Effects of social capital on self rated life satisfaction among residents evidence from vietnam

Social capital is measured as a multidimensional concept including social network, formal institution, cooperation and social trust.. Using order probit model on data for 3415 households

UNIVERSITY OF ECONOMICS ERASMUS UNVERSITY ROTTERDAM

HO CHI MINH CITY INSTITUTE OF SOCIAL STUDIES

VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

EFFECTS OF SOCIAL CAPITAL ON

SELF-RATED LIFE-SATISFACTION AMONG

RESIDENTS: EVIDENCE FROM VIETNAM

By

TRUONG NGOC ANH THU

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, November 2016

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY THE HAGUE

VIETNAM THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

EFFECTS OF SOCIAL CAPITAL ON RATED LIFE-SATISFACTION AMONG

SELF-RESIDENTS: EVIDENCE FROM VIETNAM

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

TRUONG NGOC ANH THU

Academic Supervisor:

Dr PHAM KHANH NAM

HO CHI MINH CITY, November 2016

Truong Ngoc Anh Thu

Ho Chi Minh City, 2016

In my capacity as the supervisor of this thesis, I certified that the statements above are true to the best of my knowledge

Signature

Dr Pham Khanh Nam

Date:

ABSTRACT

This dissertation provides an insight into the relationship between self-rated satisfaction and social capital, controlling for some demographic and individual characteristics in rural Vietnam as well as tests the theoretical predictions of the social capital investment model from the data of Vietnam Access To Resources Household Survey in

life-2014 Social capital is measured as a multidimensional concept including social network, formal institution, cooperation and social trust The results are quite consistent with the predictions, which supports the theory about the social capital investment of Glaeser et al (2002) that people who have higher education would have more investment in their social capital, and being stable in the residence also contributes to their investment of social capital while relatively higher income affects their social network and trust negatively Finally, the association between age and formal institution follows the inverted U-shaped trend Using order probit model on data for 3415 households from twelve provinces in Vietnam, it is found out that there is a significantly positive relationship between social capital in forms of social network, formal institution and cooperation and an individual’s subjective satisfaction However, trust, on the other hand, is examined to be insignificantly associated with a person’s self-rated satisfaction in this research

Keywords: social capital, social network, cooperation, formal institution, social trust, happiness, life-satisfaction, order probit model, rural Vietnam

TABLE OF CONTENTS

DECLARATION i

ACKNOWLEDGEMENT ii

ABSTRACT iii

TABLE OF CONTENTS iv

LIST OF TABLES vi

LIST OF FIGURES vii

CHAPTER 1: INTRODUCTION 1

1.1 PROBLEM STATEMENT 1

1.2 RESEARCH OBJECTIVES AND QUESTIONS 2

1.3 RESEARCH SCOPE AND METHODOLOGY 3

1.4 THESIS STRUCTURE 3

CHAPTER 2: LITERATURE REVIEW 4

2 1 THEORETICAL FRAMEWORK 4

2.1.1 Concept of social capital 4

2.1.2 Model of social capital investment 5

2.1.3 Becker and Murphy (2000)’s theory on the relationship between social capital and life-satisfaction 6

2 2 PREVIOUS EMPIRICAL STUDIES 7

CHAPTER 3: DATA AND METHODOLOGY 11

3.1 DATA 11

3.2 RESEARCH METHODOLOGY 11

3.2.1 Variable description 11

3.2.2 Econometric model 14

CHAPTER 4: RESULTS AND DISCUSSION 15

4.1 DESCRIPTIVE STATISTICS 15

4.2 RESULTS OF SOCIAL CAPITAL INVESTMENT MODEL 20

4.3 RELATIONSHIP BETWEEN SOCIAL CAPITAL AND SELF-RATED LIFE-SATISFACTION 24

4.4 ROBUSTNESS TESTS 36

CHAPTER 5: CONCLUSIONS 39

5.1 CONCLUSIONS 39

5.1.1 Concluding remarks from the test of model of social capital investment 39

5.1.2 Concluding remarks from the estimations of relationship between social capital and subjective life-satisfaction 40

5.2 POLICY RECOMMENDATIONS 41

REFERENCES 44

APPENDIX A 48

APPENDIX B THE RESULTS OF ORDER PROBIT ESTIMATIONS 49

LIST OF TABLES

Table 3 1: Factor analysis of social capital 12

Table 3 2: Variable Description 13

Table 4 1: Descriptive statistics 16

Table 4 2: Correlation coefficients among key variables 18

Table 4 3: Regression of demographic variables on social capital 23

Table 4 4: Results of estimations of the relationship between life-satisfaction and social capital 26

Table 4 5: Results of the test of the impact of relative income on the relationship between life-satisfaction and social capital 30

Table 4 6: Results of the test of the impact of children on the relationship between satisfaction and social capital 32

Table 4 7: Results of the test of the impact of the time of living in the same community on the relationship between satisfaction and social capital 34

Table 4 8: Results of estimations of the relationship between social capital and life-satisfaction (robustness check) 37

LIST OF FIGURES Figure 2 1: The framework of the relationships among social capital, other dependent

variables and life-satisfaction 10

Figure 4 1: Average life-satisfaction in twelve provinces in 2014 17

Figure 4 2: The relationship between social network and life-satisfaction 19

Figure 4 3: The relationship between formal institution and life-satisfaction 19

Figure 4 4: The relationship between cooperation and life-satisfaction 20

Figure 4 5: The relationship between trust and life-satisfaction 20

CHAPTER 1: INTRODUCTION

1.1 PROBLEM STATEMENT

Social capital has increasingly become an important topic in science and policy since many empirical evidences suggest that social capital is connected with various aspects of life (Putman, 2000) and may have a positive impact on the life-satisfaction of individuals and the whole countries Many findings from earlier studies have indicated that various forms of social capital are strongly associated with the level of human’s subjective satisfaction because they provide an individual with networks and connections of social interaction as well as help he or she to gain support from other people in society (Helliwell, 2003) Bjornskov (2003) showed that social capital was an important element in explaining why some countries are happier than others Putnam (2000) also found that social capital was positively related with happiness across the USA Thus, it is said that communities which are rich in social capital and individuals who have high levels of personal social capital are believed to be generally able to enjoy more satisfaction

However, conflicting results have been also reported due to the vagueness of the definition of social capital and the measurement of satisfaction Some researchers find no statistical significance of the relationship between social capital and life-satisfaction in their study Yet the relationship social capital and subjective life-satisfaction has not been investigated among residents in rural Vietnam Since the country has experienced dramatic social, economic, and political changes in the past years, which may lead to the changes in the socio-cultural context which is one of the formations of social capital, the relationship between social capital and life-satisfaction is worth researching in Vietnam Moreover, traditionally Vietnamese people tend to be very close to each other in family and the society Family, village and nations are considered to create a basic structure for Vietnamese society (Hickey, 1964) Hence, people’s accumulated social capital could then affect their life-satisfaction

Furthermore, in recent years, economists have given more attention on the empirical studies of happiness or life-satisfaction since life-satisfaction is considered as one of the indicators of life quality which is an issue that many countries care and focus on Life-satisfaction is one of the keys for a wealthy country due to many reasons A country can get many benefits from its satisfied citizens in economic, social, and psychological terms First, satisfied people are healthy one who can live longer and make full use of their strengths,

skills and abilities to contribute to their own well-being as well as that of the others and the society Second, a satisfied individual deals with depression better, he or she is less prone to anxiety, stress or anger, hence is likely to commit in fewer crimes Moreover, satisfied people are less of a burden to health services, social welfare agencies, or the whole economy

in comparison with depressed and unsatisfied people Therefore, it could be said that increasing the level of satisfaction of an individual does not only affect that person but also helps to improve the whole country Moreover, though Vietnam is still a developing country,

it is usually known as a country with a high rank in happiness index in the world According

to New Economics Foundation (2012 & 2016), Vietnam came second in Happy Planet Index (HPI) among 151 countries in 2012, and ranks fifth in 2016 globally, second in the Asia Pacific region However, is it true in Vietnam rural area? Basing on the results of the survey VARHS 2014, the level of self-satisfaction of Vietnamese people will be discussed to see if Vietnamese are really satisfied with their life

Besides, evaluating people’s social capital and estimating the determinants of rated life-satisfaction of people in rural Vietnam helps policy makers and local government

self-to make policy decisions in investment more effectively Since Vietnam is in its process of industrialization and modernization, suitable and good policies shall be introduced and implemented in order to improve people’s life as well as to develop rural area in general Therefore, this empirical study aims to contribute to the literature the case study of rural Vietnam about the relationship between social capital and life-satisfaction of Vietnamese residents

1.2 RESEARCH OBJECTIVES AND QUESTIONS

This research is to investigate the association between social capital and subjective life-satisfaction in rural Vietnam Social capital is measured as a multidimensional concept consisting of social trust, network including formal institution and informal institution, and cooperation while life-satisfaction is self-rated by Vietnamese residents

This study attempts to answer the following questions:

What are the characteristics which contribute to the investment in social capital of an individual?

Does an individual feel more satisfied with life if he or she has more social capital?

In other words, if an individual has more friends, joins more institutions, more trust and is more cooperative, is he or she more satisfied with life?

Do relative income, gender, the number of children, and the number of years living in the same community affect the effects of social capital on an individual’s subjective life- satisfaction?

1.3 RESEARCH SCOPE AND METHODOLOGY

The research uses the data from Vietnam Access To Resources Household Survey (VARHS) conducted in twelve Vietnamese rural provinces in 2014

The main methodology of this study is quantitative analysis Specifically, order probit estimation is used to estimate the relationship between self-rated life-satisfaction and social capital

Following Bjornskov (2003) and Frey and Stutzer (2002), terms “happiness”, satisfaction”, “self-rated happiness”, “self-rated life-satisfaction”, “subjective happiness”,

“life-“subjective life-satisfaction”, and “wellbeing” in this research are used interchangeably to indicate the happiness of an individual

1.4 THESIS STRUCTURE

This thesis has five chapters Chapter 1 introduces the background and motivation of the research on the effect of social capital and life-satisfaction in Vietnam Chapter 2 reviews related theories and previous empirical studies Chapter 3 presents the data as well as summary statistics and methodology for producing necessary results Chapter 4 shows the results and discussions Chapter 5 summarizes main research findings, suggests policy implications

CHAPTER 2: LITERATURE REVIEW

This chapter first reviews the definition of social capital and the measurement of rated life-satisfaction Then, two main theories relating to social capital model and the relationship between social capital and satisfaction are introduced The investment model of social capital suggested by Glaeser et al (2002) explains an individual’s decision in accumulating social capital and the optimal condition where an individual maximizes his returns to social capital The theory of utility gained by social capital by Becker and Murphy (2000) is, after that, presented to build the framework for the relationship between social capital and satisfaction At the end of this chapter, some related previous empirical analyses are presented in order to provide a deep insight into the researches of life-satisfaction and

self-social capital before

2 1 THEORETICAL FRAMEWORK

2.1.1 Concept of social capital

Social capital is generally considered as a vague and intangible term which is difficult

to describe Even until now, no single definition can give an accurate and broadly recognizable concept of social capital Hence, in order to build a good proxy for social capital

so as to analyze the relationships between social capital and individual’s subjective satisfaction, in this study, a dimensional definition of social capital is constituted basing on the theory of Coleman (1988), Putnam (2000) and Paxton (1999)

life-According to Coleman (1988), social capital is the social relation facilitating economic and non-economic goods, including obligations within a group, trust, intergenerational closure, norms and sanctions These trust and obligations then encourage coordination and cooperation among individuals In this concept, Coleman illustrates how an individual within a group can gain resource from the social structure of that group, which suggests the idea of creating proxy of trust and cooperation in this study

Putnam (2000) has widened Coleman’s social capital theory by putting more emphasize on interpersonal networks He defines social capital as a combination of trust, norms, and formal and informal social networks which exists in all relations in people’s life Formal institution is the formal groups or organizations that individuals join The chances

of meeting with others in a group enhance the frequency of interaction among individuals, thus increasing social capital Informal institution is while known as the interaction with family and friends

Paxton (1999) also suggests that there are two components creating social capital One

is the objective relations among people which bring them together and tie them to each other

in everyday life Another is the subjective type of tie which happens among individuals such

as trust, reciprocal and positive emotion

Overall, following these concepts, the proxy of social trust, and cooperation are created

in this study in order to indicate the individual’s trust in others as well as community and his

or her cooperation Moreover, social capital is also measured by formal institution and informal network that an individual engages in In this research, a proxy for general social capital index is tried to be created by the four above proxies of social capital However, it fails to establish this index Therefore, only four proxies of trust, cooperation, social network and formal institution are together used to estimate the associations between social capital and individual’s self-rated life-satisfaction

2.1.2 Model of social capital investment

Coming from the idea that though there are theories and evidences on the effects of social capital, there has not yet existed a commonly accepted theoretical framework about the determinants of investment in social capital, Glaeser et al (2002) suggested an analysis

of the formation of social capital using a model of optimal individual investment decisions According to the model of investment in social capital suggested by Glaeser et al (2002), the dynamics of social capital stock S are followed by a dynamic budget constraint and described as:

St+1 = 𝛿St + It Where S is individual social capital, 𝛿 is the depreciation of stock of social capital which is lower than 1 (𝛿 <1) So, 1 - 𝛿 is the depreciation rate It is the level of investment having a time cost C(It) in which C(.) is an increasing and convex function The opportunity cost of time is the wage rate w

In the theory of Glaeser et al (2002), social capital is defined as social characteristics comprising of social skills, and charisma that enable a person to obtain market and non-market returns from the connections with other people in the society It is said that when a person invests in social capital in order to gain market and no-market returns, the acts of accumulating social capital are positively associated with that person’s self-rated life-satisfaction If this investment does not mean to affect their subjective satisfaction, it may due to the fact that that person is seeking for market returns During a lifespan of T periods,

the individual’s maximization problem of social capital could be expressed as:

Max𝐼0,𝐼1, 𝐼𝑇∑𝑇𝑡=0𝛽t[StR(𝑆̂) – wC(It)], s.t St+1 = 𝛿∅ St + It Where 𝛽 is the future discount factor, 𝑆̂ is the aggregate per-capita social capital SR(𝑆̂) is the per period utility flow each person receives, including market returns and non-market returns Market returns may include wages or better jobs for the socially skilled person Non-market returns may include the improvements in his relationships, health and direct satisfaction R’(𝑆̂) is assumed to be > 0 because an individual will reap nothing from

a group which has no other members The symbol ∅ represents the depreciation factor coming from mobility (∅ = (1 - 𝜃) + 𝜃𝜆), where 𝜆 is the depreciation of his social capital when a person moves to another place The optimal condition related with the investment problem where an individual maximizes his returns to social capital is given by:

wC’(It) = 1−(𝛽𝛿∅)

𝑇−𝑡+1

1− 𝛽𝛿∅ R(𝑆̂) This first order condition shows that the social capital investment (1) increases with the discount factor, 𝛽, (2) rises with occupational returns to social skills, R (.), and (3) increases with communities with more aggregate social capital 𝑆̂, while it (4) declines with mobility, 𝜃, (5) decreases with the opportunity cost of time, w, (6) declines with the rate of social capital depreciation, (1 – 𝛿), (7) declines with the rate of social capital depreciation due to relocation (1 - 𝜆) and (8) declines with ages, t This study will test these theoretical suggestions of model of social capital investment

2.1.3 Becker and Murphy (2000)’s theory on the relationship between social capital and life-satisfaction

Since the beginning of human civilization, life-satisfaction is the common ambition that individuals and countries are trying to reach Though there are many ways to evaluate satisfaction, subjective measures of life-satisfaction have been believed to be reliable and valid in terms of assessing a person’s well-being because many researchers found that the subjective well-being measures or satisfaction measures are highly correlated with measurements of satisfaction from both laboratory experiments by observing attendees’ facial expression and vocal tones and sources from people closed to the attendees (Diener, 1994) Therefore, the self-judged life-satisfaction applied in this research is considered to be

reliable in terms of assessing a person’s wellbeing

As suggested by Denier (2000), individual’s satisfaction is affected by both internal and external factors The internal factors could be subjectively inner mental and physical feelings which create sense of satisfaction while external factors could be environment, and social capital The literature has provided many evidences that there may be relationships between social capital and satisfaction since social capital provides an individual with kinds

of support, chances of seeing others as well as sharing and then creates a sense of trust and belonging According to Becker and Murphy (2000), besides goods and services consumed which are believed to affect people’s utility, social capital is also considered as an input that yields a person’s utility directly The function is as below:

U = U (x, y; S) Where x and y are all kinds of goods and services, S is social capital which does not shift the utility function, but moves the utility level up or down, which means social capital

is not just a part of environment which is usually ignored or left in the background like in other theories but is a direct element affecting level of utility of an individual For example, the fact a person is less satisfied or more satisfied depends on how many friends he or she has Because individuals have more social relationships and attend more social activities, they feel more satisfied with their life This approach is appropriate when the social environment is stable When the environment changes as when a family moves, social forces may shift the utility function (Becker & Murphy, 2000) This theory is the base to build the model:

H = U (S) Where H is the level of life-satisfaction, S is social capital Social capital directly affects the level of satisfaction of an individual

2 2 PREVIOUS EMPIRICAL STUDIES

Many researchers have found that social capital positively influences self-reported satisfaction at individual level in form of social network since social capital offers people with some sort of support and chances of sharing which may provide people the feeling of belonging and integration in the society (Putnam, 2000; Helliwell and Barrington-Lee, 2010; Leung et al 2013; Rodríguez-Pose & Berlepsch, 2014) Pichler (2006) indicates that social networks help people to find their own position in the society Time with family, friends, and neighbors is estimated to be positively related with life-satisfaction (Lelkes, 2006;

Powdthavee, 2008; Peasgood, 2007) Helliwell and Putnam (2004) suggest the marriage, family, and contact with friends and neighbors, and involvement in community appear robustly related to happiness and satisfaction Powdthavee (2008) describes life-satisfaction could be increased proportionally with the increasing of frequency of social contacts However, this relationship is not robust according to Bjørnskov (2008) when he finds no statistical significance of the relationship between informal networks and satisfaction in his cross-sectional analysis for the United States

Furthermore, as a part of social capital, social trust and formal participation are found

to be highly correlated with life-satisfaction Rodríguez-Pose & Berlepsch (2014) run ordinal logistic regression from the individual data for 25 European countries and finds that social capital affects satisfaction positively, especially social interaction and trust Helliwell and Putnam (2004), Helliwell (2006), and Bjørnskov (2006) all suggest that trust has a positive impact on satisfaction Bjørnskov (2006) finds a positive relationship between generalized social trust and life-satisfaction by running regression of an international sample of more than 80 countries According to Bjørnskov (2008), a high level of social trust allowed people

to share and work with others whom they do not know, thus making a happier life This prediction was proved by the robust results in analysis of 48 the US States Chang (2009) run the order probit estimation of 35070 observations in Taiwan and finds that trust has a positive impact on the level of satisfaction Yip et al (2007) reports that cognitive social capital such as social trust is positively correlated with the psychological health as well as subjective wellbeing at the village level They further find that trust affects health and wellbeing through social network and support Moreover, Uslander (1999) suggest trust as the most essential factor of social capital creating the vibrancy in the community Hence, a positive parameter for trust and satisfaction is expected However, Ram (2010) documents that the relationship between trust and satisfaction is very fragile and statistically insignificant in different cross-country samples Leung et al (2013) run the bootstrap hierarchical regression in the survey of 24951 participants who are 15 years old and older across the ten Canadian provinces and find that only trust in family is significantly related with satisfaction, trust in neighbors or strangers do not affect people’s satisfaction

Regarding the formal network, there is little statistical and inconsistent correlation between structural social capital (organizational membership) and people’s satisfaction According to Bjornskov (2006), civic participation is found to be negatively associated with

people’s wellbeing Leung et al (2013) in their research in Canada also report that people who engage in civic participation are not happier than others who do not join Yet, in a study

of European countries, it is found that there is a strong relationship between the membership

in organizations and self-rated satisfaction even when other social explanatory factors are taken into account (Gundelach & Kreiner, 2004) Miller and Buys (2008) contribute to the empirical studies by the study in Australia which finds that social capital and participation

in community activities predict happiness, health and life-satisfaction

Cooperation, in form of social capital, also has an impact on life-satisfaction Some researchers predict that cooperation could affect people’s satisfaction because the feeling of engaging in an activity together creates people’s excitement and satisfaction (Argyle & Lu, 1990) Less cooperativeness over working with other people might be very risky to satisfaction (Lu & Argyle, 1991) Though the finding in research of sixty-three adult subjects

by Argyle & Lu (1990) shows that cooperation fails to be correlated with the level of satisfaction, which may come from the inadequate measure of cooperativeness Lu & Argyle (1991) reports a positive relationship between life-satisfaction and cooperation when they run regression for a small sample of 114 participants from 18 to 65 years old Furthermore, using a number of small samples, Crandall (1980) also finds that cooperation is associated with self-reported life-satisfaction

Besides social capital, many other factors could affect people’s life-satisfaction, Diener et al (1999) describes that demographic variables such as age, sex, income, race, education and marital status could influence satisfaction, though they only account for 8-20% of the variance in explaining satisfaction Age square is added to the model since Blanchflower and Oswald (2008) show that people around the age of fifty is reported to have the lowest life-satisfaction, which may imply that younger people have higher life-satisfaction than older people Gender is also a determinant in estimating the relationship between social capital and life-satisfaction Because women are usually found to be more socialized than men (Hudson, 2006), the effects of social capital on their satisfaction may be bigger than men’s one Furthermore, the relationship between money and satisfaction is believed to be positive but not really significant (Diener et al., 1993; Frey & Stutzer, 2010) Finally, health plays an important role in self-rated life-satisfaction (Kozma & Stones, 1983) Perneger et al (2004) explores that the feeling of satisfaction all or most of the time respondents in his cross-sectional survey of 1257 randomly selected university students in

Geneva, Switzerland is not only strongly related with better mental health, but also with the feeling of getting enough love and affection (from social capital) However, the association between satisfaction and physical health is found to be weak and statistically non-significant When analyzing the relationship between social capital and satisfaction, however, some problems need to be addressed Recent researches have found the non-robustness of the association between social capital and satisfaction According to Bjørnskov (2008), this relationship is only stable once the country being investigated has reached a certain income threshold Moreover, the vagueness of the definition of social capital leads to the inconsistency in the use of proxies, which make the differences in the empirical results Therefore, in this study, multidimensional concept of social capital comprising social trust, network including formal institution and informal institution, and cooperation is employed

to transfer as much as possible the influences of social capital on self-rated life-satisfaction 2.3 THE CONCEPTUAL FRAMEWORK

Basing on the background of the aforementioned literature review in relation to both theories and empirical studies, the conceptual framework of the relationships between social capital and life-satisfaction is proposed in below figure 2.1

Figure 2 1: The framework of the relationships among social capital, other dependent

variables and life-satisfaction

Source: author’s summarization

Demographic characteristics:

age, gender, education,

income, marital status

life-CHAPTER 3: DATA AND METHODOLOGY

This chapter introduces data, social capital, life-satisfaction measurements, and methodology First, the data used in this research is collected from Vietnam Access To Resources Household Survey in 2014 The proxies of social capital, life-satisfaction, and other controlled variables are then explained in this chapter Finally, the econometric model and variable description are given in order to explain more about the data and the methodology

3.1 DATA

Data was taken from Vietnam Access To Resources Household Survey (VARHS) which was conducted in the rural areas of twelve provinces in Vietnam (including ex-Ha Tay, Nghe An, Khanh Hoa, Lam Dong, Dak Lak, Dak Nong, Lao Cai, Dien Bien, Lai Chau, Phu Tho, Quang Nam and Long An) in 2014 by a collaborative project among Central Institute for Economic Management (CIEM) of the Ministry of Planning and Investment (MPI) and the Centre for Agricultural Policy Consulting of the Institute of Policy and Strategy for Agriculture and Rural Development (CAP-IPSARD) of the Ministry of Agriculture and Rural Development (MARD), the Institute of Labour Science and Social Affairs (ILSSA) of the Ministry of Labour, Invalids and Social Affairs (MoLISA), and the Development Economics Research Group (DERG) of the University of Copenhagen The data used for analysis was filled for only people aged 18 years and older (reduced the observations to 3415) Thanks to the large observations, it ensures that the sample is representative of the rural population within the sampled provinces and also makes the VARHS become a strong and reliable tool for getting detailed and good information about the economy and society of rural Vietnam and then doing research

3.2 RESEARCH METHODOLOGY

3.2.1 Variable description

In this dissertation, social capital is a multidimensional concept measured by social trust, network including formal institution and informal institution, and cooperation (i) Formal institution is measured by the number of formal associations that the

respondents participate

(ii) Informal institution or social network is measured by the number of people that

the respondents can ask for help when they have difficulties

(iii) Trust is calculated by the question asking “most people are basically honest

and can be trusted”, the responses are “Agree”, “Disagree”

(iv) Cooperation is measured by choosing between two options Option 1 is to get

and farm 1/2 hectare of land entirely by their own while option 2 is to join to farm 3/2 hectares of land with other family (not related to them)

Moreover, a composite social capital index is tried to be built in this model so as to see the general effects of social capital on an individual’s self-reported life-satisfaction However, when applying the factor analysis, it seems impossible to create this index because

it is not valid for estimations since no factors can be used as a representative for social capital

in this research (see details in Table 3.1)

Table 3 1: Factor analysis of social capital

Factor analysis/correlation Number of obs = 3415

Method: principal factors Retained factors = 3

Rotation: (unrotated) Number of params = 6

Factor Eigenvalue Difference Proportion Cumulative Factor1 0.18046 0.04190 1.6742 1.6742

Factor2 0.13856 0.13404 1.2855 2.9597

Factor3 0.00452 0.22026 0.0419 3.0016

Factor4 -0.21574 -2.0016 1.0000

LR test: independent vs saturated: chi2(6) = 178.90 Prob>chi2 = 0.0000

Factor loadings (pattern matrix) and unique variances

Variable Factor1 Factor2 Factor3 Uniqueness Network 0.3306 0.0742 - 0.0047 0.8852

life-Other dependent variables are demographic variables including people’s age, age square, gender, their income, education, marital status, the number of children and the number of ill days in a year to see how health affects a person’s life-satisfaction as well as the number of years that he or she lives in the same community

All variables are described in the following summary table:

Table 3 2: Variable Description

Social capital

Formal institution Number of formal associations participated

Social network Number of people that respondents could ask for help

Trust A two- point scale response to the statement “Most people are

basically honest and can be trusted” (Agree (1)– Disagree (0))

Cooperation

A dummy variable Choose between two options Option 1 (0)

is to get and farm 1/2 hectare of land entirely by their own while option 2 (1) is to join to farm 3/2 hectares of land with other family (not related to them)

Demographic characteristics

Children Number of children the respondent has

Illdays The number of days that the respondent was not able to perform

normal activities due to sickness during the past twelve months

Relativeincome The relative income of a household in the province

Yearsliving The number of years of living in this commune

Married If married, married = 1; otherwise, married = 0

Widowed If widowed, widowed = 1; otherwise, widowed = 0

Divorced If divorced, divorced = 1; otherwise, divorced = 0

Separated If separated, separated = 1; otherwise, separated = 0

Shortvocational If short vocational, short vocational = 1; otherwise short

vocational = 0 Longvocational If long vocational, long vocational = 1; otherwise long

vocational = 0 Specializedhighschool If specialized high school, specialized high school = 1;

otherwise specialized high school = 0 College If college, college = 1; otherwise college = 0

University If university, university = 1; otherwise university = 0

Master If master, master = 1; otherwise master = 0

in the same community

Ordered Probit estimations are used to investigate the relationship between self-rated life-satisfaction and social capital The estimation function takes the form:

HAPPIi = α1 SCi + α2 Xi + εiHAPPIi is the dependent variable of an individual’s self-rated life-satisfaction while

SCi stands for social capital In this study, SC1 is the social network, SC2 is the formal institution, SC3 is cooperation and SC4 is social trust

HAPPI1 = α11 SC1 + α21 Xi + εi1HAPPI2 = α12 SC2 + α22 Xi + εi2HAPPI3 = α13 SC3 + α23 Xi + εi3HAPPI4 = α14 SC4 + α24 Xi + εi4

Xi includes demographic characteristics: age, age square, gender, income, education, marital status, the number of ill days, the number of years living in the same address and the number of children

To test which factors amplify the relationship between social capital and satisfaction, some interaction terms are added into the model as below:

life-HAPPIi = α1 SCi + α2 Xi + α3 SCi Xi + εi

In this research, the factors which are tested are the relative income, gender, the number of children and the number of years living in the community

CHAPTER 4: RESULTS AND DISCUSSION

In this chapter, tables and charts of variables are presented to show the descriptive statistics, including mean, standard error, min and max Moreover, some graphs are introduced to compare the satisfaction level among twelve provinces in Vietnam and the relationship between social capital and life-satisfaction The estimation results are then presented and the main findings are, after that, discussed According to the results, except for trust, all other forms of social capital (formal institution, informal network, cooperation) are positively associated with people’s subjective satisfaction Some factors such as relative income, children, the number of years living in the community are somehow estimated to amplify the effects of social capital (in forms of network, trust) on people’s satisfaction Robustness test is finally discussed

4.1 DESCRIPTIVE STATISTICS

According to table 4.1, the average number of formal institutions that a household joins is 0.83, which means most of people join at least one association in their daily life There is someone who does not join any associations and there is someone who joins four organizations at the most A respondent has nearly five friends who can help them by average But still there is someone who does not have any friends that they trust while there

is someone who has up to fifty friends who they can ask for help More than 80% of respondents believe that most people are honest and can be trusted However, most of households are not cooperative since they choose to work alone instead of cooperating with others

In addition to social capital, the independent variables also include demographic variables such as gender, age, education level, marital status, children, income and the years

of living in the residence which reflects the mobility of households Most of the participants answering the question of satisfaction are male, and at the age of around 50, married and have two children by average Around 14% of participants pursue a short-time vocational training, which indicates the low level of education Besides, the variable Illdays is added to illustrate health of participants Mean of the number of ill days is nearly twelve days within the year, demonstrating that one is sick in at least one day per month Finally, the average satisfaction in a province is created adding to the model so as to test whether the respondents feel happier if the people who are around them are happier than others in the other communes

Table 4 1: Descriptive statistics

Source: Author’s summarization

The average score for the level of subjective satisfaction is approximately 1.36, which

is somehow closely to the level of not very pleased Around 8% of observations are not at all pleased with their life, 51.10% are not rather pleased and 37.69% are rather pleased while only 3.48% of them feel very pleased with their life Lam Dong is the province which has the highest level of satisfaction with level of 1.73, nearly “rather pleased” level The percentage of “very pleased” and “rather pleased” level account for 9.59% and 54.79% respectively in this province On the other hand, Lai Chau is the place where most of people feel not very pleased with their life, with the lowest level of satisfaction 1.11 In Lai Chau,

even noone feels good with their life while 80.41% feel not very pleased

Figure 4 1: Average life-satisfaction in twelve provinces in 2014

Source: Author’s summarization

As can be seen from table 4.2, all independent variables including social capital and demographic variables are significantly correlated with individual’s self-rated satisfaction except for variable trust Therefore, it is predicted that probably trust will have no effect on individual’s satisfaction in this study

Figure 4.2, figure 4.3, figure 4.4 and figure 4.5 illustrate the relationship between satisfaction and social network, formal institution, cooperation and trust respectively While

life-it is easy to see the poslife-itive tendency between satisfaction and social network, and formal institution, there is no clear evidence about the association between satisfaction and trust, and cooperation in figure 4.4 and figure 4.5

0 0,5 1 1,5 2

Khánh Hòa

Lai Châu ĐồngLâm

Lào Cai

Long An

Nghệ An

Phú Thọ Quảng Nam

Percentage of satisfaction level and average happiness

Not at all pleased Not very pleased Rather pleased Very pleased Average happiness

Table 4 2: Correlation coefficients among key variables

Formal institution Trust Cooperation

Totalincome 0.2345*** 0.0668*** -0.0189 -0.0147 -0.0208 Relativeincome 0.2302*** 0.0529*** -0.0042 -0.0297 -0.0044 Yearsliving 0.0514*** -0.0069 0.0455*** 0.0629*** -0.0315* Averagesatisfaction 0.2242*** 0.0712*** -0.0559*** -0.0213 0.03

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Figure 4 2: The relationship between social network and life-satisfaction

Figure 4 3: The relationship between formal institution and life-satisfaction

Figure 4 4: The relationship between cooperation and life-satisfaction

Figure 4 5: The relationship between trust and life-satisfaction

4.2 RESULTS OF SOCIAL CAPITAL INVESTMENT MODEL

In order to test the model of social capital investment, Ordinary Least Squares is applied for continuous variables (network and formal institution); Probit model is used for dummy variables (cooperation and trust) to examine the effects of social capital and the demographic characteristics of individuals Table 4.3 shows that the estimation results proving that the empirical work is somehow consistent with the theoretical predictions The

coefficients for age and age square are statistically significant with a positive and negative value, respectively for only one measure of social capital – formal institution It means that social capital in form of formal association first increases with age, then reaches the peak and starts to fall It may be explained by the fact that people join many organizations when they are young and gain social capital, benefits from those organizations, but their formal network is reduced when they get older This inverted U-shape relationship supports the theoretical predictions However, the pattern is totally inverse for coefficients between age and cooperation People seem to be more cooperative when they get older The coefficients for age and social capital in form of trust and informal network are all insignificant in this test

In regards to gender, male respondents are expected to have lower level of social capital, but the coefficient is not significant in this study On the other hand, children variable has a statistically negative effect on social capital in form of informal network, and formal institution People with more children have less time for getting social capital in these forms because they may spend most of their time earning money to raise the whole family The relationships between children and social capital in form of trust and cooperation are insignificant

There is a significantly positive relation between the number of years the household stays in the commune and social capital (formal institution and trust), which supports the theoretical prediction that social capital depreciates due to relocation It may come from the fact the living in the same place for a long time will encourage an individual’s trust in people and social capital in form of formal institution since he or she joins more associations in the community during the long period of living time

Regarding the opportunity of time, represented by absolute income and relative income, the absolute income is positively correlated with trust and informal network, however since the parameters of informal network, trust and relative income are significantly negative, it is said that people who have relatively higher income than the others in their province have more opportunity cost, hence they have less time for gaining social capital in forms of informal network and trust, which appears to be consistent with the theoretical model However, in this research, people who have a relatively higher income get more social capital in forms of cooperation and formal institution as the coefficients of cooperation, formal institution and relative income are significantly positive though the

coefficients of these forms and total income are significantly negative Presumably, when an individual who are relatively richer than others, they have a better level of cooperativeness and a strong will of participating in more organizations

As for marital status, only married people are significantly analyzed to invest in stock

of social capital in forms of network and formal institution It may because married couples can have a wider network by meeting the partner’s friends, relatives or attending some other organizations for couples However, there is no impact between the other marital status and social capital

The significant and positive coefficients of education level indicate that individuals who have higher education level have higher level of social capital, especially the ones who graduate from university because coefficients between the variable university and informal, formal network and cooperation are all statistically positive

The estimation results are overall consistent with the theoretical predictions from the model of social capital investment by Glaeser et al (2002) In form of formal institution, the level of social capital has inverted U-shaped associations with age Generally, people who have higher education level enjoy higher stocks of social capital in Also, people with less mobility have more level of social capital, which assures the importance of the community

in the effects of social capital Regarding the cost of opportunity, depending on the form of social capital, sometimes an individual with higher income has less social capital but still they can gain social capital in other forms