(Luận văn thạc sĩ) spatial analysis of income sources at province level in vietnam

VIETNAM NATIONAL UNIVERSITY, HANOI
VIETNAM JAPAN UNIVERSITY

NGUYEN THU HANG

SPATIAL ANALYSIS OF INCOME SOURCES
AT PROVINCE LEVEL IN VIETNAM

MAJOR: MASTER’S PROGRAM OF PUBLIC POLICY
CODE: …………………

RESEARCH SUPERVISOR:
Prof. MORITO TSUTSUMI

Hanoi, 2019


TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION .................................................................................... 1
1.1

Background of the study ..................................................................................... 1

1.2

Rationale of the study ......................................................................................... 2

1.3

Objectives of the study ....................................................................................... 4

1.4

Research questions ............................................................................................. 4

1.5

Significance of the study .................................................................................... 4

1.6

Design of the study ............................................................................................. 5

CHAPTER 2: LITERATURE REVIEW ......................................................................... 6
2.1

Spatial analysis ................................................................................................... 6

2.2

Income as an aspect of livelihoods ..................................................................... 7

2.3

Background of ethnicity and income structure in Vietnam ................................ 8

2.3.1

Ethnic geographical distribution in Vietnam ............................................... 8

2.3.2

Poverty distribution by ethnicity in Vietnam............................................... 9

2.3.3

Changes in Vietnam‟s income structure in Vietnam ................................. 10

2.4

Previous studies ................................................................................................ 11

CHAPTER 3: METHOD AND METHODOLOGY ..................................................... 14
3.1. Method and methodology ................................................................................. 14
3.2. Data collection .................................................................................................. 17


CHAPTER 4: FINDINGS AND DISCUSSIONS ......................................................... 19
4.1

Area of Study .................................................................................................... 19

4.1.1

An overview ............................................................................................... 19

4.1.2

Economic growth ....................................................................................... 21

4.1.3 Production of agriculture, forestry and fishery .............................................. 22

4.1.3

Industry ...................................................................................................... 23

4.1.4

Service activities ........................................................................................ 24

4.1.5

Development investment ........................................................................... 24

4.2

Descriptive statistics ......................................................................................... 25

4.3

Changes in income sources in Vietnam 2008-2016 ......................................... 77

4.4

Discussions ....................................................................................................... 82

CHAPTER 5: CONCLUSION AND RECOMMENATIONS ...................................... 85
5.1

Conclusion ........................................................................................................ 85

5.3. Limitations ........................................................................................................ 87

5.4. Suggestions for the further studies ................................................................... 88
REFERENCES ............................................................................................................... 89


LIST OF TABLES
Table 1: A comparison of Spatial regression models and OLS regression model......... 27
Table 2: A comparison of Spatial regression models and OLS regression model. Year:
2008 Dependent variable: Income from Agric (million VND) ..................................... 30
Table 3: A comparison of Spatial regression models and OLS regression model. Year:
2008 Dependent variable: Income from Nonagric (million VND) ................................ 32
Table 4: A comparison of Spatial regression models and OLS regression model Year:
2008 Dependent variable: Income from Wages (million VND) .................................... 34
Table 5: A comparison of Spatial regression models and OLS regression model. Year:
2008 Dependent variable: Income from Other sources (million VND) ........................ 35
Table 6: A comparison of Spatial regression models and OLS regression model. Year:
2010 Dependent variable: Total income (million VND) ............................................... 37
Table 7: A comparison of Spatial regression models and OLS regression model Year:
2010 Dependent variable: Income from Agric (million VND) ..................................... 39
Table 8: A comparison of Spatial regression models and OLS regression model. Year:
2010 Dependent variable: Income from NonAgric (million VND) ............................... 41
Table 9: A comparison of Spatial regression models and OLS regression model Year:
2010 Dependent variable: Income from Wages (million VND) .................................... 43
Table 10: A comparison of Spatial regression models and OLS regression model. Year:
2010. Dependent variable: Income from other sources (million VND) ........................ 45
Table 11: A comparison of Spatial regression models and OLS regression model. Year:
2012. Dependent variable: Total Income (million VND) .............................................. 48


Table 12: A comparison of Spatial regression models and OLS regression model.

:

2012. Dependent variable: Income from Agric (million VND) .................................... 50
Table 13: A comparison of Spatial regression models and OLS regression model Year:
2012 Dependent variable: Income from Nonagric (million VND) ................................ 52
Table 14: A comparison of Spatial regression models and OLS regression model. Year:
2012 Dependent variable: Income from Wages (million VND) .................................... 54
Table 15: A comparison of Spatial regression models and OLS regression model Year:
2012 Dependent variable: Income from other sources (million VND) ......................... 56
Table 16: A comparison of Spatial regression models and OLS regression model Year:
2014 Dependent variable: Total Income (million VND) ............................................... 58
Table 17: A comparison of Spatial regression models and OLS regression model Year:
2014 Dependent variable: Income from agric (million VND) ....................................... 60
Table 18: A comparison of Spatial regression models and OLS regression model Year:
2014 Dependent variable: Income from Nonagric (million VND) ................................ 62
Table 19: A comparison of Spatial regression models and OLS regression model Year:
2014 Dependent variable: Income from wages (million VND) ..................................... 64
Table 20: A comparison of Spatial regression models and OLS regression model Year:
2014 Dependent variable: Income from other sources (million VND) ......................... 66
Table 21: A comparison of Spatial regression models and OLS regression model Year:
2016 Dependent variable: Total Income (million VND) ............................................... 68
Table 22: A comparison of Spatial regression models and OLS regression model Year:
2016 Dependent variable: Income from Agric (million VND) ..................................... 70


Table 23: A comparison of Spatial regression models and OLS regression model Year:
2016 Dependent variable: Income from NonAgric (million VND) ............................... 72
Table 24: A comparison of Spatial regression models and OLS regression model Year:
2016 Dependent variable: Income from wages (million VND) ..................................... 73
Table 25: A comparison of Spatial regression models and OLS regression model. Year:

2016 Dependent variable: Income from other sources (million VND) ......................... 75

LIST OF FIGURES
Figure 1: The changing rate of total income between 2008 and 2016 ........................... 77
Figure 2: The changing rate of income from wages between 2008 and 2016 ............... 78
Figure 3: The changing rate of income from Agriculture, Forestry, Fishery between
2008 and 2016 ................................................................................................................ 79
Figure 4: The changing rate of income from Non-Agriculture, Non-Forestry, Nonfishery between 2008 and 2016...................................................................................... 80
Figure 5: The changing rate of income from other sources between 2008 and 2016 .... 81


ABBREVIATION
SDM

Spatial Durbin model

SLM

Spatial Lag Model

SEM

Spatial Error Model

OLS

Ordinary Least Square

GIS

Geographic Information system

GDP

Gross Domestic Product

CPI

Consumer Price Index

ASEAN

Association of South-East Asian Nations

WTO

World Trade Organization

HDI

human development index

UNDP

United Nations Development Programme

Agric

Income from Agricultural, Forestry and Fishery activities

NonAgric

Income from Non- Agricultural, Non-Forestry, Non-Fishery activities

Other

Income from other sources

Wages

Income from Wages

Total

The total income

VHLSS

Vietnam Household living standards survey


ACKNOWLEDGEMENT
In order to complete my thesis, I have received many advices and guidance from my
supervisor - Professor Morito Tsutsumi as well as my friend Rim Er-rbib. Thank to
professor Morito Tsitsumi, I can acquire more knowledge and more skills. Before
coming back to VietNam, my supervisor gave me a valuable book that helps me a lot to
complete this thesis. With all my respect and gratitude, I would like to express my
sincere appreciation to:
My supervisor, Professor Morito Tsutsumi for his inspiring guidance and great support
throughout my thesis procedure. His insightful advices and scientific knowledge has

inspired me and helped me in improving research and preparation for my Master thesis.
He also supported me a lot to get the data of FDI licensed projects which seemed really
hard to acquire. Without his great support, I cannot finish my thesis.
My academic tutor, Ms. Rim Er-rbib, for her useful support and encouragement, who is
always willing to help and gave me so many useful and constructive instructions
especially for how to use GIS software.
University of Tsukuba and Vietnam Japan University for giving me such a excellent
environment with so many amazing people.
Finally, I would like to thank my family for being a wonderful moral support that gives
me so much motivation and enthusiasm to overcome the challenges and difficulties in
writing this thesis.

Student,
Nguyen Thu Hang.


CHAPTER 1: INTRODUCTION
1.1 Background of the study
Vietnam has been through a rapid economic growth in the last three decades. The
characteristics of this rapid growth are the decline of the number living in poverty and
the rising average income. Since the 1990s, there has been nearly 30 million people
overcoming the poverty line. More specifically, the GDP per capita from 1990 to 2015
has increased from $100 to $2,300, respectively (Oxfam, 2017). In the last 30 years,
the average of the economic growth has increased from 5-6 percent to 6.4 percent. The
rapid growth especially the increasing economic has several impacts on the
Vietnamese. On the one hand, it improves people‟s living standards. However, it also
causes the economic inequality as well as the uneven opportunity among people.
Which means the equal distribution of income of the people has an important role in a
society with high equality. So now the challenge is that in the situation of the rapid
economic growth how does Vietnam make solutions so that the distribution of income

across Vietnam becomes much more equal.
The rapid economic growth and the good policies in the last 30 years have significant
contribute to poverty reduction. However, the gap between the rich and the poor has
been expanding seriously. This gap has been causing many social problems and need to
be solved as soon as possible. So the Government need to issue new policies that
ensure the poverty and the inequality will be controlled. According to Saumik et al
(2016), while Vietnam experiencing the economic structural transformation as well as
the poverty reduction, the growth is more beneficial for the rich than the poor. This is
realized as the returns to manufacturing and to agriculture increasing only for the top
10th - 20th percentiles. In general, the economic inequality has been rising dramatically
in the last twenty years.

1


According to Oxfam (2017), in one day, the Vietnamese richest man earn more than
the poorest earns in 10 years. This man possessed assets worth $2.3bn which could be
used to help 13 million poor people to get out of poverty. According to the World
Bank (2013), from 1992 to 2012, the Gini index has risen from 35.7 to 38.7, showing
that the income inequality rose. However, this kind of data may underestimate the
serious impacts that inequality can have on Vietnam. For example, the expenditures or
the income of rich individuals may be under-reported in the household surveys, so the
empirical measures of inequality may be biased.
Since 2004, among the first four quintiles (the bottom 80 percent) there is a small
difference in the income distribution. However, in comparison between those quintiles
with the richest quintiles (the top 20 percent), the income distribution has been
widening significantly. In other words, the benefit of growth has been distributed
unequally in recent years. This is consistent with the report conducted by Oxfam in
2016. The survey did depth-interview with 600 respondents from three provinces (Lao
Cai, Nghe An, Dak Nong). The results showed that the income of the 20 percent of the

richest households is 21 times higher than that of the 20 percent of the poorest
households.There is one point suggesting that income at the province level is serious
and has been increasing over time, especially in the remote areas where agriculture is
the main source of income (Lam et al., 2016). Therefore, it is necessary to look into the
income sources at province level to justify the income disparity.
1.2 Rationale of the study
It is revealed by the evidences in the research by Nguyen (2016) that reductions in
poverty and dividends from growth have been spread unevenly across Vietnam,
increasing income inequality between regions and to some extent within regions. By
region, the Red River Delta and the South East are considerably overrepresented in
middle income groups, whereas the Mekong River Delta is overrepresented in the nearpoor group. The North West and Central Highlands are the two regions where most of

2


the poor live. According to VHLSS (2012), the South East has the highest monthly
income per capita in the country (VND3,016,000 or $150), which is more than three
times the average monthly income found in the North West region (VND999,000 or
less than $50).
Using VHLSS data (2004–2014), the findings by McCaig &Brandt (2015) show that
households in the South East (the richest region in Vietnam) have the highest income
mobility of any region. Compared with households in the Red River Delta (the
reference group), households in the North East, South Central Coast, and Central
Highlands are less likely to move up from the lowest quintile. Households in the South
East are more likely to move up from the lower 40 percent. With downward mobility,
households in the North Central Coast and Central Highlands are more likely to move
down from the high-income quintiles.
Such regional variation is also the product of ethnic factors in Vietnam (McCaig
&Brandt, 2015). Vietnam is an ethnically diverse country: there are 54 ethnic groups,
in which the Kinh majority accounts for 85 percent of the population. Kinh tend to live

in delta areas, and have higher living standards than other ethnic minorities. Hoa
(Chinese) are also a rich group, and also live in delta areas. Thus, Hoa are often
grouped together with Kinh in studies on household welfare, although they may face
ethnic discrimination in other areas.
Income poverty is disproportionately higher among ethnic minority groups. Members
of ethnic minority groups make up less than 15 percent of the country‟s population but
account for 70 percent of the extreme poor. According to the 2014 survey conducted by
the Ministry of Labor, Invalids and Social Affairs, the incidence of poverty among
ethnic minorities was as high as 46.6 percent, compared to 9.9 percent for the Kinh and
Hoa groups. The gap in income mobility among ethnic groups is also large, and there
are signs that this gap has been increasing over time. Between 2010 and 2014, around
19 percent of ethnic minorities in the bottom quintile moved to a higher income

3


quintile, while for Kinh and Hoa, this figure was 49 percent. In addition, ethnic
minorities are more likely to move down but less likely tomove up, compared with
Kinh and Hoa. It is revealed that both the absolute and relative income gap between
Kinh/Hoa and other ethnic groups has increased over time. The ratio of per capita
income of Kinh/Hoa to that of other ethnic groups increased from 2.1 in 2004 to 2.3 in
2014.
The income disparity sourced from the ethnic and regional differences has led to the
income inequality at the provincial level. Therefore, it is meaningful to analyze the
income sources at province level from 2008 – 2016 and how various factors affect
them by using spatial analysis.
1.3 Objectives of the study
The overarching aims governing this current study is to obtain the insights into the
current income distribution and to reduce the income disparity in Vietnam at the
provincial level. Therefore, the thesis‟s objective is to promoting income

diversification by examining what economic and demographic variables affect the
income sources among provinces in Viet Nam using spatial approach.
1.4 Research questions
The following research questions are derived in this current study:
(1) Is there the presence of spatial autocorrelation of income sources among provinces in
Viet Nam?
(2) How do economic and demographic factors affect the income sources in Viet Nam?
1.5 Significance of the study
This study has several contributions to the literature. Firstly, it is the first to identify the
composition of sources of income in Vietnam and the contribution of various income
sources to total income inequality with reference to the use of spatial analysis.

4


Secondly, this study provides an analysis of long-term changes in income sources in
Vietnam in the last ten years from 2006 to 2016. This research expects to provide the
income inequality decomposition by income sources, based on the Vietnam Household
Living Standard Surveys (VHLSS) carried out every two years, to reduce the errors
that resulted from data aggregation process. Lastly, the recommendations generated in
this current study expect to make contributions to policy development to diversify the
sources of income, contributing to minimize the income inequality in Vietnam.
1.6 Design of the study
There are five chapter included in this current study, including:
Chapter 1 – Introduction – presents the background and rationales of the current study.
The research aims and objectives, research questions and design of the study are also
generated in this chapter.
Chapter 2 – Literature review – critically explores the theoretical fundamentals
concerning the spatial analysis and income inequality and sources. This chapter also
looks in the previous literatures to identify the literature gaps.

Chapter 3 – Methodology - presents research methodology. The research method, data
collection measures and how such models as Spatial Durbin Model (SDM, Spatial lag
model (SLM), Spatial error model (SEM) are used for data analysis. This chapter also
discusses the validity and reliability of the research instruments.
Chapter 4 – Findings and discussions – shows the results of data analysis and discusses
the income sources of Vietnam with the provincial levels.
Chapter 5 – Conclusion and recommendations – summarizes the whole study and
research findings. In addition, limitation of the thesis and suggestions for further
research are also given out.

5


CHAPTER 2: LITERATURE REVIEW
2.1 Spatial analysis
Spatial analysis or spatial statistics includes any of the formal techniques which study
entities using their topological, geometric, or geographic properties. Spatial analysis
includes a variety of techniques, many still in their early development, using different
analytic approaches and applied in fields as diverse as astronomy, with its studies of
the placement of galaxies in the cosmos, to chip fabrication engineering, with its use of
"place and route" algorithms to build complex wiring structures. In a more restricted
sense, spatial analysis is the technique applied to structures at the human scale, most
notably in the analysis of geographic data.
Complex issues arise in spatial analysis, many of which are neither clearly defined nor
completely resolved, but form the basis for current research. The most fundamental of
these is the problem of defining the spatial location of the entities being studied.
More specifically, spatial autocorrelation can be defined as the coincidence of value
similarity with location similarity (Anselin, 2003). There is positive spatial
autocorrelation when high or low values of a random variable tend to cluster in space
and there is negative spatial autocorrelation when geographical areas tend to be

surrounded by neighbors with very dissimilar values. There are at least three possible
explanations. One reason is that there is a simple spatial correlation relationship,
showing what is causing an observation in one location also causes similar
observations in nearby locations. Another possibility is spatial causality, meaning that
something at a given location directly influences it in nearby locations. A third
explanation is spatial interaction: the movement of people, goods or information
creates apparent relationships between locations.

6


Spatial heterogeneity means in turn that economic behavior is not stable across space
and may generate characteristic spatial patterns of economic development under the
form of spatial regimes: a cluster of forward States (rich regions, the core) being
distinguished from a cluster of backward States (poor regions, the periphery). The
methodology of exploratory spatial data analysis (ESDA) is applied to find the
evidence of spatial autocorrelation and spatial heterogeneity. The estimation of global
spatial autocorrelation (Moran‟s I) and local spatial autocorrelation (LISA) will
indicate how economic activities are located in India during the reform period 1993–
2004. Moreover, local spatial statistics confirms the existence of spatial heterogeneity
and, consequently, raises an agenda behind the differential growth profile of forward
States and backward States.
2.2 Income as an aspect of livelihoods
Although income from agricultural activities is the base of livelihood strategies for
rural households in developing countries (Ashley, 2000; Dolan, 2004; Ellis, 1999;
Sandbrook, 2006), empirical evidence suggests that households regularly engage in
nonagricultural activities as a source of income (Ashley, 2000; Ellis, 1998, 1999;
Hartter, 2007; Kaag et al., 2008; Lepper and Schroenn Goebel, 2010; Smith et al.,
2001). It has been found that high levels of non-agricultural income are often
associated with higher levels of agricultural productivity and higher overall household

income (Dolan, 2004; Ellis, 1999; Ellis and Bahiigwa, 2003; Evans and Ngau, 1991).
According to Ellis (1999: 2) livelihoods are defined as „the activities, the assets and the
access that jointly determine the living gained by an individual or household‟. The
concept of livelihoods was originally coined in the 1940s to describe people‟s
strategies of making a living (Kaag et al., 2008). However, with the macroeconomic
development literature of the 1980s and the pioneering literature of Robert Chambers,
the livelihood approach transformed into its current meaning by including the complex

7


dimension of poverty (Ahebwa, 2012; Brocklesby and Fisher, 2003; de Haan and
Zoomers, 2005; Kaag et al., 2008).
In Vietnam, the majority of rural dwellers secure their livelihoods primarily through
small-scale subsistence agriculture (Nguyen, 2004; Lam et al., 2016). Most Vietnamese
households are dependent on small-scale agriculture activities for their earnings (GSO,
2016). Vietnamese households use their goods from their agricultural surpluses to sell
in the local markets for cash generation. Some other affluent households invest in large
scale crops or livestock farming for commercial purposes. Livestock is also used as
savings for covering difficult periods or extraordinary expenditures for celebrations or
holidays, such as the payment of dowries. However, it is indicated by the previous
studies in the sources of income in Vietnam that it is critical to expand the horizons of
sources of income with the focus of non-agricultural sector (Nguyen, 2004; Hartter,
2007; Mackenzie, 2011; Lam et al., 2016). Drawing on the country‟s abundant
resources, the Vietnamese government is stimulating the development of nonagriculture sectors to diversify the sources of incomes.
2.3 Background of ethnicity and income structure in Vietnam
2.3.1 Ethnic geographical distribution in Vietnam
Vietnam has 54 officially recognized ethnic groups, with more than 85% of the
population made up of Kinh people. The rest of the population, 15%, is distributed
among 53 ethnic minorities. Most of these ethnic groups, however, have a few

thousand people each. According to the General Statstisitics Office Vietnam (GSO,
2015), of the ethnic minority group, the most numerous are the Tay (1.9%), Thai
(1.8%), Muong (1.5%), Kho Me (1.5%), H‟Mong (1.2%) and Nung (1.1%). Most
ethnic minority groups reside in mountainous areas, while the Kinh and Chinese are
found in the lowland areas in Red River delta, Central Coast and Mekong Delta. By

8


comparison, the minority groups are primarily located in the East and West Northern
mountains, in the Central Highlands, and in the North Central Coast.
2.3.2 Poverty distribution by ethnicity in Vietnam
Since the economic reform introduced in 1986, known as Doi Moi, both majority and
minority ethnic groups have experienced an improvement in living standards, which
has been reflected in increasing average expenditure per capita, falling fertility rate and
household size, and declining in the level of malnutrition (Epprecht, Müller, & Minot,
2011). However, Vietnam‟s ethnic minority groups lagged behind the Kinh ethnic
majority. Initially, early in the last decade, the ethnic minority groups achieved a
significant success in poverty reduction, e.g. poverty rates fell from 75.2% in 1998 to
50.3% in 2008. Nevertheless, ethnic minorities have increasingly accounted for most of
the poor in Vietnam. Although they contributed only 15% of Vietnam‟s total
population, ethnic minorities accounted for about half of the poor and 68% of the
extremely poor (Kozel, 2014). Poverty rates among ethnic minorities average between
four and seven times higher than that of the Kinh people. The malnutrition rate of
children from ethnic minority households is also considerably higher than among
children from ethnic majority households. Vietnam‟s poverty map shows that the
majority of the poor live in the upland regions, whereas the better off households are
found in Vietnam‟s urban centres along the coast. There existed an increasing disparity
between the ethnic majority and ethnic minorities among income percentiles in
Vietnam from 1998 to 2010. In 1993, the ethnic minority was 1.6 times poorer than the

ethnic majority. This gap increased to 2.4 times in 1998, 4.5 times in 2004 and 5.1
times in 2010. The proportion of the poor from Vietnam‟s ethnic minorities in 2010
was considerably higher than in 1998.

9


2.3.3 Changes in Vietnam’s income structure in Vietnam
Income structure in Vietnam has changed over time. The proportion of income from
agriculture has declined, while wage income has contributed to an increasing share of
total household income in 2000s as well as in the previous decade. In rural areas, crop
income and agricultural side-line income remained two main sources of household
income, but together they contributed one third of total household income for top ten
percentile income households. However, income from cultivation declined sharply by
half compared with its level a decade ago (Benjamin et al., 2017; McCaig, Benjamin,
& Brandt, 2009). The proportion of income from wages in rural areas increased faster
than in urban areas.
The share of wage income of the bottom-income household group increased faster than
that of the top-income households. In the meantime, in urban areas, changes in income
structure have not been as fast as in rural areas in 2000s. However, wages had already
become the main income source of urban households since the 1990s. The share of
agricultural side-line income in total household income has remained stable at a small
share in urban areas during the 2000s. The top income quartile households experienced
a faster increase in income than the other quartiles. The income share from remittances
and other income sources in 2000s has moderately decreased compared to the 1990s.
There was also a shift in the employment structure among ethnic minorities toward
wages in nonfarm employment and nonfarm self-employment in the early 2000s (Pham
& Bui, 2010). However, the ethnic minorities still received a smaller amount of their
income from non-agricultural wages and nonfarm businesses. In the meantime, the
ethnic majority received a higher portion of their income from wages (Cuong, 2012;

Dang, 2012; Kozel, 2014).
The main income source for the ethnic majority was from wage employment, whereas
for the ethnic minority, the main source was crop income. Poorer ethnic minority
households had a larger proportion of their total income from crops (Cuong, 2012). In

10


terms of employment, in 2006 agriculture accounted for 30% of ethnic majority
employment, but made up 55% of ethnic minority employment (Kozel, 2014). There
was a significant rise in income share from wages, while the level of income from the
agricultural sector has declined. However, the change toward wage-earning
employment of ethnic minorities was slower than those of the ethnic majority. There
are several studies on income inequality between ethnicities in Vietnam (Benjamin et
al, 2017; Kozel, 2014; Cuong, 2012; Baulch, Pham, and Reilly, 2012; Baulch, 2011;
Epprecht et al. 2011; World Bank, 2009; Van de Walle and Gunewardena, 2001).
However, most of them focused on various characteristics to explain the widening
income or income inequality gap. Although ethnic minorities have made significant
progress in improving living standards, health and education in recent years, this group
still lag behind the ethnic majority in terms of household per capita expenditure and
income. The absolute gap between the ethnic majority and ethnic minorities widened
dramatically in the 2000s (Benjamin et al., 2017). The main causes of the disparity
between the ethnic groups are differences in educational attainment, residential area,
accessibility to public services and household assets (Cuong, 2012; Dang, 2012;
Tuyen, 2016; van de Walle & Gunewardena, 2001; World Bank, 2009). Furthermore,
Benjamin et al. (2017) and Cuong (2012) find that the main contributors to the
widening income gap are the ethnic minority‟s lower wages and lower non-farm
business income. In addition, the income structure of the ethnic majority people has
shifted from the agricultural sector to non-agricultural sectors more quickly than that of
the ethnic minority. This income source disparity is also the drivers of the larger

income gap between ethnic minority groups (Cuong, 2012).
2.4 Previous studies
Concerns about increasing income inequality with reference to the sources of income
have become the areas of focus in many researches which investigating the income
issues in the world and Vietnam. Dabla-Norris et al. (2015) have investigated the

11


factors influencing the increasing inequality worldwide. It is realized that the issue of
income equality under the effects of sources of income has become alarming in not
only such developed countries as the US, European countries, Japan and Korea, etc. but
in developing and poor nations as well (Dabla-Norris et al., 2015; Furrer, 2016).
The findings by Milanovic (2013) have revealed that the effects generated by the trend
of globalization provide benefits for those with middle and high income levels rather
than those with the low level. By using the data obtained between 1988 and 2008, he
concluded that while those who have the top 1% income experienced a 70% increase in
their income over the given period, their poor counterparts hardly enjoyed any increase
in their income. Oxfam (2017) emphasized that the top 1% rich people are those who
own the majority of global wealth. These findings are also supported by the researches
concerning the expanding income inequality in such others countries in BRICS by
Berg (2015) and Haldane et al. (2015). Additionally, these researches identified that
among the most powerful factors influencing the income inequality, the source of
income have significant impacts on income inequality
For the past decades, owing to the stable and skyrocketing economic development
Vietnam has experienced the significant increase in the amount of average income per
capita, contributing to the poverty reduction. However, the studies by McCaig et al.
(2009) and Kozel (2014) indicates that despite the economic development and income
increase, the income gap in Vietnam has been continuously widened. These scholars
provide the evidences with different measures to indicate that there is a rise in the

absolute income gap between the top and bottom income groups in Vietnam. Oxfam
(2017) also reported that the daily income of the top 1% is at least ten times more than
the annual income of the top 5% bottom in Vietnam. The difference between the
income level of rural and urban households has also witnessed the same pattern.
Consequently, the income disparity has remained as one of the most problematic issues
in Vietnam. Kozel (2014) also attempts to prove that the increasing income gap across
the country is significantly attributed to the sources of income which are different
province by province and region by region in Vietnam. Therefore, it is critical that the

12


policymakers and scholars look into income sources as a significant and meaningful
factor to the income inequality in Vietnam.
In Vietnam, the income gap is regarded as one of the most challenging barriers to the
attempts to obtain the sustainable development of the Government. Despite the
reduction of the poverty rate to less than 10%, the income inequality has still lowered
the progress as the whole (Kozel, 2014). It is planned by the Government that the
development policies will target to earn a 2% decrease per year in the poverty rate
(Gibson, 2016). Dealing with the inequality requires the investigation into sources of
income in Vietnam.
According to Abdulai, A., & CroleRees, A. (2001): “the income of agricultural
households is affected by various factors such as land, the level of education, the
number of labors”. The research titled Effect of Resources on Incomes of Agricultural
Households in Thanh Hoa Province: A Case Study at Tho Xuan and Ha Trung Districts
by Chu Thi Kim Loan & Nguyen Van Huong (2015) also points out that sources like
the scale of production (lands, farms), the The research by Nguyen & Tran (2018)
concerning the effects of various income sources on income inequality also points out
that. They also revealed that among the sources of income wages and non-agriculture
incomes are the most influencing drivers of income gap in Vietnam. Their counterpart

from agricultural activities were also relatively evenly distributed. The research
findings also imply significant changes in the structure of incomes in Vietnam with a
shift from agriculture reliance to non-agriculture reliance economy. Therefore, it can be
concluded that the income sources have significant impacts on the income equality.
However, in all the studies concerning this issue there has been no significant research
concerning the spatial analysis of income sources in Viet Nam at province level
especially after Viet Nam‟s signing WTO in 2006. This study will focus on the income
sources based on different economic and demographic variables at the province level
and will explain the influence of those variables on the income sources using these
variables.

13


CHAPTER 3: METHOD AND METHODOLOGY
3.1.

Method and methodology

The major difference between spatial econometrics and standard econometrics is that
spatial econometrics requires diffrent sets of information. It relates to the observed
values of the variables and it also relates to the particular location where the variables
are observed. This means spatial regression takes into account the spatial correlation.
This study uses Moran I‟s test to test the presence of spatial autocorrelation of the
income sources among provinces. If this index is significant at 5% then applying
spatial model is necessary. The Moran I‟s test takes the form like this:
𝑛

𝑛
𝑛

𝑖=1 𝑗 =1 𝑤𝑖𝑗 𝑋𝑖 − 𝑋
𝑛
𝑛
𝑛
𝑖=1 𝐽 =1 𝑤𝑖𝑗
𝑖=1

𝑋𝑗 − 𝑋
𝑋𝑖 − 𝑋

With the hypothesis:
Ho: no spatial correlation among provinces
H1: there is spatial correlation among provinces
Where:
Xi : Observed variable at the province i
Xj : Observed variable at the province j
𝑋 : Average variable of X
n: observations
Wij : spatial weight matrix between two provinces.

14


In this thesis, spatial weight matrix (Spatial contiguity weights) indicating whether
spatial units share a boundary or not is used to summarize the spatial relation among 63
spatial units (provinces). The spatial weight matrix contains 63 columns and 63 rows
associated with 63 provinces in Viet Nam:
𝑤11
⋮
n𝑤𝑛 =

𝑤1𝑛

⋯
𝑤𝑖𝑗
⋯

𝑤𝑛1
⋮
𝑤𝑛𝑛

(1)

and has the standard form as following:

Wij =

1, 𝑏𝑛𝑑(𝑖)𝑏𝑛𝑑(𝑗)
0, 𝑏𝑛𝑑(𝑖)𝑏𝑛𝑑(𝑗) = 

(2)

Where:
i, j: provinces taken into consideration
bnd: boundary
The weight matrix receives the value as 1 when these two provinces share the border
and as 0 otherwise.
Besides the OLS, this paper also runs three other spatial models which are Spatial
Durbin Model, Spatial lag model and spatial error model and then compares between
them to choose the best model for analyzing.
The Spatial Durbin model takes the form:

Y = 𝝆WY+ X𝜷(1) +WX𝜷(2) + 𝜺

(3)

Where:
X: a matrix of non-stochastic regressors
β(1), β(2), ρ: parameters to be estimated

15


W: the weight matrix exogenously given
WY : the spatially lagged variable of Y
WX: the spatially lagged variable of X
The spatial lag model takes the form:
Y = 𝝆WY + 𝜷 X +u

(4)

Where:
U: stochastic disturbances
β, ρ: parameters to be estimated
W: the weight matrix exogenously given
WY: the spatially lagged variable of Y
The spatial error model takes the form:
Y = X + u

(5)

U = Wu + 

 ~ N (0, 2 In)

Where:
X: a matrix of non-stochastic regressors
U: stochastic disturbances, β: parameters to be estimated
Wu: the weight matrix of stochastic disturbances

16


Finally this paper uses the software QGIS for drawing maps to see the changing rate of
income sources across provinces as well as the main regions.
3.2.

Data collection

Dependent data: Per capita monthly income at current prices by income sources and by
provinces. (deflated by CPI with year 2008 as base year)
Per capita monthly income is calculated by dividing the household‟s total income by
the number of family members then dividing by 12 months. This income includes:
income from wages; income from agriculture, forestry and fishery (after tax); income
from non-agricultural, forestry and fishery (after tax); and the last is other income
sources such as donations, gratuities, savings interest, etc. Items which are not taken
into account in the income include savings, debt collection, asset sale, debt financing,
transfers, capital from joint ventures, associates in production and business.
Explanatory variable data
-

Immigration and migration rate:

a. Immigration rate:
The number of people from another territorial unit (original place) immigrating to a
territorial unit during the study period (usually one year) on average per 1000
inhabitants of that territorial unit.
IMR ‰ =

𝐼
× 1000
𝑃𝑡𝑏

Where: IMR: immigration rate (‰).
I: The number of immigrants.
Ptb : average population.

17