Multidimensional poverty in the southern region in vietnam in the south of vietnam

GSO: General statistics offices WB: World Bank WHO: World health organization MPI: Multidimensional poverty index VHLSS: Vietnamese Household Living Standard Survey GDP: Gross domestic p

TRUONG DAT ANH

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, OCTOBER 2016

Page | i

INSTITUTE OF SOCIAL STUDIES

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

MULTIDIMENSIONAL POVERTY IN THE SOUTHERN

REGION OF VIETNAM

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

TRUONG DAT ANH

Academic Supervisor:

DR TRAN TIEN KHAI

HO CHI MINH CITY, OCTOBER 2016

Page | ii

I assure that the entire contents of this thesis (Multidimensional poverty in the Southernregion of Vietnam) are written by me and the contents are based on my knowledge and myreview of literature This thesis has not been submitted for any degree and I am responsiblefor entire contents I wrote

TRUONG DAT ANH

Page | i

I would like to express profound thanks to A/Prof TRAN TIEN KHAI because of the dedicatedguidance and supports during thesis time The opinions and comments of supervisor play asignificant role in my thesis

Moreover, I sincerely thank Prof NGUYEN TRONG HOAI, Dr TRAN KHANH NAM, Dr TRUONGDANG THUY and lecturers – officers of Vietnam Netherland program Thanks to lecturers, I haveaccumulated a variety of knowledge to finish study modules and support for my thesis

Finally, I wish lecturers had many success and advantages in career in the future and agrowing number of research projects to contribute for the economy of Vietnam

Page | ii

GSO: General statistics offices

WB: World Bank

WHO: World health organization

MPI: Multidimensional poverty index

VHLSS: Vietnamese Household Living Standard Survey

GDP: Gross domestic product

Page | iii

This thesis researches about multidimensional poverty in rural area of the Southern region

of Vietnam The research scope is 19 provinces, separated into area the Southeast regionand the Southwest region Based on VHLSS 2014 data set of General Statistic Offices ofVietnam, 1977 households have filtered to examine Alkire & Forster approach have appliedand established a framework include 20 indicators and 6 dimensions, which are suitable forthe reality and data availability in the Southern region of Vietnam Using the methods ofChi2 test and descriptive statistics, this thesis found some relationships between householdcharacteristics and poverty status Besides, there are some differences in poverty statusamong provinces and areas Hence, poverty policies used for tackle poverty status shouldfocus to reasonable targets to ensure efficient and effective

Page | iv

TABLE CONTENTS

CHAPTER I: INTRODUCTION 1

1.1 Research problem 1

1.2 Research objective 3

1.3 Research question 3

1.4 Research hypothesis 4

1.5 Research scope 4

1.6 Organization of research 4

CHAPTER II: LITERATURE REVIEW 5

2.1 Poverty and measurement 5

2.1.1 The concept and measurement of uni-dimension poverty 5

2.1.2 Concept and measurement of Multidimensional poverty 7

2.2 Empirical studies 11

CHAPTER III: RESEARCH METHODOLOGY 15

3.1 Research methodology 15

3.2 Multidimensional poverty index 15

3.3 Conceptual framework 18

3.4 Dimensions and indicators 23

3.5 Data sources 24

CHAPTER IV: RESULTS 26

Page | v

4.1 The choice of cutoff ratio k 26

4.2 Multidimensional poverty and uni-dimensional poverty 30

4.3 Multidimensional poverty across regions 31

4.4 Multidimensional poverty across provinces 33

4.5 Dimensions contribution to multidimensional poverty index 36

4.6 Household characteristics and multidimensional poverty 38

4.6.1 Poverty status and demographics of household 38

4.6.2 Poverty status and dimensions (indicators) 44

CHAPTER V: CONCLUSION 55

5.1 Summary 55

5.2 Conclusion and recommendation 56

5.3 Limitation and further research 58

References 60

Appendix 63

Page | vi

Appendix 1: source of data in VHLSS 2014 63

Appendix 2: Spearman test result of five characteristics of a household 64

Page | vii

AP LIST OF TABLES

Table 2.1: Poverty threshold of Ministry of Labor, Invalid and Social Affairs in Vietnam from

1998 to 2015 6

Table 2.2: Poverty threshold of General Statistic Office in Vietnam from 1998 to 2010 and later 7

Table 2.3: 5 dimensions and 10 indicators of General scheme 2015 11

Table 2.4: Livelihood assets and dimensions of Tran & Nguyen (2012) 13

Table 2.5: Set of dimensions and indicators of Salaza et al (2013) 14

Table 3.1: Choice of dimensions and indicators 19

Table 3.2: Dimensions and indicators 23

Table 3.3: Number of households in provinces 25

Table 4.1: The influence of cut off ratio k 26

Table 4.2: Spearman correlation result of indicators and income of household and p-value 29 Table 4.3: Gender of head of household and multidimensional poverty status 39

Table 4.4: Ethnic of head of household and multidimensional poverty status 40

Table 4.5: Number of member in household and multidimensional poverty status 41

Table 4.6: School year of head of household and multidimensional poverty status 42

Table 4.7: Age of head of household and multidimensional poverty status 43

Table 4.8: Poverty status and deprivations of indicators in dimension of education 45

Table 4.9: Poverty status and deprivations of indicators in dimension of employment 46

Table 4.10: Poverty status and deprivations of indicators in dimension of health 47

Table 4.11: Poverty status and deprivations of indicators in dimension of living standard 49 Table 4.12: Poverty status and deprivations of indicators in dimension of monetary 52

Table 4.13: Poverty status and deprivations of indicators in dimension of social capital 54

Page | viii

LIST OF FIGURES

Figure 2.1: Theory framework of Alkire & partners (2014) 9

Figure 3.1: Conceptual framework 18

Figure 4.1: relationship of headcount ratio H and cutoff ratio k 27

Figure 4.2: The relationship of average intensity A and cutoff ratio k 28

Figure 4.3: Rate of uni-dimensional poverty and multidimensional poverty in nineteen provinces 30

Figure 4.4: The comparison of four dimensions between the Southeast region and the Southwest region (Education, employment, health and social capital) 32

Figure 4.5: The comparison of two dimensions between the Southeast region and the Southwest region (Living standard and monetary) 33

Figure 4.6: Headcount ratio H and Average intensity A in 19 provinces 35

Figure 4.7: Adjusted headcount ratio (Multidimensional poverty index) for provinces 36

Figure 4.8: Contribution of dimensions 37

Page | ix

CHAPTER I INTRODUCTION

In this chapter, there are five main parts, research problems, research objectives, researchquestions, research hypothesis and research scope This chapter focus to propose the realsituation of the Southern region of Vietnam and the role of research targets Besides, thefollowing parts will answer the questions: “Why this research is necessary?”; “What is thepurposes of this research?”

1.1 Research problem

Poverty is a considerable problem of every nation, especially for developing countries Based

on the poverty status of a nation, policy maker would able to declare suitable decision Inrecent years, poverty in Vietnam is measurement depend on monetary perspective It meansthe particular household would be considered poor if the income or expenditure were lowerthan social standard (which is set by the Government) Specially, the Southern region ofVietnam plays a very important role in nation’s economy in term of its contribution to GDP,labor force, food supply, etc Consequently, the most important work is determine povertystatus effectively, entirely and representatively

Locating at the South of the country, this region is separated into two smaller areas, theSoutheast region and the Southwest region In the Southeast region, there are six provinces

in this region: Binh Phuoc, Binh Duong, Tay Ninh, Dong Nai, Ba Ria – Vung Tau and Ho ChiMinh city The other counterpart includes 13 provinces: Long An, Tien Giang, Ben Tre, VinhLong, Tra Vinh, Dong Thap, Hau Giang, Soc Trang, An Giang, Kien Giang, Bac Lieu, Ca Mauand Can Tho city The Southern region of Vietnam borders with Cambodia and capture a longcoastline of South China Sea and Gulf of Thailand Total area of the Southern region ofVietnam is about 64,151 thousand square kilometers Natural features of this region areconducive for agriculture, livestock and tourism with dense river systems (total length of up

to 5,700 kilometers), high rainfall (1,500-2,000 millimeters a year) and fertile soil (slob andbasaltic soils)

Page | 1

Besides, the Southern region of Vietnam is a major economic region of the country separatedinto two key economic zones Southern Key Economic Zone and Mekong Delta Key EconomicZone are contributing to the whole economy significantly, about 59% to total GDP of Agriculturalsector of Vietnam in 2011 Infrastructure and logistic systems in Southern are developed withTan San Nhat international airport and many seaports connected by convenience road systems.Specially, manufacturing and services are the major sectors of the Southern and contribute 60%

of manufacturing and services sector of county in 2011

However, behind the scenes, according to the Vietnam Human Development Report of UNDP in

2015, the ratios of poor people are 13.4% in the Southeast region and 41.6% in the Southwestregion compared to 21.3% in total of the country The infant mortality rate is high, occupying8/1000 births in the Southeast and 11/1000 births in the Southwest In addition, childmalnourish proportion is about 20% in this area These figures pose the issues that a large part

of residents in the Southern are now living under poor status and having poor life condition.Hence, the measurement of poverty is needed to illustrate and analyze the poverty status in thisarea not only in economic perspective but also in social perspective

Moreover, poverty status is different among provinces and areas, especially urban area andrural area According to General Statistics Office (GSO), about 70% of Vietnamese are nowliving in rural area and a large part of them are not having good living conditions This studywill focus only on rural area because poverty is much more severe and broaden compared tourban area

According to Sen (1976), poverty could able to measure in multidimensional approach,which is called multidimensional poverty This is a broader indicator to represent for povertystatus of a country Poverty is not only comes from monetary problem but also come fromeducation, health, living standard, etc This paper will focus to estimate multidimensionalpoverty of the Southern region of Vietnam; build up suitable dimensions and indicators set;compare the poverty status among households and areas

Tran & Nguyen (2012) showed there are nine dimensions of poverty in rural area ofVietnam, or Tran (2014) also point out equivalent result, both of methodologies are PrincipalComponent Analysis & Multiple Correspondence Analysis

Page | 2

The latest research in national scale published in 2015 is the cooperation of Ministry ofLabor, Invalid and Social Affairs and UNDP This research called “General scheme oftransformation approach from measuring uni-dimensional poverty based on incomeapplicable to multidimensional in period of 2016-2020” The purposes of this research are:(i) to estimate the deprivation of social societal needs during periods, (ii) to determinemultidimensional poverty households, (iii) to determine the root cause of multidimensionalpoverty Moreover, this research also suggest 5 dimensions and 10 indicators to measuremultidimensional poverty status in Vietnam.

However, these previous empirical studies for Vietnam only estimate which is the suitabledimensions, no relationship or affect among Multidimensional poverty and thesedimensions have found In short, the main purpose of this paper will measure theMultidimensional poverty for the Southern region of Vietnam and find out the differences ofMultidimensional poverty and household’s characteristics More detail will be described inthe following section

1.2 Research objective

To prove the important role of Multidimensional poverty measurement and find the bestmethodology for the measurement, this paper focuses on four following objectives:

1 Determine suitable dimensions to measure multidimensional poverty household

2 Compare multidimensional poverty between the Southeast region and the

1.6 Organization of research

This research includes five chapters Chapter 1 is the introduction part of the thesis Chapter

2 gives basic concepts in poverty and some empirical studies relevant to multidimensionalpoverty Moreover, there are also a summarize table of literature review and theoryframework in chapter two Chapter 3 will present the methodology and approach ofmultidimensional poverty Chapter 4 and 5 will contain the result of this research andrecommendation to policy maker

Page | 4

CHAPTER II LITERATURE REVIEW

The previews of literature and empirical studies will be presented in this chapter Someconcept about poverty and multidimensional poverty will be defined and summarized.Moreover, this chapter also reviews about methods, approaches and findings of severalempirical studies on over the world and in Vietnam

2.1 Poverty and measurement

2.1.1 The concept and measurement of uni-dimension poverty

There are several definitions of poverty on over the world, these definitions is differentbecause of the aspects and standpoint However, the similarity is an individual or ahousehold is suffering insufficient conditions (physical and/or mental)

In early 19th century, Rowntree (1910) suggested that poverty is status of lack of money topurchase basic things (see Nguyen, 2010) This definition is similar to monetary povertyperspective, which have been using in some developing countries Later on, in 1995, at the

“World summit for social development Copenhagen” held in Copenhagen, Denmarkproposed that a person is considered as poor if his or her income is lower than $1 a day

World Bank suggests the definition of poverty in a much more specific way The poor sufferthe deprivation in basic goods and services such as education, health care, nutrition; theyare vulnerable and have no power in national institution (World Bank) Therefore, in thisdefinition, poverty can exist if living conditions of people are under the average levels of thecommunity

Absolute poverty and relative poverty

There are several points of views in definition of absolute poverty and relative poverty, whichare widely used around the world The similarity of these two concepts is the deprivation ofbasic need and/or the deprivation of rights and social position of an individual or household

Page | 5

To deal with this issue, World Bank (2000) suggested that absolute poverty is the deprivation

of minimum basic need of life and might face diseases, malnourish, etc Relative poverty isthe deprivation of a person, a group of people in this area compared to another area or thiscountry compared to another country In addition, Bellu and Liberati (2005) also mentioned

to this issue and suggested that relative poverty is an unequally of income and the quality oflife Statistics and measurements used to compare between observations are thereforerelatively Following the same perspective of inequality, Asia Development Bank (2012)suggested, “The poor are those who gain when income becomes more evenly distributedand the non-poor are those who lose” Hence, poverty is the gap between the poor and therich, the bigger gap the more poverty When it comes to absolute poverty, it is a deprivation

of a certain figure or statistic, for example, according to “World summit for socialdevelopment Copenhagen” in 1995, absolute poverty is the status of a person who is livingwith under $1 income a day

Poverty threshold

In Vietnam, poverty threshold (poverty line) is a threshold used to determine whether ahousehold is under poverty or not There is a difference in poverty threshold between theapproaches of General Statistic Office and Ministry of Labor, Invalid and Social Affairs

The government issues this threshold in a period of five-year For example, according to

“Decision No 170/2005 / QD-TTg”, a household will be classified as poor if the average income of

a member per month is less than 200,000 VND for rural area In 2011, the threshold was

increased up to 400,000 VND by “Decision No 09/2011 / QD-TTg” for period of 2011 – 2015.

Table 2.1: Poverty threshold of Ministry of Labor, Invalid and Social Affairs in Vietnam

from 1998 to 2015 Unit: VND

Rural

Urban

Page | 6

In other method, General Statistic Office uses the technique of World Bank to calculatepoverty threshold This threshold is the amount of money required a month to ensure 2,230calories a day per person and nonfood goods such as clothes, habitat, etc This figure will bemodified after some years In 2008, total income needed by a person to be not considered

2.1.2 Concept and measurement of Multidimensional poverty

Multidimensional poverty is a broader concept compared to uni-dimensional poverty.Decades ago, some studies (such as Sen et al., (1976); Sen et al., (1998); Boltvinik, (1998);etc.) have laid a groundwork for multidimensional poverty concept, which is the deprivation

of various aspects in human life A household might face the problem not only in monetaryaspect but also in many other health, education, information, etc Boltvinik (1998) concludedthat a household might not monetary poverty but that household might get trouble with therest dimensions

In a broader scale, according to United Nation Declaration (2008), the concept of poverty isnow extending Poverty is now defined as the deprivation of nutrition, school attendance,access to health care service, no productive asset and cannot access to credit Moreover,households considered poor have no rights and have a weak relationship to the community.This new definition about poverty is more specific than only in income aspect compared tothe previous definition decades ago

Hence, based on the declaration of United Nation and some recent studies, many policy makersand researchers on over the world are now believe poverty is multidimensional issue

Page | 7

However, poverty is varying among countries and areas; it also depends on demographic,geographic, religion and economy of each region.

Methods of multidimensional poverty measurement

Decades ago, the method of poverty measurement by one dimension is widely used, such asmonetary dimension (income or expenditure are indicators) However, this uni-dimensionmethod reveals the issue that the poverty status is just only reflected by monetarydeprivation status Health, education, living standard, social capital, etc is also important for

a live of everyone in the world The idea that a person might have plenty of money but have

no health or relationship in the community is also considered as poor Sen et al (1976)suggested the concept to solve this problem, which is the predecessor of multidimensionalpoverty Thirty years after, Alkire & Foster have announced a new method to calculatepoverty by a theory and concept of more than one dimension The set of three dimensioninclude dimension of Health, dimension of Education and dimension of Living standard,which will fully reflect the quality and status of a household

There are some common dimensions used by many authors around the world, health,education and living standard Health is one of the most common dimension besidemonetary dimension because it is the key factor of increasing productivity A country withhigh productivity must be have a good population health To estimate this dimension, manyauthors used various indicators to represent for the status of a household or individual.Salazar et al (2013) used two indicators “Health insurance” and “Access to health service” torepresent for health status of a household Dhongde and Haveman (2014) added the otherindicator called “Disability” to represent for any burden of a household member, which willcause direct impact to finance and working productivity of household In some developingcountries such as Latin American, a research done by Nawar (2014), who used mortality ofchildren as an indicator This indicator is reasonable and significant in case of developingcountries, where health of children still not be fully care in poor household

Education is similar to health in case of contribution to productivity, hence there arenumerous indicators used by authors There are some common indicators: “Adult literacy”,

“Year of schooling”, “Child education”, etc These indicators are used by many authors

Page | 8

estimate the status of education (Salazar et al., 2013; Nawar, 2014) Living standard isconsidered as one dimension in many researches, for example, Vijaya et al (2013) usedsome indicators to represent for this dimension, electricity, cooking fuel, water source, asset

of household These indicators will draw a full picture of household living condition Adeprivation in these indicators is the status of low living condition However, these indicatorsmight various among researches because of the variety of religion, living standard andgeography

Alkire & Foster approach

Nowadays, this method is widely used in various countries and institutions on over theworld Arguing that poverty is not a problem of monetary deprivation but also many otherissues, hence there are many dimensions and indicators were published to estimate poverty.The most prominent is the set of Alkire et al (2014) with three dimensions includingeducation, health and living standard that are proxied by 10 indicators

Figure 2.1: Theory framework of Alkire & partners (2014).

Page | 9

In this framework (figure 2.1), three dimensions were used to measure the poverty; eachdimension included two or more indicators Each indicator is selected and calculated from thedata and reality status of an individual or household Among these indicators, if a household islack of the particular indicators (base on cutoff ratio k) will be considered as poor Themeasurement of the deprivation of indicators is to accumulate the total weights of all indicatorsdeficiency (each indicator has a separate weights and if an individual or household fail to hold anindicator, that weights will be record) If an individual or household has total weights of allindicators deficiency higher than cutoff ratio k, that individual or household will be considered asmultidimensional poverty After that, the percentage of multidimensional poverty in the samplewill be called as H (headcount) and the average intensity of deprivation will be called as A (theintensity of poverty) Multidimensional poverty index (MPI) is the multiple of H and A Back tothe framework, dimension of education is including two separate indicators The first indicator is

“Years of schooling”, which will be considered as deprived if no one in the household has at least

of one year in school The second indicator is “Child school attendance”, which will beconsidered as deprived if there are at least one child in the household cannot go to schoolbecause of any reason In this dimension, these two indicators might reflect the large part of theeducation status in the household, the first indicator represent for the education status of adult

in the household, who are the breadwinner of family and the second indicator represent for theeducation status of the future generation of a family Dimension of Health is including twoindicators; “Child mortality” and “Nutrition”, these indicators cannot fully reflect the healthstatus of adult in the family because this research is in worldwide scale, which will get a lot ofconstrain in data However, the mortality of children will represent the skill, knowledge,monetary ability and living standard of parents “Nutrition” will indicate how well the health ofthe next generation of the family is To estimate the deprivation of living standard, the authorsused six indicators These indicators can fully reflect the living status of the household; electricand water indicators is the basic need for a household, “Flooring” and “Cooking fuel” representfor the quality of the habitant and daily diet, while the rest will indicate for the asset they own.Hence, this dimension is including a part of monetary dimension poverty method (some assetindicators)

However, there is an issue still unresolved, how to calculate the weights of indicators anddimensions reasonably? These weights will influence directly the poverty status of an

Page | 10

individual or household Hence, because of no reasonable method yet, the use of equalweights for all indicators with in a dimension is acceptable, and all dimensions are equal inweights.

2.2 Empirical studies

After review for key concepts of poverty and multidimensional poverty (including povertydefinition, poverty line, methods and approach), the approach to measure poverty is veryimportant in this research

The research done by Vietnamese Government and UNDP named “General scheme oftransformation approach from measuring uni-dimensional poverty based on incomeapplicable to multidimensional in period of 2016-2020” have fully reflect the situation andthe urgency of multidimensional poverty measurement In this research, authors reveals thatVietnam poverty rate is declining during the period of 2010-2015, from 14.2% to around 4%.However, this figure is based on uni-dimensional poverty measurement and it is much lowerthan multidimensional approach Authors applied five dimensions and ten indicators suitablefor all provinces in Vietnam (table 2.3)

Table 2.3: 5 dimensions and 10 indicators of General scheme 2015.

Dimensions

EducationHealthHabitatLiving conditionInformation accessibility

These entire dimension is weighted equally (0.2 each) and each indicator will get equal weight inits dimension Authors also mention about “extreme poverty”, which is the worst situation

of poverty Household is considered extreme poverty if that household gets deprivation inboth uni-dimension (monetary poverty) and multidimensional poverty.

Monetary approach is uni-dimensional approach, it answer the question “Does thathousehold has enough money for their living standard?” This approach is popular in manycountries (Vietnam is using this approach in many years) and easy to calculate However, thisapproach seem unsuitable now because it carrying various problems

In 1970s, multidimensional poverty concept is initially introduced to the world (Sen et al.,1976) This method is broader version of monetary approach More than thirty years later,Alkire and Foster published an international research of multidimensional poverty; theresearch including three dimensions: education, health and standard of living (eachdimension has many indicators) Besides that, Bibi (2003) showed a great paper tosummarize about the methods to measure Multidimensional poverty There is two methods

in general Non-axiomatic Approach and Axiomatic Approach.

There are some critical imperial studies Tran & Nguyen (2012) found 10 suitable dimensionsrepresent for four livelihood assets (including human resources, natural resources, physicalresources and social resources) (table 2.4) in Vietnam by using data of VHLSS 2008 In thisresearch, authors used new approach to calculate some very new indicators and dimension.For example, dimension of social capital, which is not common use on the world The proxy

of this dimension is the ethnic of household and this might affect the behavior and reaction

Table 2.4: Livelihood assets and dimensions of Tran & Nguyen (2012).

Alkire et al (2015) suggest the approach to measure Multidimensional poverty is Foster Counting Methodology, and probability model – Probit model to estimate whichcharacteristics of household will affect Multidimensional poverty status In fact, Nguyen(1998) used probit model to estimate the relationship between poverty status andhousehold characteristics in Ben Tre province Ballon and Apablaza (2012) applied Logitmodel to find out the relationships of Multidimensional poverty status with Demographicand Socioeconomic of a household head, the authors used five variables included education,gender, household size, living in urban area and religion The authors also found that, fourover five of the variables is strongly significant, only religion is non-significant Ataguba et al.(2011) applied regression to find out determinants of Multidimensional poverty in Nigeria,the result is also similar Hence, regression is a useful tool to find out the determinants ofmultidimensional poverty within a country (Alkire et al 2015)

Alkire-In the broader scale, numerous countries around the world have applied multidimensionalpoverty for many years Presidents of Colombia, Juan Manuel Santos have declared to applymultidimensional poverty (Alkire & Foster approach) to Colombia (Salaza et al., 2013) In thisproject, the target is to reduce poverty rate from 35% to 22% in a period of 2008-2014 Thiscountry also combine two measurement of poverty, income poverty measurement andmultidimensional poverty measurement Cutoff ratio “k” is set at 0.33 and extreme poverty

is set at 0.47 There are 5 dimensions and 15 indicators in the project (table 2.5)

Page | 13

Table 2.5: Set of dimensions and indicators of Salaza et al (2013).

Education achievement

Education

LiteracySchool attendance

Health

Access to health servicesAccess to water sourceAdequate elimination of waste

Access to public utilities and housing

Adequate floors conditions

Adequate external walls

No critical overcrowding

Concluding remarks

Based on empirical and theory reviews, some suitable dimensions and indicators have beenfound A set of dimensions included six dimensions, 20 indicators have established, and it will

be presented in the next chapter Besides, Alkire & Forster approach used to calculate

multidimensional poverty is a suitable approach for the Southern region of Vietnam withmany advantages The multidimensional poverty result of this region will be analyzed bydescriptive statistic methods because there are some problem with cause and effect in therelationship of poverty status and other factors of a household

Page | 14

CHAPTER III RESEARCH METHODOLOGY

In this chapter, there are five main parts: research methodology, multidimensional povertyindex, conceptual framework, choice of dimensions and indicators and the last part is datasource After several empirical studies, the set of dimensions and indicators have establishedand suitable methods are chosen to calculate multidimensional poverty index Besides, theprocess of calculating multidimensional poverty and the summary of data are also presented

 Applying Descriptive statistic (table summarize, crosstab, etc.) to analyze the

different between monetary poverty perspective and Multidimensional poverty

perspective

 Using Alkire-Foster multidimensional measurement methodology (AF) to

determine Multidimensional poverty, its dimensions, indicators and threshold foreach indicator/dimension

3.2 Multidimensional poverty index:

In recent years, the measurement of multidimensional poverty (Multidimensional povertyindex) suggested by Alkire & Foster are attracting international attention because of itsconvenience and advantages This is an overall index used to indicate or compare theproportion of poverty within or between the countries Initial introduced in 2007 by Alkire &Foster, this method is the result in multidimensional poverty index M0 There are three mainsteps to calculate the index:

Page | 15

Step 1:

Based on empirical studies and theory, establish all of the dimensions and indicators withthe suitable weighted for each indicator These dimensions and indicators should be stick tothe status of the studied area

Step 2:

Define the cutoff threshold for every single indicator based on empirical studies andgovernment regulations The cutoff threshold will determine the point of deprivation foreach indicator Each observation will be assigned a weighted sum of deprivations among allindicators called by “ci”

Define the poverty cutoff threshold among dimensions called by “k” If a weighted sum ofdeprivations of any observation higher than k, that observation will be considered asmultidimensional poverty This poverty cutoff threshold is the highest ratio of deprivationsthat a non-multidimensional poverty observation experience As the following function:

H is headcount ratio;

Page | 16

q is total multidimensional poverty observations and

n is total observations in sample

 Determine the average intensity by the following equation:

q is total multidimensional poverty observations

Finally, multidimensional poverty index can be calculated easily by the following equation:

= ×

Decomposition of MPI

According to Alkire et al (2011, p.6), MPI can be broken down to investigate the

contributions of components by the following formula:

= 11 + 2 2 +⋯+

Where:

wn is weight of indicator n and

CHn is censored headcount ratio of indicator n (ratio of total household deprived in indicator

n over total household of the sample)

Page | 17

3.3 Conceptual framework

Conceptual framework

There are many dimensions and indicators identified in literature review, however, this paperchoose 6 dimensions and 20 indicators, which are suitable and data-available to build-upconceptual framework (figure 3.1)

Figure 3.1: Conceptual framework.

Monetary

Social capital

Page | 18

Choice of dimensions and indicators

Table 3.1: Choice of dimensions and indicators.

Education is a major factor of the society and social welfare High level of education willbenefit not only productivity but also income and living condition According to the purpose

of Vietnam in 2015 to ensure that all children must attend in school, the indicator

“Education of children” is used to estimate education status of a household A household will get a deprivation in this indicator if at least one child cannot attend in school The other indicator is “Education achievement”, which represent for the whole education level of all

member in household A household will get a deprivation in this indicator if no householdmember have never go to school for a single year

dimension is “Male unemployment”, male is generally considered main income source of

household and take any burden of household Hence, if there is a male unemployed inhousehold, that is waste of source A household will get a deprivation in this indicator if atleast one male unemployed in household

Health

Productivity is determined not only by education but also by health of employee Employeeswith good health condition are able to suffer higher work pressure and capacity, which willincrease productivity of the whole community In addition, an individual with health cannotmake income, purchase goods or enjoy life Hence, this dimension is very important and must be

measure carefully by four indicators The first one is “Health insurance”, which will ensure the

financial ability of household when any member get problem in health A household will get adeprivation in this indicator if no member in household has health insurance The next indicator

is “Number of member got problem in health” If a member of household got

Page | 20

problem in health within 12 months, they might cause influences to their family and mightdecrease the productivity of their household Moreover, the productivity might not berecovered A household will get a deprivation in this indicator if at least one member got

problem in health within 12 months “Nutrition” is the third indicator of this dimension, it

must be required if a household is considered good health status However, nutrition ismainly depends on income of household A household will get a deprivation in this indicator

if a household has not enough nutrition within 12 months The last indicator is “Insufficient money for health service” Health service cost is a large part of health expenditure; poor

household might not have enough money to afford A household will get a deprivation in thisindicator if a household has not enough money for health service

Living Standard

Living standard is a need of basic condition for at least Many aspects, education, income,health, etc might determine living standard An individual must have basic conditions, which isequal to the characteristics of their community To fully estimate this dimension, seven

indicators were used First, “Utilities” is a very basic requirement of life, which will ensure an

individual have basic condition to work and live Safe water source and electricity is veryimportant to any individual A household will get a deprivation in this indicator if a household

cannot access to electricity grid or safe water source “Sewage” system is now becoming

important because it will affects health status of the whole community If a household do nothave waste treatment system, they might cause bad effect to the environment A household willget a deprivation in this indicator if a household do not have waste treatment system The next

indicator is “House level”, which represents for the quality of their house and the basic

conditions House level is separate into five groups in Vietnam, from level 1 is highest conditionand level 5 is lowest A household will get a deprivation in this indicator if a house of household

is level 5 “Asset ownership” is a very important indicator, which represent how well a

household is There are some basic assets a household to ensure normal conditions such astransportation vehicle, TV and phone A household will get a deprivation in this indicator if

household owns none of the previous asset Next, “Crowding” is an indicator represented for

the average square meter for a member An individual cannot have good living conditionswithout acceptable habitat A household will get a deprivation in this

Page | 21

indicator if average square meter for a member less than 10 “Saving ownership” represent

for that status of well financial condition of a household A household that has enoughmoney for expenditure and still has surplus might own a saving account A household willget a deprivation in this indicator if a household has no saving account The last indicator in

this dimension is “Toilet facility”, which is the basic need of every individual House with no

toilet might be considered poor and low living condition Nowadays, toilet is a basic facilityand should present in any house A household will get a deprivation in this indicator if ahousehold has no toilet

Monetary

Monetary dimension is very important because it used to a measurement in uni-dimensionalpoverty Decades ago, policy makers applied income or expenditure of a household toestimate poverty status However, poverty is a multidimensional issue so monetary can be

only considered a dimension “Below poverty line” is the most important indicator in this

paper It is used to estimate both multidimensional poverty and uni-dimensional poverty Ahousehold will get a deprivation in this indicator if average income per member a month is

lower than 653,000 VND The second indicator of this dimension is “Agricultural land”,

which is considered productive asset because agriculture is the main sector in rural area ofthe Southern region of Vietnam Household might get severe living condition if they do nothave much cultivated area A household will get a deprivation in this indicator if theircultivated area with 12 months less than 3,364 m2

Social capital

The last dimension is social aspect; this dimension will be a proxy for the relationship ofhousehold Information and relationships might direct impact to behavior of every householdmember and then influence other dimension Hence, information accessibility plays a crucial role

in household behavior because it will navigate the thinking and reaction of household to the rest

of the world This indicator named “Information accessibility” and a household will get a deprivation in this indicator if there is no phone in household The last indicator is “Join any group”, which represent for the relationship of household to the community Household might

get some helps and opportunities when they are member of any social group to

Page | 22

overcome difficulties in life and get better income A household will get a deprivation in this indicator if no household member is in any group.

3.4 Dimensions and indicators

Table 3.2: Dimensions and indicators.

Page | 23

3.5 Data sources

To satisfy for the research problem and objective, this paper needs an entire data set, whichcan represent for the Southern region of Vietnam in many aspects Hence, using data fromVietnam Household Living Standard Survey (VHLSS 2014), which contained 1977 households

in rural area of the Southern region of Vietnam would be the best solution The whole dataset VHLSS 2014 will be classify and eliminate to archive only rural area of Southern region ofVietnam’s data

Page | 24

Table 3.3: Number of households in provinces.

No.

12345678910111213141516171819

Total:

Page | 25

CHAPTER IV RESULTS

In this section, the main findings of this thesis will be presented More specifically, there aresome differences in poverty status among provinces and households Almost findings arestrongly significant and reasonable Besides, the result of uni-dimenisonal poverty andmultidimensional poverty are completely different The choice of cutoff ratio k is alsopresented in this chapter

4.1 The choice of cutoff ratio k

In this approach of multidimensional poverty measurement, cutoff ratio k is a crucialelement of the result In other words, cutoff ratio k will direct influence themultidimensional poverty result Because cutoff ratio k is the minimum proportion required,any household’s deprivation higher than cutoff ratio k will be considered poor According toAlkire et al (2009), this approach includes two cutoff ratio; the first one is cutoff thresholdfor each indicator to classify the deprivation of household The Second cutoff ratio is called k;therefore, k varies from the lowest weight of one indicator to the total weights 1 Thefollowing table depicts the value of k vary from 0.05 to 0.60

Table 4.1: The influence of cut off ratio k.

k 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60

Page | 26