An explanation for vietnams recent migration pattern based on cumulative causation theory and relative deprivation theory

UNIVERSITY OF ECONOMICS ERASMUS UNVERSITY ROTTERDAM HO CHI MINH CITY INSTITUTE OF SOCIAL STUDIES VIETNAM THE NETHERLANDS VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECO

UNIVERSITY OF ECONOMICS ERASMUS UNVERSITY ROTTERDAM

HO CHI MINH CITY INSTITUTE OF SOCIAL STUDIES

VIETNAM THE NETHERLANDS

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

AN EXPLANATION FOR VIETNAM’S

RECENT MIGRATION PATTERN BASED ON

CUMULATIVE CAUSATION THEORY AND

RELATIVE DEPRIVATION THEORY

BY

Vu Thu Ha

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, DECEMBER 2017

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

AN EXPLANATION FOR VIETNAM’S

RECENT MIGRATION PATTERN BASED ON CUMULATIVE CAUSATION THEORY AND RELATIVE DEPRIVATION THEORY

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

Vu Thu Ha

Academic Supervisor:

Prof Dr Nguyen Trong Hoai

HO CHI MINH CITY, DECEMBER 2017

DECLARATION

I hereby declare that this thesis paper entitled “An explanation for Vietnam’s recent migration pattern based on cumulative causation theory and relative deprivation theory” has been written by me, representing the result of my own work, in compliance with the guidelines and disciplines of Vietnam Netherlands Programme The contribution of my supervisor and collaboration with others, specified in the acknowledgement, to fulfill this research are consistent with normal supervisory practice This dissertation, or any substantial part of this dissertation, has not been submitted for any previous qualifications or any other institutions Date: 01 December 2017

It is impossible for me to finish my thesis without the guidance from my supervisor, Prof Dr Nguyen Trong Hoai, knowledges provided by Vietnam Netherlands Programme, help from my classmates and supports from my family First and foremost, I would like to send my special thanks to my supervisor, Prof Dr Nguyen Trong Hoai, who is always ready to steer me to the right direction,

to give me useful advices when I have trouble with my research and to motivate me

to explore my ability I am so grateful for all of his dedication and devoted attention

in every single step in my journey of conducting this research Having a chance to work with him is really an unforgettable experience in my life

I would also like to thank Dr Pham Khanh Nam, Dr Truong Dang Thuy and all other lectures and mentors from Vietnam Netherlands Programme for their valuable lessons, guidance and knowledge that they shared to me throughout the program I appreciate all of wonderful experience and fantastic friends in class VNP22 that I’m luckily obtained thanks to attaining this course

Last but not least, I must express my profound gratitude to my family for providing me unconditional supports and unceasing encouragement that I can pursue my learning desire

From bottom of my heart, I want to say thank you all

This paper combines cumulative causation theory and relative deprivation theory to explain incentives and influences on decision to send migrants of a household, concerning interaction of the household’s characteristics, especially factors related to income, and the features of the origin community Two subsets of data from two successive Vietnam Household Living Standards Surveys (VHLSS)

in period from 2010 to 2014 are established and employed to avoid endogeneity in the analysis of migration pattern in Vietnam in 2012 and 2014 By using logit method and combination of difference-in-difference (DID) and propensity matching score method (PSM), we find some interesting results that have never been found in previous empirical studies for Vietnam case First, the competitiveness competence level of the origin province may reduce possibility of work migrants Furthermore, relative income status before migration event, represented by ratio of income per capita of a household over the average income per capita in the origin province, may positively affect the possibility of sending migrants of the household Then, returns from migration, could create a significantly positive effect on the relative income status of the household at the origin via remittance channel, generating an automatic mechanism in migration progress as expected in the cumulative causation theory

Keywords: migration, cumulative causation theory, relative deprivation, remittance, relative income, PCI

JEL classification: O15, F24, D31

TABLE OF CONTENTS

I INTRODUCTION 1

1.1 Problem statement 1

1.2 Research questions 3

1.3 Research objectives 6

1.4 Data and methods 7

1.5 Thesis structure 7

II LITERATURE REVIEWS 8

2.1 Migration theory reviews 8

2.2 Relative deprivation theory 15

2.3 Remittance – the link between the migrants and their origin 17

2.4 Conceptual framework 18

2.5 Empirical studies 21

III METHODOLOGY 26

3.1 The dataset 26

3.2 Main applicable concepts 27

3.3 Stage one: examining determinants of migration 29

3.3.1 Model 29

3.3.2 Variables 31

3.4 Stage two: testing effects of migration 32

IV RESULTS AND DISCUSSION 35

4.1 Determinants for migration in Vietnam in 2012 and 2014 35

4.2 Impacts of migration on income and expenditure in Vietnam in 2012 and 2014 50 V CONCLUSION 60

5.1 Main findings 60

5.2 Policy implication 62

5.3 Limitation of the study 63

REFERENCES 64

APPENDIX 68

Annex 1: Descriptive statistics of explanatories in year 2010 68

LIST OF TABLES

Table 1: Logit regressions of determinants of migration in period 2010-2012 36

Table 2: Logit regressions of determinants of migration in period 2012 -2014 42

Table 3: Impact of migration on income and expenditure in 2012 52

Table 4: Impact of migration on income and expenditure in 2014 55

ABBREVIATION

DID Difference-in-Difference

PCI Provincial Competitive Index

PSM Propensity Matching Score

US or USA The United States of America

VHLSS Vietnam Household Living Standards Survey VMS Vietnam Migration Survey

VNDHS Viet Nam Demographic and Health Survey VNMHS Viet Nam Migration and Health Survey WTO World Trade Organization

2015 This fact challenges authorizes and policy-makers in the European Union to take pr ompt actions and provide relevant policy to better manage immigration issues and protect the living standard of native people In the United States of America, one of the most powerful country in the world, the new president, Donald Trump, takes office, pledging a fasten immigration policy in the US that including a crackdown on illegal immigration, construction of border walls, reductions in refugee and immigrant admissions, and greater screening of newcomers All of these facts and actions in the last recent years illustrate for the increasing importance meaning of migration problem in association with the socio-economic situation.

The report of the World Trade Organization (WTO, 2008) notices the first explosion of migration happened in the 1950s as the consequence of the World Wars, then gradually increasing before booming in the latter half of the 20th century.Castles, De Haas, and Miller (2013) describe the general patterns of International population movements in the modern world as globalization of migration relating to the diversification of sending countries and the spread in spectrums of economic, social and cultural background; having changes in the direction of dominant

migration flows overtime; and growing the politicization of migration with more and more intervention from the politics relationships

As migration becomes an increasing importance factor in the development process, investigation on the incentives and influences of migration to human beings is necessary Many researchers are attracted to find its characteristics, origin, mechanism and impacts (Harris & Todaro, 1970; Katz & Stark, 1986; Lee, 1966; Lewis, 1954; Ranis & Fei, 1961; Ravenstein, 1889; Sjaastad, 1962; Stark, 1984, 1991; Stark & Levhari, 1982; Stark & Taylor, 1989; L Taylor, 1986; Todaro, 1969) Generally, there are three main schools of theory on migration, called macro-level, micro-level and meso-level regarding different perspectives of views (Hagen‐Zanker, 2008) A lot of empirical studies on determinants and effects of migration has been conducted, especially regarding patterns in the two biggest countries in the world including the USA (Douglas S Massey et al., 1993; Stark & Taylor, 1989) and China (Sicular, Ximing, Gustafsson, & Shi, 2007; Zang, 1998; Zhao, 1999; Zhu, 2002) Although each school of theory has specific perspective of views on migration, their common objectives are revealing the determinants and effects of this phenomenon While classical theories generally concentrate on the economic viewpoint, the modern theories expand the scope of research to other social aspects, including networks, migration institutions and relative deprivation (Hagen‐Zanker, 2008)

In Vietnam, the flows of migration have been dramatically increased through times, especially since the country became independence and adopted renovation policies in the 1980s As reported in The World Bank data from 1960 to 2015, the number of Vietnamese people migrated internationally in 1975 was about 4,661 individuals, starting to rocket nearly ten times in 1980, reducing slightly in the late

of the 80s before rising rapidly to reach the volume of 72,793 international migrants

in 2015 Interestingly, in terms of fraction of international migration in the total population, the World Bank data presents the migration pattern in Vietnam having

percentage points These facts suggest for rousingly remarkable influences of

international migration to the socio-economic circumstance in Vietnam

Apart from international migration, the phenomenon of internal migration in Vietnam in recent years is also noticeable According to Vietnam Labor and Employment Survey conducted by the General Statistics Office of Vietnam (GSO), the amount of internal migrants aged 15 and over has exceeded 1 million in 2015 and 2016, about 1.245 million and 1.037 million or 1.8% and 1.5% of the population respectively, much higher than the volume in the previous period 2012 -

2014 with approximately 832 thousands people aged 15 and over changed their place of residence or 1.2% of the whole population

1.2 Research questions

Despite of description by the World Bank as a dynamic country in East Asia region with strong economic growth, remarkable mitigation of poverty and other achievements in social improvements, Vietnam is facing with variety of difficulties and challenges of an emerging country, especially fragility of poverty alleviation and limitations in labor productivity Apart from application of advanced production models in different scales, the facts and figures on rising of migrants in all over Vietnam suggest that migration has been selected by the authorities, many people and households with the expectation of enhancing their income and living standards

In the 1970s and 1980s, with the aims of exploring new opportunities of cultivation and reallocating population across the country, Vietnamese government carried out migration policies to send a remarkable number of Red River Delta inhabitants up to Northern areas or downward to the Mekong River Delta and Central Highlands, generating motivation for movement wave in Vietnam (Cu, 2005) Cu (2005) indicates that the government-planned migration not only brings many noticeable enhancement for migrants and their family in terms of employment, income and housing but also creates remarkable changing in economic

structure in local and entire of Vietnam As a consequence, attraction of fruitful outcomes from migration encourages other residents to participate into this trend, creating internal migration boom in many provinces in Vietnam After that, at the beginning years of 21st century, when Vietnam having higher integration with other countries and areas around the world via opening the economy, actively participating in regional and international organizations, Vietnamese authorities continuingly employ advantage in human resource of the country by promoting international migration In particular, Vietnam rapidly becomes a major source of unskilled workers for high-income East Asian countries, including Japan, Korea, Taiwan and Malaysia (Ishizuka, 2013) In 2009, a project supporting poor districts

in promotion of labor export for purpose of mitigating poverty sustainably in phase

of 2009 - 2020 is approved by the Prime Minister The main target of this project is

to send about 120,000 workers from 61 districts abroad for working; besides, poor and ethnic minority workers is fully funded This fact implies high expectation of Vietnamese policy makers for outcomes of migration to improve living standards of Vietnamese people

Migration, in general and specifically in Vietnam, may have influences on variety of aspects at different levels, including well-being of migrants and their family and societies at the origin and the destination, in both positive and negative ways (Cu, 2005; Dang, Tacoli, & Hoang, 2003; Ishizuka, 2013) On the one hand, a huge number of unemployed and poor people, particularly in rural or less developed regions, could get jobs with better earnings via migration Simultaneously, labor force in urban and more developed regions is supplemented, which plays as an additional motivation for economic development in these areas Hence, the economic structure that contains resources and industries in both urban and rural areas could be shifted towards more specialized, modernized and more efficient Additionally, thanks to migrants, there are more chances to exchange the cultures of the origin and the destination, triggering diversity and enrichment in cultures However, there is existence of potentially negative impacts on movement of people

First, migration could be a reason for lacks of rural labors in particular occasions, such as on harvest, causing temporary imbalance and changes in allocation of labor resource within the family, which may have significant effects on agriculture, traditional occupations and other activities in the rural area Second, the rapid increase of immigrants in big cities, such as Ha Noi, Ho Chi Minh City or Binh Duong, could trigger high pressures in terms of providing social services at the destination, especially for accommodation and health services Furthermore, the development of rural-to-urban migration may lead to pollution, overcrowded and traffic jams at the destination Thus, the flow of migration may increase pressure on the authorities concerning management of social order and assurance of social welfares.

Therefore, along with the rapid development of migration phenomenon, many researchers pay attention into investigation of determinants and influence of migration in Vietnam (Coxhead, Nguyen, & Vu, 2015; Dang, 1999, 2001; Dang, Goldstein, & McNally, 1997; Dang & Nguyen, 2006; Dang et al., 2003; Fukase, 2014; T P Nguyen, Tran, Nguyen, & Oostendorp, 2008) However, most of these studies concentrate in the receiving place rather than in the sending regions Besides, previous determinants of migration in Vietnam are mainly defined directly

in absolute value, instead of relative comparison Furthermore, the socioeconomic circumstance changes very often, data need to be updated Therefore, a fresher look

is needed to shed a new light on the picture of migration in Vietnam, concerning the migration decision in not only characteristics of migrants but also their relations with the mutual interests of the household as well as the linkages of the household sending emigrants and the surrounding circumstance at the origin

Furthermore, circumstance of the neighborhood has great influences on their members as the control environment for all the group members (Wilson, 2012) In the book The truly disadvantaged (Wilson, 2012), the vicious cycle of poor education, high unemployment and low income is assumed to exist in the family,

joblessness will suffer a social insolation which blocks them out of the job network

As the consequence, they have less chance to get a good job or gain high salary, the education and other social welfare will seldom receive sufficient attention and investment Therefore, hardly do the awareness, linguistic and work related skills of people in these communities as good as in places with high level of education and income Moreover, the socioeconomic situation itself generates depressed attitude for people and disincentives them to escape from the hardship Neighborhood effects is also revealed in terms of game theory (Akerlof, 1997) and consumption externalities, referred as phenomenon “keeping up with Jonese” or jealousy (Dupor

& Liu, 2003) Hence, it is possible that decision to send migrants of household may

be significantly affected by their position in their commune, especially concerning wealthy or living standards

Derived from these concerns, two questions are arousing up for Vietnam situation:

 Is there any relationship between the movement decision of household and their relative income status at the origin community as in the hypothesis of relative deprivation?

 Do the returns from migration create reversed effects on the relative income status in the reference group as expected in the cumulative causation theory?

1.3 Research objectives

The main objectives of this research are to expand investigation on the possible determinants of migration patterns in Vietnam, in terms of relative income with the origin community, and then to determine the feedback of movement action

to that community In this paper, migration is expected to be motivated by the relative deprivation in income, after that operates as a self-feeding process in accordance with the cumulative causation theory of migration

This research aims to contribute an empirical study in literatures of cumulative causation theory on migration and relative deprivation theory The potential results

of the research are expected to provide some helpful information to policy maker in Vietnam and other emerging countries dealing with increasingly complicated migration issues

1.4 Data and methods

The data collected in Vietnam Household Living Standard Survey (VHLSS) from more than 9000 households all around Vietnam in year 2010, 2012 and 2014 will be employed in quantitative assessments Because different purposes may have different motivation and actions, in order to specifically response for the former question, dummy variables for existence of work, non-work and general migrants in household will play as dependent variables in three different logit regressions The explanatories will contain general characteristics of the household, the commune, the sending province and the relative income per capita of household in comparison with the average income per capita of the sending province, as the origin community Subsequently, the answer for the latter query will be found by combined application of the propensity matching score (PSM) technique and the difference-in-differences (DID) approach, that comparing the relative income status

of the two households which have almost similar features, except from migration decision

1.5 Thesis structure

The remaining of this paper would like to be organized as following After general introduction on the research problem in the first section, the second section provides reviews on basic theories and previously outstanding empirical studies In Section 3, data descriptions and applicable methodology are demonstrated Next, the representation of the results and discussion are coming in Section 4 The final section concludes the paper by providing the overall summary, some policy implications and suggestions for further researches

II LITERATURE REVIEWS

This paper approaches to migration field through two streams of literature: one

is the cumulative causation migration theory and the other is the relative deprivation theory, in combination with the impact mechanism of migration via remittance The main ideas of these schools of theory will be summarized below However, rather than mechanically employing the theoretical ideas to Vietnamese case, I suppose to use these perspectives as guidelines to generate and test hypotheses to explain recent migration pattern in Vietnam

2.1 Migration theory reviews

With the aim of having deeper look on migration phenomenon, some reviews

on theoretical literature of migration is needed

From different perspectives of views as well as looking at different aspects of migration, there is variety of definitions for this terminology The Oxford dictionary

defines “migration” as “Movement of people to a new area or country in order to find work or better living conditions” With National Geographic Organization, the

human migration is the movement in accordance with the purpose of permanent, semi-permanent or seasonal residence Migrant can be voluntary or be forced to move In other words, more specific in socio-economic area, the term “migrant” is

referred to “person who is to be engaged, is engaged or has been engaged in a remunerated activity in a State of which he or she is not a national” in the United

Nations Convention on the Rights of Migrants (1998) Through different expressions, “migration”, generally, can be understood as an action of people to move from one place to another with the aims of working or residence Due to this movement changes the living circumstance of people, it is expected to have influences on the human life in various aspects

The early dawn of the migration theory starts with Sir Ernest George Ravenstein (1889) The first theoretical perception on relationship between distance and propensity to move, which is also known as “seven laws of migration”, is

considered as the empirical rules of thumb on migration literature These laws are general conclusions based on his observations in the 19th century; however, most of them still have specifically significant meaning in modern society The laws of migration state that (1) the majority prefers migration in short distances than long distance and towards larger centers; thus, (2) displacement and development process shall occur in both sending and destination regions; (3) the process of dispersion and absorption relate to each other; (4) migration chain develops overtime; (5) the movement could exit when having balance in commerce and industry; (6) urban residents have less motivation to migrate than rural people; and (7) so do the females Along with the explosion of migration in the 1950s, the vast of economic migration theories has been developed to explain incentives of migration; analyze and predict effects of this phenomenon to human being

In general, the theories of migration can be divided into three main schools, including macro-level, micro-level and meso-level While in macro-level, the migrant is considered in terms of interaction between aggregate demand and supply

of labor; micro-level theories explain migration as an individual decision The meso-level is connection between the two other schools, where it is believed that migration decisions are highly affected by the living context

Hagen‐Zanker (2008) visualizes the framework of migration decision making Figure 1 briefly shows possible causes and effects of migration; in which the causing factors are classified into three levels from micro to meso and macro The micro level factors, which are able to be incentives of migration decision, consist of individual and household’s characteristics, including demographic information, factors related to value such as income and risks, and differences in needs of power (structural tensions) and prestige (nominal tensions) On the macro level, migration

is considered as a consequence of interaction between demand and supply for labor

of the nation, effects of political institutions such as migration policies and laws, or development of integration and globalization which may change industrial and

more likely to be motivated by socio-economic context, for example, existence of

migration networks, or relative deprivation or institution promoting migration in the

origin community Furthermore, there are many effects of migration which may aslo

send feedback to selection of future migrants For illustration, remittance as a main

outcome of migration, may improve well-being of receivers at the origin and better

finance for costs of further movements On the other point of view, experience of

previous migrants may be lively evidence for migrant institutions, social capital and

networks to expand their influences Additionally, cumulative and circular migration

refers to the self-feeding process of migration over time once it could commence

These above determinants will be discussed in more details in the following

literature reviews of migration theories

Figure 1: A general framework of migration decision making

Source: Hagen‐Zanker (2008)

 Macro - theories of migration

At first, the macro migration models consider migration decisions in relation

with “pull” and “push” factors These factors are the economic characteristics at the

destination and original place, respectively, which are supposed to create incentives

to migrate of people The neoclassical macro migration theory (Harris & Todaro, 1970; Lewis, 1954; Ranis & Fei, 1961) is developed from trade theory, combined with the geographical discrepancy in the supply and demand of labor The dual labor market theory (Piore, 1979) defines migration as the consequence of a pull factor in receiving countries, called structural labor demand This theory concerns to the social – political - cultural structures at the destination, where people work not only for income but also for social status and prestige Based on these models, the direction of migration can be determined; however, the weakest point of this model stays on its assumption of the equilibrium point of wages, which cannot be found in reality, leading to the consequence that hardly can the volume of migration be predicted

On another perspective of view, Wallerstein (1974) builds up the world systems theory which argues that international migration goes along with the political and economic globalization Under the influence of trade liberalization, the movement of capital and natural resources generates flows of migrations (Massey, 1989) Linked with the connection between countries, Portes, Walton, and Labor (1981) indicate that international migration is highly possible between past colonial countries and their former colonies because of the early connection in cultural, linguistic and so on

Basically, the above theories basically analyze contextual structures of aggregate level to explain the event of migration in one point of time However, there are arguments that migration should be considered as a dynamic process (Boyd, 1989; Fawcett, 1989) In order to further support for this point of view, the concept of migration is narrow from global level in system theory into migrant networks in the network theory (Fawcett, 1989; Douglas S Massey, 1990; L Taylor, 1986)

In general, macro-level theories of migration focus on relation and interaction

of countries more than people to identify the direction and development of

migration Nevertheless, since migration is an action of human, it would be determined by decision making mechanism concerning the human behavior in connection with their context situation Thus, the individual factors and interaction

of individual with intermediate structure such as family, community are likely a remarkable problem of the macro theories on migration, leading to development of micro - and meso - level literature

 Micro - theories of migration

Besides, a school of theory paralleled with the macro theory of migration is microeconomic model of individual decision making (Lee, 1966; Sjaastad, 1962; Todaro, 1969) In this scheme, migration is considered as a personally rational choice after assessment of costs and benefits related to this decision Lee (1966) is a pioneer in formulation of a push-pull framework, evaluating positive and negative factors on migration at both the supply and demand side of migration The neoclassical micro-level treated migration as a form on investment in human capital People will move to where they can be most productive; however, in order

to capture the higher earnings, they are supposed to involve in some investments, such as costs of traveling, effort and opportunity cost in learning new language, getting acquaintance with new culture and applying for a new job (Douglas S Massey et al., 1993) There are many variations of model can be applied in cost-benefit calculation of this approach, for example, probability of legally acceptance

in the area of destination, probability of employment at the origin and the destination, possible earnings in the destination and the cost of movement If the total benefit is expected to dominate the total cost of moving, then the migration decision will be taken into action, except from other preferences

Notwithstanding the human capital approach is useful in explanation of migration decision making behavior of people, it is hard to examine this method In fact, not every people exactly show their rationality, and barely does such kind of decision have effects on only the actor, but other members in their family or

community The structural influencing factors are mentioned as external factors; thus, the interaction between structural opportunities and individual is still ignored

 Meso – theories of migration

The latest school of migration literature that appears as a bridge connecting the two other strands is meso–level theory Despite having different branches, meso-level theory generally determines migration as an interpersonal or inter-group decision, and the basic assumption of this scheme is continuing relationship between potential migration group and other social structures (Schiller & Faist, 2010)

After some milestones in migration perception of family decision of Sandell (1977) and Mincer (1978), a new stage of migration theory begins, called “The New Economics of Labor Migration” (NELM) The outstanding features of the new approach are not only at the decision-makers but also a broader view of migration effects The family may not migrate together; but each member’s movement decision should be considered for the wellbeing of the family as a whole The attribute of migration action are supposed not only to maximize expected income or social status; but also to minimize risks and to loosen restrictions related to market failures of capital markets, insurances or governmental programs (Katz & Stark, 1986; Stark, 1984, 1991; Stark & Levhari, 1982; Stark & Taylor, 1989; L Taylor, 1986)

This approach is likely a remarkable advance in migration theory J E Taylor and Fletcher (2001) declare that the NELM is the only migration theory that explicit the connection between migration and impact of migration via remittance Moreover, this approach investigates both absolute income of household and relative income in comparison with a reference group to reveal how implicit social status incentives migration (Stark, 1991; Stark & Taylor, 1989) According to the NELM, the motivation and influence of migration imply the possible effects of

government policies and economic changes concerning income distribution on the relative deprivation of household income

Afterwards, with the increasing development of network theories and expansion of the NELM, another school of migration theory has been formulated and improved to interpret the internal dynamics of migration, called cumulative causation Cumulative causation refers to the interaction process between individual behavior and community structure that migration may contribute to alteration of the social context, then likely facilitating the additional movement over time (Douglas

S Massey, 1990; Stark & Taylor, 1989; Stark, Taylor, & Yitzhaki, 1986; J E Taylor, 1992)

One of the most famous researchers, having crucial contributions in foundation of cumulative causation migration theory, is Douglas S Massey who explains the patterns of migration in connection with the networks which links people and household with larger structures; then creates increasing effects over space and time (Douglas S Massey, 1990) Migration is supposed to be facilitated via two main manners First, the risks and costs of migration would be lower for potential migrants thanks to the network connections formed after the pioneer movement Second, migration would be considered as a household strategy to increase the household’s absolute income by receiving remittance; to diversify risks

as adding one more source of income; and furthermore, improve their economic position in comparison with the reference group The fact of the U.S immigration from Mexico is employed as empirical evidence for his arguments After that, Douglas S Massey et al (1993) point out six socio-economic factors that potentially affected by migration, including: (1) the distribution of income; (2) the distribution of land; (3) the organization of agriculture production; (4) culture; (5) the regional distribution of human capital; regional distribution of human capital; and (6) the social labeling

socio-Adding up all the ideas of meso-theories on migration, especially the cumulative causation, there are some noticeable points on migration as following First, migration should be considered as a decision a group sharing the common costs and benefits, for example a family, although that action may not be performed

by all the members Second, migration is not only a way to improve absolute income but also plays as insurance strategy for household to mitigate the dependence of the household on the socio-economic context at the origin Third, the social structure in the sending community may create incentives to migrate and migration can alter the context of that community afterwards, causing the expansion

of migration in momentum In particular, the distribution of income of the origin could generate motivation for emigrants before their movement sends feedback to change features related to income at the origin, including local consumption, local production, living standards and social label

As this branch of theory is established later than the NELM as well as other theory in micro- or macro-level, its development is still as not completely as the others For example, the limits of cumulative causation have been questioned concerning in demographic characteristics of different structural contexts The function of principal mechanisms is argued only having effects in small cities, rural town and villages, but not in large urban areas of Mexico (Fussell & Massey, 2004) Overall, there are not many empirical tests conducted in association with this literature, especially in an emerging country like Vietnam, hence, this research will employ cumulative causation theory as theoretical frame to investigate the relationship between relative income of household and migration in Vietnam

2.2 Relative deprivation theory

It is a common saying that the grass is always greener on the other side of the fence To analyze different behaviors of people in conjunction with relative income

in a reference group, the theory of relative deprivation should be mentioned This theory concerns the feelings or experiences of an entity under comparison the

situation between that entity and the rest of society Relative deprivation may refer

to the dissatisfaction of an individual when being deprived of something that he believes of his entitlement himself

One of the first formal definitions of the term “relative deprivation” is launched out by a British historical sociologist named Garry Runciman Runciman (1966) describes four preconditions for an individual A to feel relatively deprived of

an object X as following:

(i) Person A does not have X;

(ii) He sees some other persons (possibly himself at some other points of time as having X (whether or not that is or will be in fact the case);

(iii) He wants X; and

(iv) He believes that it is feasible to obtain X in reality

It is notable that the feeling of deprivation comes from conditions (i) and (iii), while the relativity of the concept is presented by the others

Supposed that there are two households having the same level of absolute income but they are living in two different villages One village is more prosperous than the other Hence, in consideration of the village’s income distribution, the household in the richer village is supposed to have lower rank than its counterpart

If household utility is a function of not only absolute income by also income position compared to the other households in the community According to the theory of relative deprivation, this household is more likely to have a stronger motivation to improve their position status than the other in order to increase their level of relative satisfaction Assuming that migration is an effective mean to enhance the family’s income, the possibility that the former commits to migration should be higher than that of the latter, ceteris paribus

The concept of relative deprivation has been employed into the cumulative causation migration theory to explain the incentives of international migration decision by Stark and Taylor (1989) Based on data of Mexico-to-US migration, Stark and Taylor (1989) find evidence for a significant role of relative deprivation

in migration decisions, and the effectiveness of migration strategy in gains of household not only in terms of absolute income but also enhancement in family’s income position relative to other in the reference group reversely They argue that the awareness of members (or households) in the migrant-sending community about the income differences is an important factor affecting migration; hence, there is, perhaps, a higher incentive to emigrate in societies that experience much economic inequality In the short run, remittances may increase inequality, but in the long run, the disparity may reduce

In addition, the literature provides no consensus on the effect of remittances

on inequality at the origin Some studies (Docquier & Rapoport, 2003; Stark, Taylor, & Yitzhaki, 1988) recommend an inverse U-shaped curve representing to the inequality in the sending society with regard to the effect of remittances, instead

of a monotonic line Therefore, this paper is going to apply quadratic function form

to examine variable representing relative income

2.3 Remittance – the link between the migrants and their origin

Each theory of migration is more or less related to the issue of income, in which, remittance plays as an important bridge to connect the migrants at the destination and their relatives at the origin In order to take an overall view on causes and effects of migration, Jones (1998) summarized a system of migration and remittance, which is an expanded version from Russell (1986) The “Migration-Remittance System”, made by Jones (1998), presents not only a long range of possible determinants and consequences of decision making process related to migration for an individual or a household, but also feasible impacts for origin community Particularly, the key factor in this system is the migration selectivity process First, a particular group of population with specific socioeconomic and demographic features will choose to migrate Then, the next decision that migrants should make is how much to remit While the amount of remittances sending back home is indicated that having relation with the migration experience of the migrants

depend on the spending manner of the families Additionally, some studies show that the tendency of expenditure or investment is different between families in accordance with their level of income (Jones, 1995; Russell, 1986; Wiest, 1984) Wiest (1984) and Jones (1995) find out that households with above-average income are more likely to spend outside the origin village or consume imported goods, causing leaking of remittances out of their local; whereas, families with income lower than average are more possible to consume locally Hence, remittances could play much more significant role in money circulation at origin, having more positive impacts on development of the local economy Besides, several researchers have argued that if migrants come from lower-income households, their remittances are expected to mitigate income inequality within household at the origin Otherwise, if they come from upper-income families, income inequality among families in origin community may have propensity to increase Furthermore, factors affected to income at the destination also have vital influence on remittances and inequality, as a consequence (Jones, 1992; Douglas S Massey, 1986) In Vietnam, Niimi, Pham, and Reilly (2009) also find a strong link between migrants at the destination and their family or relatives at home via remittance Thus, with remittance as an effective motor in the migration mechanism, it is optimistic to expect that the relative income status of household sending migrants will be enhanced after having migration and become a catalyst in the self-feeding process

of migration

2.4 Conceptual framework

All things considered, the conceptual framework for the migration pattern in relation with the relative income status of a household and the origin community in this research is sketched as following (See Figure 2)

As presented in the Figure 2, the first stage considers motivation of migration The socio-economic context at the origin, such as infrastructures, available industries and local investment policy, and the demographic characteristics of the

household, for instance, household size, education and skills of household head, will determine the possible production activities of the household at the origin Next, the outcomes of these activities include the income, expenditure and expectation of the family on their living standards In spite of staying at the same region, different demographic characteristics generate different productivity, causing discrepancy in income distribution of household in the community However, as sharing the same socio-economic context, especially in a country with more than two thirds of population living in rural areas like Vietnam (GSO statistics), the households are highly expected to have awareness of living standards of others neighbors in some extent This perception maybe a premise for relative deprivation affect people’s behaviors As acknowledged by contemporary economists, people may put many attention on their relative position by comparing what they possessed with the others, causing higher feasibility of emulation in sense of keeping-up-with-the-Joneses (Luttmer, 2005) An implication of relative deprivation theory is that people could improve their satisfaction if they achieve others’ living standard level From this point of view, combined with assumption that migration may make people better off, migration may help poor households improve their position by increasing income, while members of wealthy households can experience better education or enjoy superior welfare in a more advanced society Hence, possible determinants in making migration decision are socio-economic and demographic status of household at the origin, existence of substitute production activities and living standards or income status of the family at the origin

Then, in the second stage, after decision of moving is made, migrants will have two options that is remitting home or not If there is remittance, this source of income may contribute to allocation of income at the origin via consumption or investment of the family at home In contrast, the cost of movement and lack of working labor at the origin due to migrants, who are not sending remittance back home, may decrease the total earnings of the family, making reduction in relative income position in relation with the initial reference group Although the existence

of remittance depends on many factors, including the purpose of moving, the influence of migrants on the relative income status of their family at hometown is likely to be remarkable Not only that, based on the cumulative causation theory, the impacts of migration via remittance on not only the sending household but also the sending community are expected to encourage the continuation of emigration process

Figure 2: The conceptual framework for the migration pattern in relation with the relative income status of a household and the origin community

Source: Author derived from the literature

2.5 Empirical studies

Besides general impacts of migration for sending and receiving countries as

in poverty mitigation practice in developing countries (Adams & Page, 2005; Deshingkar, 2006; Semyonov & Gorodzeisky, 2005) Before looking at empirical studies in Vietnam, it is necessary to have a glance at migration researches in other neighborhood countries

China is the fastest growing economy in the world that also suffers the highest level of inequality in East Asia (Ducanes & Balisacan, 2005); hence, the Chinese authority has employed migration strategy to relax and control such differences in the population (Chen & Cowell, 2017; Sicular et al., 2007; Zhu, 2002) By statistical analysis using panel data from China Health and Nutrition Survey, Chen and Cowell (2017) point out the evidence for increasing changes of income mobility and inequality overtime Besides, the rise in inequality appears to be accompanied

by an escalation in inequality of opportunity in China (Y Zhang & Eriksson, 2010)

In the mid-1980s, the world observes the largest labor movement flow from rural to urban areas in China; the dramatic income gap between Chinese rural and urban areas is considered as a main incentive of migration; and the Chinese government has to apply longstanding policies to restrict this kind of mobility However, by applying a theoretical model of labor allocation within rural houses in accordance with institutional constraints to reduce migration of the government, Zhao (1999) finds that although big family could provide more migrants, the majority of workers and families do not participate in migration at all; and the Chinese rural families are unwilling to migrate permanently to urban regions In addition, by examination of time series and cross-sectional data in internal migration in China, K H Zhang and Shunfeng (2003) remind that interprovincial migration in China may be encouraged

by the rural-urban income gap and discouraged by their geographical distances, and the volume of intra-province migrants has a positive relationship with the urban-rural income gap and urban population in that province As Vietnam and China sharing many common characteristics in culture, demography and economic development, these empirical results from China case generally suggests for migration pattern in Vietnam, that migrants should be considered as a representative

of the household participating in migration and the connection of migration and income gap is noticeable

Furthermore, in a study of 40 villages in India, Lipton (1980) argues differences in effect of migration on inequality level of different group Migration is likely to enlarge the inequalities intra-rural areas by pulling the better-off migrants toward better jobs and pushing the poor into poverty trap and labor-replacing methods Besides, Indonesia is referred as an illustration for successfully employment of migration as effective strategy to develop rural areas (Liu & Yamauchi, 2013) By using panel data with district-level demographic data in Indonesia, Liu and Yamauchi (2013) check the determinants of migration in with several types of model, such as linear probability model and probit model The estimated results suggest that the more educated, males and young people have higher tendency to migrate; hence, migration in conjunction with urbanization could absorb a huge number of rural population, largely contributing to success of industrialization and improvement of human capita in Indonesia Therefore, it is an optimistic expectation to find evidence for positive impacts of migration on welfares in Vietnam

In Vietnam, previous migration studies often use large-scale surveys, including the 1989 census (Banister, 1992; Dang et al., 1997) , The Viet Nam Demographic and Health Survey (VNDHS) (Groleau, Soulière, & Kirmayer, 2006), the Viet Nam Migration and Health Survey (VNMHS) (Dang & Le, 2001), the Vietnam Migration Survey (VMS) (Fukase, 2014) and the Vietnam Household Living Standards Survey (VHLSS) (Fukase, 2014; M C Nguyen & Winters, 2011) Although all of these surveys are conducted on large scales, the specific samples, respondents or contents of questionnaire shall be differently designed for each survey in equivalent to its main objectives For example, VNDHS mainly focuses

on married women, or both VNMHS and VMS are only conducted in a limited number of provinces normally having high rates of migration By containing very

the Vietnam Household Living Standards Surveys are considered to be the most representative survey for the whole population (T P Nguyen et al., 2008) Hence,

in this study, the data from the most recent VHLSS, from 2010 to 2014, shall be used to investigate the determinants and impacts of migration

Furthermore, a variety of research methodologies are applied in previous migration studies in Vietnam By using macro gravity model, Dang et al (1997) find that migrants are more attracted by highly developed provinces, causing some failure in population movement allocation of the government in the 1980s The similar method is applied in study of Phan and Coxhead (2010) on factors affected

on migration patterns and the role of migration on inter-province income disparity, using data from the 1989 and 1999 censuses Besides, micro approaches are applied

by T P Nguyen et al (2008) to explore factors leading to out – migration for both economic and non-economic reasons and terms of migration, using panel data of VHLSS 2002 and 2004 The regression results from a probit model show the strong link between migration decision and the characteristics of the sending community Additionally, evidence of an inverted U shape relationship between the possibility

of migration and per capita expenditures is indicated Fukase (2014) uses the multinomial logit and conditional logit models to analyze the effects of employment opportunities that foreign-owned firms created and choices of destinations for internal migration, using the data from the VMS and VHLSS in 2004 The research revealed that foreign job opportunities have great attraction for migrants, especially for female Recently, C V Nguyen, Van den Berg, and Lensink (2011) employ panel data of household in VHLSS 2004 and 2006 to estimate the influences of work and non-work migration on families’ welfare, poverty and inequality, applying different – in – different (DID) method combined with propensity matching (PSM) method In this paper, the decision of migration is argued to be decision of the whole family in short- term This argument and methodology is quite suitable with the migration pattern in Vietnam Unlike in financial or investment decisions, migration or not is a decision of a household, related to their

expected costs and benefits, based on their available information; hence, hardly a household could collect a long-range of historical data to analyze a wide range of possible outcomes of this decision The selection process is much more likely related to their preference, experience or influenced by others practices Hence, the same assumption that migration is a short-term decision and application of PSM and DID into event study will be employed in this research

Last but not least, with the purpose of having more understanding about migration pattern in Vietnam, it is necessary to identify some outstanding features

in characteristic and lifestyle of Vietnamese people Vietnam is a country located in South East Asia with favorable features in geographic and nature resources for agriculture, which plays as vital production activities for the whole country through centuries Majority of population live in rural areas, households closely live in villages Laboring activities in conjunction with other social and historical factors create the specific culture and personality of Vietnamese that is notably with patriotism, solidarity and respect attitude to the community (Thêm, 2004) However, the limitations of village culture are presented with decisions by emotion, lack of self-confident and easy to get jealous, that could promote the effects of relative deprivation in decision related to mutual income of the household such as migration

Thus, the above socio-economic features associated with a noticeable situation

of poverty and inequality, simultaneously with increasing trend of migration, the migration pattern in Vietnam is highly expected to follow the cumulative causation migration theory and relative deprivation theory Based on these basic theories, for the sake of investigation the relationship between income disparity and migration trend in Vietnam, the following hypotheses are supposed to be tested: 1) relative position in terms of income intragroup plays a significant role in migration decision

of household, and then 2) migration could send feedback to the that relative income status of the reference group via remittance The latest datasets of Vietnam

econometric and analytical techniques, which will be specifically define in the next section, will be employed Hence, this research is expected to contribute one more empirical work into cumulative causation literature of migration as well as being the first paper applying relative deprivation theory to investigate the relationship of income disparity and migration in household level in Vietnam

III METHODOLOGY

3.1 The dataset

Apart from data for Provincial Competitive Index (PCI) is collected from annual report conducted by Vietnam Chamber of Commerce and Industry, all other variables in this research employs the data from VHLSS 2010, 2012 and 2014 These three sets of data make up two research groups for adjacent years, including

2010 -2012 and 2012-2014, with the sake for test the possible sustainability of the relationship between relative income status of the family and their migration decision The first year in such periods would be referred as the reference year, whereas the second year will be the research year As the Vietnam Household Living Standards Surveys collect information of the enumeration areas within the last 12 months, the separation of reference year and research year is needed to determine characteristics of the individual, households and community related to migration prior to the time migration decision made, avoiding possible bias and technical errors in the analysis

The surveys contain information of different group level, from individual to household and commune, in all of 63 provinces in Vietnam The total number of household participated in 2012 and 2014 surveys is 9399 and 9399, and the 2010 survey is conducted in 9402 households The enumeration areas are randomly designed from master sample population census; however, a half of the sample is rotated by years For instance, the enumeration areas in VHLSS 2014 contain 50% areas surveyed in VHLSS 2012 and the rest are newly selected from the master

the subsample of panel data only contains 4173 and 4217 households, respectively for the former and latter research groups

3.2 Main applicable concepts

 Migrant

As defined in Vietnam Household Living Standards Surveys, household members are those who share accommodation and meals from 6 months or more over the last 12 months and share a pool of incomes and expenditures with others in the family In addition, as regulated in the Labor Code number 10/2012/QH13 approved by Vietnam National Assembly, employees are defined as “people from

15 years old and above, capable of working, working under labor contracts, receiving salaries and subject to the management of the employers” Thus, for transparency, migrants in this research are identified as alive people aged 15 or more, have not lived in the household for more than 06 months more during the past year of the survey Household having migrants is the household that sending at least one migrants outside of the household for any reasons, except from death The variable for household having migrants in this paper will be in form of a dummy variable assigned value 1 (one) if there is at least one member aged 15 or more staying in the household less than six months in 12 months before the survey time, and 0 for the rest

 Two types of migration: work migration and non-work migration

Because there is evidence that initial purpose of migration may affect the possible determinants and later effects of migration in Vietnam (C V Nguyen et al., 2011; T P Nguyen et al., 2008), migrants will be classified based on reasons of movement for deeper investigation of their behaviors In the VHLSS questionnaire, there is a question for reasons of living in the household less than 6 months, including medical treatments, education, work purposes and others Hence, in economic aspect, the purposes of migration might, generally, be classified into two types of work and non-work The number of households sending migrants in 2012

is 1,075, consisting of 208 and 929 cases properly for work and non-work purposes These number in 2014 are 987, 158 and 887, correspondingly Dummies for presence of work and non-work migrants in the household are created by respectively assigning value 1 for subjected household and 0 for the others Since statistical figures show approximately 6% of households having both two types of migration at the same time, this paper will not apply the form of categorical variable for different types of migration

 Relative income status

Concerning the relative income status, despite of lacking of empirical bases for proxy of this conception, the relative comparison between income per capita in household and average income per capita of local province will be employed in this research This representation has inspiration from the idea of Blanchflower, Van Landeghem, and Oswald (2009), examining relative utility of individuals in relation with imitative obesity when their sensation and actions effected by comparisons with others in a society The empirical work on cross-sectional data in 29 European countries shows evidence for existence of imitative keeping-up-with-the-Joneses in body weight occurring among people who cares about the ratio of their weight to the mean weight of the population Additionally, because the household members share a pool of incomes and expenditures with others but the household size is non-homogeneous feature, the living standard in a small household may different from a crowded one; hence, the relative income status of household is better considered on the basis of per capita income Moreover, the province will be considered as the level of reference group or the neighborhood community in this research

Hence, to compare income position of household in their province, accounting for the specific characteristics of different provinces, the relative income status of household i in province j, denoted RYij, will be calculated as following formula:

Where Yi refers the income per capita in household i, and Y j

is the average income per capita of province j where the household i located in

As mentioned above, this paper is going to examine two stages of migration First is to find out possible determinants affects the decision-making process of a household on migration, especially relative income status of the household Then, the second stage is to determine reversal effect of migration on the relative income status of the household Following is descriptions on specific analysis methods applying in each stage

3.3 Stage one: examining determinants of migration

3.3.1 Model

In the first stage, to estimate the impact of relative income status on migration

decision, the logit model for decision selection is employed The dependent variables are dummies of whereas household having migrants or having work migrants or non-work migrants in the research year Each specific dependent variable will be checked in an appropriate model; however, the general form of regression model is supposed to be as below:

Although there is still having controversial discussion on significant level of specific determinants in relationship with migration decision in particular period of time and place, the role of demographic characteristics on decision of sending migrants of a household is undeniable

T P Nguyen et al (2008) find evidence that age of household head , income and expenditure of household may have effects on migration decision in the inverted U-shape form, called “migration hump” For example, the poorest and richest households are observed to have less likelihood of migration Lucas (2005) also discoveries evidence that the proportion of short-term and long-term migrants having concave relationship against per capita household expenditure quintiles The suggested reason is while the poor is supposed to be blocked access to move by high costs of migration, the wealthy is expected having no incentive to migrate Hence, in the logit regression for determinants affecting migration probability, the quadratic form for variables concerning income and age of household head will be added

In a research for migration pattern in China, Zhao (1999) discoveries interesting evidence on migration pattern in China Although China suffered the largest flow of labor movement in the world history in the mid-1980s, the majority

of workers in the family did not involve in migration Furthermore, the migration is found as a temporary decision of the household by examining the responses of household consumption to income from migration Additionally, when considering production function of a household, the theoretical model suggests that more land in the household will reduce the supply of migrant labor, while the household with more labor could provide more migrants Because the socio-economic and political contexts in Vietnam have many similarities to the Chinese context, the same migration pattern is expected

Fukase (2014) indicates young and more educated people could have higher possibility to engage in migration Thus, the demographic characteristics considered

in this research consist of household size and structure, absolute per capita income and consumption of the household (measured in logarithm for normal distribution), age and education of members and assets possessed by the household at the origin

Regarding features of the community, notwithstanding the geographical factors, such as regions and rural, are expected to have influences on migration decision, the socio-economic context of the origin is expected to have a strong relationship with migration in accordance with meso-theories of migration The Provincial Competitive Index (PCI), which is a sophisticated index annually calculated by Vietnam Chamber of Commerce and Industry since 2005, is able to be

a good proxy for the local business environment The overall PCI reflects the economic governance quality of provincial authorities in creating favorable business environment for development in the private sectors The higher PCI is, the higher potential development of local enterprises is, attracting more investors, creating more jobs for locals, and possible reducing emigration Similarly, in the commune dataset, the information on existence of enterprises, factories or traditional occupation villages that local people can commute daily is available If there are such organizations or institutions, the possibility of local employment, as a substitutional mean of income improvement with lower costs, increases; thus, migration, especially work migration, is expected to mitigate

3.3.2 Variables

As stated above, apart from relative income per capita of the household, the independent variables contain two main aspects, including household characteristics and socio-economic features of the origin community All data of these explanatories are collected in the survey in the reference year, which is two years earlier than the data of dependent variables

In the same line with investigation of C V Nguyen et al (2011) and T P Nguyen et al (2008), the household characteristics include dummies for ethnic minorities of the household, gender of household head, marriage status of household head, training and education level of the head and the spouse and the construction status of the house; and some quantitative variables of household size, age of household head, ratio of female members, ratio of members out of labor force, ratio