the l th

province)

- β0ijkl is the slope

- Error εijkl is a random factor with an average expectation of 0 and constant

variance of error

- γ0l - is a random factor at the provincial level, assuming that its average is 0, the

variance of error is constant and independent ofεijkl Similarly f 0klis the random factor

at the district level, its average is 0, the variance of error is constant and independent of

ijkl

ε v 0 jkl is a random factor at the commune level, its average is 0, the variance of the

error is constant and independent of εijkl u is a random factor at the household level, 0ijkl

its average is 0, the variance of the error is constant and independent of εijkl

- X pijkl is the p th independent variable at the first level of the i th household of j th

commune, the k th district, the l th province

2.4.3 Panel data Tobit model for analyzing the factors affecting household

expenditure on higher education

In order to estimate the Vietnamese household expenditure on higher education

for the period of 2010 - 2014, the study uses an panel data Tobit model with the

censored sample (households without expenditure on education is censored)

The specific model in this study is defined as follows:

*

0 (**)

p it

Y





∑

Where: (*) in case of *

0

it

Y > and (**) in case of *

0

it

Y is the dependent variable, calculated by the natural logarithm of the total

household expenditure on education

X p is independent variable

βp is the coefficient of the independent variables

2.4.4 Analytical models for impacts of education expenditure on enrollment

rates

0

p

Y =β +∑ β X +ε

10

In which Y is a dependent variable calculated by the natural logarithm of the total

number of students in the province

p

X is independent variable

p

β is the coefficient of the independent variables

CHAPTER 3

CURRENT SITUATION OF HOUSEHOLD EXPENDITURE ON

EDUCATION IN VIETNAM FOR THE PERIOD OF 2010 - 2016

3.1 Overview of current Vietnamese education

Vietnam's education system is divided into 3 types: regular education, vocational

education and continuing education; with 3 educational levels: preschool education,

general education and higher education

3.2 Current situation of public expenditure on education

State budget is given priority to invest in education and training Vietnam's annual

budget expenditure on education is approximately 20%, equivalent to 5% of GDP This

is a very high level compared to many countries in the world, including countries with

much higher economic development than Vietnam

3.3 Percentage of households with school children and current situation of

household expenditure

3.3.1 Percentage of households with school children

In 2004, the percentage of households with at least one school children was

68.24% However, in the 2016 survey, the percentage of households with school

children was 52.04%, 16.2% less than the 2004 survey

3.3.2 Current situation of household expenditure

The average expenditure of Vietnamese households increased gradually from

2004 to 2016 in all expenditure quintiles Average expenditure of households increased

from VND 19821.72 thousand (2004) to VND 95178.15 thousand (2016)

Households with at least one school children at any level tended to spend more than

households without school children

3.4 Household expenditure on children's education

3.4.1 Household expenditure on children's education based on socio-economic

characteristics

3.4.1.1 Based on gender of householder and urban/rural areas

The average education expenditure of the households in 2016 was VND 4742.38

thousand Compared to 2004, the level of household expenditure had increased

significantly (an increase of 4.22 times, from VND 1121.63 thousand to VND 4742.38

thousand) In general, over the years from 2004 to 2016, we find that all households

trend to spend more and more on education

From 2004 to 2016, urban households have a much higher level of education

expenditure than rural households

3.4.1.2 Based on age group of householder

12 Householders who are less than 26 years old often spend less on education than householders whose age is older Householders in the age group from 40 to 54 years old spend more than the other groups The relationship between the householder age group and the household education expenditure is nonlinear

3.4.1.3 Based on ethnic group of householder

Households with Kinh householders often spend more on education than other households, while households with Chinese households also have higher expenditure levels, but lower than the households with Kinh householders (In 2006 and 2016, Chinese household expenditure on education was spent more than the other households) Households with ethnic minority householder have relatively low expenditure on education

3.4.1.4 Based on qualification of householder

The householders who have no qualifications provide very low expenditure on education compared to the householder completed from college or higher and this gap is increased over the years

3.4.1.5 Based on income of householder

Over the years from 2004 to 2016, it can be seen that in all income quintiles, household expenditure on education tends to increase significantly (In 2012, only the households in the highest income quintile had higher expenditure on education than that

in 2014 and lower than that in 2016) The households in the higher income quintile

often spend a greater amount on children's education

3.4.1.6 Based on region/area

The Northern midland and mountainous is the region with the lowest education expenditure in the country, followed by the Mekong Delta From 2004 to 2008, the Central Highlands had higher education expenditure than the Red River Delta region, but from 2010 to 2016, it was the opposite

In general, from 2004 to 2016, there was a gradual increase in education investment in all regions, the largest increase was the Red River Delta region (from VND 1188,219 thousand to VND 5447,582 thousand)

3.4.2 Household expenditure on education based on educational levels

Before 2008, households spent more on general education than higher education (even it was higher in 2004 and 2006), but by 2010, expenditure on higher education appeared to be increasing and the households spent more on higher education than on general education

3.4.3 Household expenditure on education based on expenditures

Among education expenditures, school fees accounted for the highest proportion and jumped since 2010 compared to previous years In addition, the results show that households spend a considerable amount of money in the education expenditure for

13 private tutoring; in 2016 the proportion of expenditure on private tutoring accounted for

18.11% of the household education expenditure structure

14

CHAPTER 4

ANALYTICAL MODELS FOR INDICATORS IMPACTING VIETNAMESE HOUSEHOLD EXPENDITURE ON EDUCATION 4.1 Factors impacting Vietnamese household expenditure on education

4.1.1 Research data

Data sources used in this chapter are extracted from the results of Vietnam Households Living Standard Survey (VHLSS) in 2010, 2012, 2014 and 2016 conducted

by General Statistics Office of Vietnam (GSO) The output data corresponding to the factors that are likely to impact household expenditure on education are presented in chapter 1 Due to the nature of the dataset, it is not possible to combine the data from

2010 to 2016 into panel data, so in this thesis, the data from 2010, 2012 and 2014 are combined into panel data for analysis, and the thesis has a separated analysis for the year

2016

4.1.2 Scale for variables

All variables in the research model are defined as follows:

Dependent variable:

(Y) TONGCHIGIAODUC: total household expenditure on education for 12 months Unit: thousand VND

(Y1) CHIPHOTHONG: Expenditure on secondary education Unit: thousand VND

Independent variables:

householder and values 0 in case of female householder

educational background, values 2 if the householder has primary education, values 3 if the householder has lower secondary education, values 4 if householder has upper secondary education, and values 5 if the household has higher or further education

being unmarried and values 0 in other cases

being widowed and values 0 in other cases

being divorced and values 0 in other cases

being separated and values 0 in other cases

variable values 1 if the householder is a salaried employee, and values 0 in other cases

and fishery, the dummy variable values 1 if the householder works in the fields of agriculture,

forestry and fishery, and values 0 in other cases

the dummy variable values 1 if the householder works in the service & business sector, and

values 0 in other cases

thousand VND

person

living in urban areas and values 0 in case of living in rural areas

values 1 if the household has at least 01 member receiving education grant, and values 0 in

other cases

studying at public schools Continuous variable Unit: person

studying at independent schools Continuous variable Unit: person

private schools Continuous variable Unit: person

in case of living in the Northern midland and mountainous region, and values 0 in other cases

values 1 in case of living in the North Central Coast and South Central Coast, and values 0 in

other cases

Central Highlands, and values 0 in other cases

Southeastern region, and values 0 in other cases

Mekong Delta, and values 0 in other cases

if the household is at the 2nd income level, values = 3 if the household is at the 3rd income

level, values = 4 if household is at the 4th income level, and values = 5 if household is at the

5th income level

4.1.2 Statistics and description of sample variables

4.1.2.1 Dependent variables

4.1.2.2 Independent variables

16

a Householder gender

b Educational background of householders

c Householder age

d Marital status of householders

e Ethnic group of householders

f Householder occupation

g Total household expenditure for 12 months

h Household size

i Place of residence in urban/rural area

j Education grant

k Total number of members attending schools of different types

l Living area

4.1.3 Analytical models for factors impacting Vietnamese household expenditure on education

First, the thesis conducts model estimation with the first sample (panel data from

2010 to 2014) Before estimating the model results, the thesis conducts necessary tests such as testing the multicollinearity phenomenon, autocorrelation phenomenon and changes in error variance To overcome defects in the aforementioned model, the thesis uses the weighted least squares method to estimate For the 2016 data sample, the thesis uses the Tobit model for estimation

Householder gender has influences on household expenditure, and the findings suggest that male-headed households tend to spend less on education than female-headed households

Education attainment of householders is one of the crucial factors impacting household expenditure on education; the higher education attainment the householder has, the higher expenditure on education will be spent, and the householders themselves also expect all members to achieve good academic results, whereas when householders have low education attainment or no degree, they often overlook education and education investment for family members while paying attention and spending time and budget on other expenses

Coefficient of the householder age variable is positive, which indicates that the higher the householder age is, the higher expenditure households tend to spend on education, but the coefficient of the squared householder age variable is negative, which can be explained that when householder age increases to a certain level, their household expenditure on education decreases

Marital status of householders is statistically significant in model 1 Results of regression estimate for panel data shows (married householder whose spouse lives with him/her is used as a reference group) provided that other factors remain unchanged, unmarried householders tend to have the lowest expenditure on education, less than