Impact of remittance on households consumption patterns in vietnam

This study analyzes the impact of remittance on consumption patterns of food, education, health care, education, electricity, durable consumption, and other consumption which means to find the impact on share and level of those consumption categories in Vietnamese households from 2012 to 2016

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

VO THI HUYEN NHI

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

VO THI HUYEN NHI

Hanoi, 2020

ACKNOWLEDGEMENTS

I would never be able to complete this Master thesis without the guidance of my supervisors, committee members, support from my family members, and help from friends

First of all, I am extremely grateful to my supervisors, Dr Vu Hoang Linh for the

thoughtful dedication with timely and crucial comments in completing this study

I am also grateful to Dr Dang Quang Vinh, Professor Akashi Junichi, Professor Nguyen Thuy Anh, and Professor Naohisa Okamoto for your constructive opinions and suggestions

Most importantly, I am indebted to the Public Policy Program and Vietnam Japan University for providing support when I need it most I am deeply indebted to

my family and my colleagues as their support give me the motivation to complete this studying journey

Any inadequacies or errors in this dissertation are entirely my responsibility

TABLE OF CONTENTS

ACKNOWLEDGEMENTS i

LIST OF TABLES iii

LIST OF FIGURES iv

LIST OF ABBREVIATIONS v

CHAPTER 1: INTRODUCTION 1

1.1 Roles and impacts of remittance 1

1.2 The context of migration and remittance in Vietnam 2

1.3 Objectives and Scope of Research 5

1.4 Research questions and hypotheses 5

CHAPTER 2: LITERATURE REVIEW 6

2.1 Theoretical Framework 6

2.1.1 Definitions and determinants of migration and remittance 6

2.1.2 Impacts of remittance on received household and home country 9

2.1.3 Consumption patterns theories 11

2.2 Empirical Studies 13

CHAPTER 3 RESEARCH METHODOLOGY 16

3.1 Model specification 16

3.2 Data description 19

3.2.1 A brief of Vietnam Household Living Standard Survey in 2012, 2014, 2016 19

3.2.2 Data processing 20

3.3 Structure of regression 20

3.4 Hypotheses 24

CHAPTER 4 RESEARCH FINDINGS 26

4.1 Descriptive results 26

4.2 Regression results 33

4.2.1 Regression results of impact on consumption shares 33

4.2.2 Regression results of the impact on consuming level 38

CHAPTER 5 DISCUSSION AND CONCLUSION 42

REFERENCES 45

Table 3.3 Summary of independent variables in the regression model

Table 4.1 Summary of consumption share by types of household

Table 4.2 Summary of consuming level in seven consumption categories

Table 4.3: Observations categorized by each type of household, spending

patterns, and interviewing year

Table 4.4: Ratio of remittances to expenditure by different groups

Table 4.5 Regression result for consumption share

Table 4.6 Regression result for the level of consuming

LIST OF ABBREVIATIONS

FDI Foreign direct investment

KNOMAD Global Knowledge Partnership on Migration and

Development IFAD International Fund for Agricultural Development

ILO International Labor Organization

IMO International Migration Organization

IOM International Organization for Migration

ODA Official development assistance

OECD Organization for Economic Co-operation and Development PSO Provincial Statistic Office

UN FPA

VHLSS

United Nations Population Fund Vietnam Household Living Standards Survey

CHAPTER 1: INTRODUCTION

1.1 Roles and impacts of remittance

Migration and sending remittance seem to be a globally popular phenomenon that contributed to solving many issues of both home and host countries An estimated US$6.5 trillion in remittances will be sent to low- and middle-income countries between 2015 and 2030 Besides that, remittance is a very important capital source for developing countries The trend of remittance tends to increase over time despite the downward and fluctuation of FDI and ODA

Figure 1.1 Remittance Flows to Low- and Middle-Income Countries are larger than

Official Development Assistance and more stable than private capital flows,

1990-2019 Sources: Cited from World Bank staff estimates, World Development Indicators,

and International Monetary (IMF) Balance of Payments Statistics The global business witnessed many unexpected events which increased the risk

of oversea investment and the official borrowing from other countries also ties with political conditions so the ODA and FDI flow could not exceed for a certain amount Besides, the migration also lightens the pressure about job creations of many governments in low- and middle-income countries Most of the international remittance will be used by remittance-receiving families to reach well-being goals of

a family: increased income, better health, and nutrition, educational opportunities, improved housing and sanitation, entrepreneurship, thus it would help to reduce

inequality (IFAD, 2017) Ideally, remittances or any additional source of income potentially give migrants, households, and communities greater freedom to concentrate their activities and to allocate investments to fulfill further goals in long term rather than spending for recurrent spending

Even though, not every household could exploit successfully this resource so it

is easy to recognize that remittance and migration do not naturally lead to more development at the household level and the national level Such differences in a political scheme, social value, and economy of host and home countries have intervened in migrants' and household's behavior which then results in different levels

of consumption categories of migrant families

Remittances might negatively correlate with GDP growth because the shortage

of labor supply, labor market participation, and consumers from emigrants is greater than remittance flow received by left behind households so in this case remittance could not become a source for economic development (Chami et al, 2005) Remittance could be seen as an additional income for households and there are arguments that it was fungible but in some other study shows it has an additional effect on the consuming attitude of households Consumption patterns could reflect the welfare of a household when it tells the preferences of a household and predict partially about the future of the household From the change in household consumption trend, the government could interpret the change of industrial structure which caused remittance flow, household consuming trend would shape the industrial composition in those emerging economics (Kuznets, 1973)

1.2 The context of migration and remittance in Vietnam

Most of the Vietnamese migrants choose developed countries for their migration destination despite the long distances and completely different cultures (see Table 1.1) The United States is the country that attracted the largest Vietnamese migrant stock and it is five times larger than the second host country- Australia The major source of remittance to Vietnam also comes from the United States Some countries have a large number of migrant stocks like Japan or Cambodia but the amount of remittance sending is not equivalent to the rank of migrant stock so it could

be interpreted that average remittance per capita is relatively low compared to other remaining countries Canada is the 3rd contributed remittance to Vietnam but the number of Vietnamese migrants is not so high to be ranked as the top ten host countries of Vietnamese migrants

Table 1.1: Top ten main remittance-sending countries and main host countries of

Host country Migrant stock

Source: Data generated by author from the United Nations database on migrant

stocks and World Bank on remittance updated until 2019 Among the three main foreign financial streams, personal remittance has the largest foreign capital flow in comparison with other flows (Figure 1.2) This flow

also shows a steady increase in contrast with ODA, which refers to the official arrangement of government with the creditors The flow also presents a higher and more constant increase rather than FDI

Figure 1.2: Vietnam—Remittances and other foreign capital inflow, 2000–2017

(US$) Source: Summarized by the authors with: ODA International Development Statistics database; FDI from IMF, Balance of Payments database; Remittance:

KNOMAD database on remittance inflow

In front of the emerging importance of remittance, besides following the global regulation on international migration the regional regulation on migrant worker Government of Viet Nam has issued several legal documents to adapt to the change

of migration trends with the rest of the world International organizations such as IOM, ILO, or UN FPA have vital roles in consulting for the government with regular research and projects to improve the life of migrant and migrant households as well

as protect the security of host and home countries There are many documents from governments that are issued and revised frequently such as Decision No 170/1999/QD-TTg, dated 19 August 1999, and its Amendment No 78/2002/QD-TTg, dated 17 June 2002 and Circulars No 34/2015/TT-NHNN dated December 31,

2015, introduced services received and spending foreign currency The main purposes of these documents are to facilitate remittances from overseas Vietnamese, ensure the interests of the senders and the receivers and expand forms of money transfers

1.3 Objectives and Scope of Research

This study analyzes the impact of remittance on consumption patterns of food, education, health care, education, electricity, durable consumption, and other consumption which means to find the impact on share and level of those consumption

categories in Vietnamese households from 2012 to 2016

1.4 Research questions and hypotheses

Main question: How does international remittance impact on household consumption patterns?

Sub question 1: How does international remittance impact the share of each consumption category?

Sub question 2: How does international remittance impact the level of each consumption category?

Sub question 3: What policy initiative from the findings?

CHAPTER 2: LITERATURE REVIEW

2.1 Theoretical Framework

2.1.1 Definitions and determinants of migration and remittance

Remittance is the result of migration thus obtaining definitions and classification of migration is important to understand the mechanism of sending and utilizing motivation of remittance Although migration is an ancient activity of human beings until the end of the 80s, the role of migration has been discussed as a part of the development process One of the theories in this period is the new economics of labor migration considers migration as part of a household strategy to deal with limitations of finance and other developing opportunities at the home place It is no longer an individual issue and the factors affecting migration decisions or relates only economic aspects, but also non-economic issues such as knowledge and social capital that make migration networks important (Stark & Bloom, 1985)

Migration is a spatial issue involving movements of people from their place to

a new destination by crossing administrative borders for a while Migration could be classified by different dimensions If based on the period, migration would be separated as life-time for those who were born in a place different from their resident address; current migration is for those who migrate to a present place for appropriately 5 years around; frequent (circular) migration is for those who often travel in a particular time Based on the intention to stay, migrants could be recognized as permanent migrants or temporary migrants Meanwhile, based on the coverage area, migration can be grouped into international – for those people who cross the border from one nation to another, internal migration is for those who reside

in another place but still the same nation (Sugiyarto, 2014) IOM (2015) defines an international migrant as a person that for any reason has lived away from the country

of origin for more than a year

There are two main forces of migration to make people and their family's decision to leave their homeland The push factors belong to the economic aspect of home country are poverty/low wages, high taxes, high unemployment, overpopulation while factors belong to non-economic are discrimination, poor health

care, war or oppression, corruption, crime, compulsory military service, natural disaster, and hunger The pull factors belong to the economic aspects from the host country is the demand for labor, high wages, generous welfare benefits, good healthcare, and education system, strong economic growth, advanced technology, low cost of living while the factors of non-economic are family and friends/network

in this place, right and freedoms, property rights, law and order, amenities (Bansak, 2015), 2015) cited in (Simpson, 2017))

Remittance is known as the result of migration when those migrants leave their family behind but then send back some type of resource which could be visible like money, consumer goods, or invisible like the culture or knowledge (IMF, 2009)

In the case of Vietnam, the household living standard survey has already expressed the remittance income from abroad of each household so with Vietnam context this study would apply the definition of Koser (2007) which describes that remittances are the flow of money from migrants abroad to their families in their country of origin

Many factors relate to the willingness of migrants to send back money which

is from individual motivation, family arrangement, and macroeconomic circumstances The most direct determinants on migrant remittance flow depend on the migrant's ability and their instinct, motivation, and relationship with members left behind Working migrants often see themselves as a part of larger decision-making units like their family or commune, their move to a new labor market is typically driven by the interests of that larger unit Remittances present as a bridge to sustain migrant’s social support system with family and community at home countries, which

is common in the Asian context

The readiness to remit is also determined by which type of migration that the migration belong to such as the duration of migration (how long do migrants intend

to stay abroad, temporarily or permanently?), the family situation of migrants (single, married, with or without children?), and network effects (do migrants move alone, with family members, and do they keep attachments to those left behind?) (see (Munshi, 2003) One of the most intuitive motivations for remitting money back

home is what has been characterized in the literature as "altruism": the migrants care about relatives who left in the home country Under an altruistic model, the migrant could enjoy the satisfaction when observing the welfare of his/her relatives is being fulfilled by their contributions This theory could explain some results in an empirical study about poverty reduction and household income First, the number of remittances should increase with the migrant's income Second, the number of remittances should decrease with the domestic income of the family And third, remittances should decrease over time as the connection to the family gradually fading Thus, altruism was found to be insufficient for explaining the motivations to remit, at least for Botswana (Robert E B Lucas, & Stark, O., 1985)

Another motive for remitting money to family members in the home country may be pure self-interest First, a migrant may remit money to his/her parents driven

by the aspiration to inherit, if it is assumed that bequests are conditioned by behavior Second, the ownership of assets in the home area may motivate the migrant to remit money to those left behind, to make sure that they are taking care of those assets This could be explained with the result that the usage of the family is focusing on asset cumulative activities and investment for the migrant

Broader than the motivation of sender-migrants is the force from the family agreement: co-insurance and loan This theory is also named by other authors such as

“tempered altruism” and “enlightened self-interest” Lucas and Stark (1985) The difference of this theory from the two above is not about the person who is the decision-maker but a family who is a determinant unit with remittance being endogenous to the migration process The migrants and remittance are applied as a Pareto-utility maximized strategy to allocate the family and remittance is seen as the payoffs for this process Thus, the family could obtain both objectives: risk-spreading and investment in the education of young family members

Another popular thought in explaining remittance is to assume that the migrants always have the intention to return to their homeland with a certain amount

of savings so they have to negotiate with family in the home country about the share

of the savings

The last layer of determinants is the factor from the macro-factor which is related to the economic performance of both the host and home place The macro-level contains those factors from economic, political stability, and natural factors The economic segment includes real exchange rate inflation, GDP growth, the trade balance in both host and home countries (IMF, 2006); the stock of migrant workers

in a host country (Freund, Caroline & Spatafora, Nikola, 2008) Those elements belong to political stability are the indicators that evaluate government stability, socio-economic conditions, investment climate, internal and external conflict, corruption, the involvement of the military in politics, religious tensions, law and order, ethnic tensions, democratic accountability, and bureaucracy quality The impact on remittance also depends on the natural disaster, when the incident occurs the migrant has to send more subsidy to the homeland than in normal circumstances (Ratha, Dilip, and Sanket Mohapatra., 2009)

According to the General Statistic Office of Vietnam, remittance equals to the total value of money received in cash or in-kind by household members survey from individuals (e.g., child or other relatives in Vietnam or abroad) who did not live in this household in the past 12 months before participating the VHLSS survey This study would investigate more on the impact of international remittance which means the sources from abroad into Vietnamese households Therefore, the term

“remittance” is used for international remittance only

2.1.2 Impacts of remittance on received household and home country

One prominent view empathizes with the importance of remittance when showing that the increase in remittance flows from migrants favors developing countries’ development via directly enhancing household’s living standard or by the utilization of these flows by financial institutions before in hand of received households Remittance helps households to protect from external shocks, or to invest

in the profitable business or to improve community infrastructure such as schools or clinics (Docquier & Rapoport, 2003) As findings of Adams and Page (2005) among

74 low and middle –income countries, a 10 % increase in per capita remittance would lead to a 3.5% decline in the poverty rate Mckenzie & Rapoport (2007) found that

as more people migrate abroad, the distribution of remittances can begin to reduce

income inequality Abu et al (2011) found the economic growth in Bangladesh was impacted by the growth of remittances meanwhile in India there is no causal relationship between remittance and economic growth; in Sri Lanka- an Africa country- a mutual directional causality is found Remittances to developing countries mostly excel as a tool to reduce the poverty depth and scale (Simone & Francesca, 2014)

Pessimistic authors argue that the impact of remittance to the home country would be indicated with income inequality, more dependent on host countries; the economy would be more unsustainable with the result of inflation and exchange rate appreciation due to remittance inflow There are arguments for the cause of inflation (because of demand in imported goods and services from host countries to home country) or the "Dutch Disease" (a large amount of home country currency in form

of remittance flow lead to exchange rate appreciation) with the significant impact of remittance on exchange rate appreciation found in Cape Verde (Bourdet, Y., & Falck, H., 2006) cited in Ratha 2013; Chami et al 2003) Remittance could contribute to deepening the development gaps through the creation of dependency on developed countries (for instance, the received remittances were mostly directed use to satisfy the household consumerism then increase of imports could be harmful to exports and domestic production) (Dean & Claudia, 2006)

In most households, migrant's remittance compensates for the absence of a family member, and to evaluate whether migration is a wise decision or not, the household often compares living standards and well- being of families before and after the left Most families would have plans to utilize the remittance depending on the negotiation of both migrant members and members left behind Remittances may therefore lead to an increase in the general prosperity of migrant-sending areas in the longer run But still, there is the case that a member staying at home does not commit

to the arrangement about remittance which soon later results in a fragile future for households Given the enormous wage gains possible from moving to a high-wage labor market, it would be expected that for the household as a whole (including the migrant), these opportunity costs should be less than the gain If the duration of migration is short and there are high transaction costs required to access the high-

wage labor market, then there is the possibility of little or no net gain Moreover, how these gains are distributed depends on the types of migration households selected into which then lead to the variation in the opportunity costs and transaction costs that they have to face off

2.1.3 Consumption patterns theories

Fundamental literature about consumption patterns which both mentioned Engel's theory are Working (1943) and Leser (1963) Working (1943) focuses more

to find the determinants of household expenditure patterns The household patterns

in this study are food, clothing, recreation & transportation, home maintenance, and other items

Using statistics data, he has found evidence for relations between the type of community, occupation, household size as well as age composition with proportions

of different household expenditure patterns The number of families shows effects on the budget distribution as explained from the economics of scale The large family can make more efficient use than the small family in terms of normal housing facilities or utilities and living space across all members of the home Leser, (1963) concentrates to construct the models to suit different levels of household income from other scholars and testing on larger household expense patterns (food, clothing, fuel

& light, housing, tobacco, household durables, miscellaneous goods & service, leisure)

By adopting the model of Engel functions about household consumption with summarizing the prior work on consumption patterns, (Leser, 1963) has revolted the model of Engel into the most functional model in regards to regress on consumption patterns At this study, C Leser denoted 𝑣𝑖 is expenditure on commodity group i and

M is household income

Model (1) was adopted by (Allen & Bowley, 1935) cited in Leser (1963) is generally valid for high M (unless applied to an inferior good) but not for low M; M: household income;

𝑣𝑖 = 𝛼𝑖+ 𝑀 + 𝜖𝑖 (1)

Then they adopted (2) with a different error specification with 𝜔𝑖 = 𝑣𝑖/𝑀 and called consumption pattern

They imply that demand elasticity for necessities increase with income

𝜔𝑖 = 𝛼𝑖+ 𝛽𝑖𝑙𝑜𝑔𝑀+ 𝛾𝑖/𝑀+ 𝜖𝑖 (4)

Nearly 20 years later, in 1980, Deaton and Muellbauer developed the concept

“Almost Ideal Demand System" (AIDS) as a demand system containing a series of models with many desirable properties The model also implies that an increase in income will lead to a decrease in the share of income allocated to a particular commodity, as well as a decrease in the income elasticity of that good when it is less than one The AIDS model can be estimated as follows:

𝑤𝑖 = 𝛼𝑖+ ∑ 𝛾1𝑗 𝑖𝑗𝑙𝑛 𝑝𝑗 + 𝛽𝑖𝑙𝑛 (𝑥

𝑃) + 𝜇𝑖 (5) i=1…11 where 𝑤𝑖 is the budget share of good i; 𝑝𝑗 is the price of good j; x is the total expenditure of the goods in question; m i is the random disturbances assumed with zero mean and constant variance, and P is a translog price index defined by:

The difference of this model in comparison with the Working- Leser model is the combination of price index which calculates the AIDS model be more difficult because the price index is not linear in terms of parameters (Deaton & Muellbauer (1980a, 1980b) cited in FAO, 2003)

Prais and Houthakker (1955) find out that lowering income results in a household budget dedicated more on “nourishment” (Daily nourishment from meals and beverages, spices, stimulants (e.g alcohol, coffee), tobacco, occasional dining out, etc.) Hoodfar (1997) found that saving is the priority of households and spending

on durable goods is a form of investment or asset accumulation of households, especially migrant households As a saving strategy, households spent their anticipated income accumulating assets (durable goods) before the money was even earned and when they spent more for recurrent or nourishment, the less chance a household could generate assets People saved through investing in household goods, which also enabled households to enjoy their utility in actual and prestige terms

2.2 Empirical Studies

The role of remittance is an additional source of earnings for households, but remittance especially abroad remittance might bring other invisible value like the culture or knowledge to influence household consumption behavior How remittances are used, or how consumption is oriented, depict the strategy of spending

of households and help them evaluate the effectiveness of migration

The most recent view concludes remittance is transitory income and should be spent more on human and physical capital investment – instead of consuming goods, but still some are showing both increases in long-term investment (education, health, housing) while still increase in some "luxury" spending Empirical studies showed that remittances had contributed to human capital investment such as children's education spending (Yang, 2008; Mansour et al 2011), health (Taylor & Mora, 2006) In the case of the Philippines, the household spends both on conspicuous consumption like a consumer good and leisure, but also more on education and housing (Tabuga 2007)

Unlike the above view, in some countries, the impact of remittance is not perceived as a special income source rather than wages or salaries because remittances do not support any expenditure behavior change caused by migrant transfers because they are fungible and families perceive them as another source of income In Senegal, extremely low-income households perceive income from remittances as merely wage income and they use it to satisfy nourishment needs (Randazzo et al 2014) There is no difference in behavior between those who receive and non-receive remittances (Zarate-Hoyos G A., 2004; Ang et al, 2009)

In Albania, households who receive remittances from abroad have a lower share of their expenditure on food and a higher share on consumer durables compared

to households who do not receive any type of migrant remittances (Castaldo & Reilly, 2007)

The studies of Adams and co-authors suggest that how remittances are perceived depends on the socio-economic conditions of the received household Middle-income households may better understand the value of remittances and use this source productively Besides, remittance receipt is associated with less spending

on food and greater spending on investments like education and housing (Adams et

al 2008a; Adams & Cuecuecha, 2010a)

Donghui et al (2019) investigate the impact of remittance on the share of different types of consumption in Kyrgyzstan with panel data from 2011-2013 The result found that increases in remittance receipt are associated with a lower share of food and a larger share of health care but with small changes and no changes found

on other expenses

When the studies from other countries have implemented analysis on consumption patterns, the studies in Vietnam mostly focus on the impact of remittance on household welfare or the marginal amount of consumption patterns Nguyen (2008) investigated household data in 2002-2004 and found that international remittances tend to help the recipients increase saving and production investment rather than on consumption or education while internal remittances are more likely to

be used for consumer spending

In a recent study of Nguyen & Mont (2012) on Vietnamese received remittance households, most of the international remittances are spent on real estate, financial services (paying debt and saving), and used a small share for durable consumption The effect of international remittances on poverty reduction is insignificant and those categories such as food, education, or health care spending are

remittances on household welfare using the result that remittances have a positive impact on per capita expenditure not per capita income, while migration reduces work motivation of members left behind in migrant-sending households

CHAPTER 3 RESEARCH METHODOLOGY

3.1 Model specification

To answer research questions, this study would follow the model of Donghui

et al (2019) designed a single year model households' consumption patterns, which adjust from Working–Leser specification to a simpler model:

𝐶𝑖𝑗𝑡 = 𝛼𝑖 + 𝛽𝑖𝑙𝑜𝑔(𝑀𝑗𝑡) + 𝜃𝑖𝑙𝑜𝑔(𝑟𝑒𝑚𝑖𝑡𝑗𝑡) + 𝛾𝑖𝑋𝑗𝑡 + 𝑢𝑖𝑗𝑡 (8)

Thus, we can use the panel nature of the data to avoid the endogeneity bias which is unobserved characteristics of received and non- received remittance households To incorporate the structure of the panel data, the model, incorporating time-varying components

Asides the share of consumption categories, the research also investigate on impact of remittance on seven consumption categories are expressed as follows:

𝑙𝑜𝑔(𝐸𝑖𝑗𝑡) = 𝛼𝑖+ 𝛽𝑖𝑙𝑜𝑔(𝑀𝑗𝑡) + 𝛿𝑖𝑙𝑜𝑔(𝑟𝑒𝑚𝑖𝑡𝑗𝑡) + 𝛾𝑖𝑋𝑗𝑡 + 𝑢𝑖𝑗𝑡 (9)

Where 𝐶𝑖𝑗𝑡 is the share of the expenditures on goods i in the total expenditures

for household j at time t The consumption share i for each household j is required

∑ 𝐶𝑖𝑗𝑡 = 1; 𝑙𝑜𝑔(𝑟𝑒𝑚𝑖𝑡𝑗𝑡) is the natural logarithm of total remittance amount

household j received at time t; 𝐸𝑖𝑗𝑡 is the level value of expenditure i of household j

at time t The variable 𝑙𝑜𝑔(𝑀𝑗𝑡) is the logarithm of other income (except for

remittance) per capita for household j, 𝑙𝑜𝑔(𝐸𝑖𝑗𝑡) is the logarithm of all consumption categories 𝑋𝑗 is a vector of household-related characteristics, and 𝑢𝑖𝑗𝑡 is the error term 𝛼𝑖 is the intercept, which captures the baseline expenditure share on goods i

while holding other covariates at zero value; while 𝛽𝑖, 𝜃𝑖 and 𝛾𝑖 are the marginal effects of log(𝑀𝑗𝑡) , 𝑙𝑜𝑔(𝑟𝑒𝑚𝑖𝑡𝑗𝑡) and 𝑋𝑗 respectively Model (5) is original from Donghui et al (2019) but in this study would adjust the 𝑋𝑗 indicators according to the case of Vietnamese households

In order to control the impact of omitted variables, panel data is used The fixed effect model would account for those time-invariant unobserved household

characteristics that may be correlated with households’ remittance-receiving status but are lacking in the models Assuming that only unobserved time-invariant characteristics (such as household members’ remittance-receiving norms) would affect households’ remittance-receiving statuses, the fixed-effects models get unbiased estimates

Table 3.1 Descriptions for expenditure categories:

Expenditure

category

Description

Food Expenditure for food and drinks in festival occasions or in daily

consuming in the last 12 months Health Total cost including fees for check-ups, treatment, medicines,

allowances for physicians, travel, purchase of facilities, relating

to visits for check -up/treatment in last 12 months Education Total of school fees, expense for studying instruments,

supplementary course in last 12 months Housing Payment for rent in last 12 months

Electricity Payment for lectricity in last 12 months

Durable Depreciation expenses of household appliances or house

maintenances (The buying cost minus estimated remain value of the asset)

Other Expense for clothes, garbage and water expense, personal care,

transportation, travel…

Source: Summarize of author from literature review

Table 3.2 Description of independent variables in regression model

Hhsize Number of family member (the number of people who were

considered parts of the household) Urb =1 if household resident place in urban area, 0 if not

Fmhead =1 if household head is female

age0_5 Number of members below 5 years old

age6_14 Number of members from 6-14 years old

age15_60 Number of members from 60-79 years old

Wage =1 if household head be hired and receives salary

Agri =1 if household head has business in agriculture sector

Nonagri =1 if household head has business in non-related agricultural

sector Head_age = Age of household head

Head_schooling = years of schooling completed

Married =1 if household head is married

RR =1 if household live in Red River Delta

MNM Midlands and Northern Mountains

3.2.1 A brief of Vietnam Household Living Standard Survey in 2012, 2014, 2016

The Vietnam Household Living standard Survey was conducted by the General Statistics Office every two years The survey collected information during four periods, each period in one quarter from the first quarter to the fourth quarter in that year through face-to-face interviews with household heads and key commune

officials in communes containing sample enumeration areas

In 2012, 46,054 households were sample size for a nationwide survey, 36,655 households were asked about income and other issues; 9,399 households were asked about income, expenditure, and other issues The VHLSS 2014 and 2016 have the same size of sample households with 46,995 households in 3,133 communes/wards which were representative at national, regional, urban, rural and provincial levels Of 46,995 surveyed households in 2016, 37,596 households were asked about income and other issues; 9,399 households were asked about income, expenditure, and other issues

Statistics offices in all 63 provinces take responsibility for choosing those survey resident areas Each resident area chooses 20 households (15 officials, 5 back-ups) with 15 households surveyed about income (12 officials, 3 back-ups), 5 households’ income and expenditure (3 officials, 2 back-ups) From the list of households in each region, 20 households would be chosen systematically random

3.2.2 Data processing

To collect data for this VHLSS survey, it takes two months for each round Those rounds combine sampling for the income-expenditure survey and income survey A direct interview is applied to those surveys Interviewers visit households, direct meet with the heads of households, and related household members to interview and fill information in questionnaires for each household After the data entry is completed at PSO (Provincial Statistic Officer), it is sent to the Center of Information and Statistics as well as to the Social and Environment Statistics Department Then the final result of the survey would be summarized by GSO to conduct the final data set GSO is also the government body that presents the data to the public

The data in this study is taken from VHLSS from 2012-2016 has been prepared

to conduct the same variables in each year then append them at the household level

In each year, the author generates variables by merging the master data file with other data files according to the description of models which then each year has only one merging file After that three files of 3 years would be appended and be declared as one-panel data with panel identification variable is household identification (only households who have the same identification remained from 2012 to 2016 data)

3.3 Structure of regression

Household characteristics included the occupation of household head, household size, age structure, total income per capita (except for remittance) The study uses the log-transformation for income remittances and expenditures because they are highly skewed and do not follow a normal distribution Aside from the income level, the social and material position that households envisaged for themselves influenced the allocation of their expenditure

The marriage status of the household head, the married is used as a dummy variable to represent this characteristic The education level of the household head is represented by the number of years completed school The age of household or the occupation also helps to define the trend of household consumption patterns

The community characteristics are living areas of households that are described by dummy variable urban/rural areas and the living region of households which are Red River Delta, Midlands and Northern Mountains, Northern, and Central Coastal, Central Highland, Southeast, Mekong Delta After running fixed effect regression, almost all regional indicators have been omitted from models such as Midlands and Northern Mountains, Northern and Central Coastal, Central Highland, Southeast, Mekong Delta, only Red River remains

The age structure of a household also contributes to the consumption structure The composition of age structure for this study is number of child dependents (0-5 years old), number of schooling dependents (6-14 years old), numbers of working force members (15-60 years old) The gender structure is calculated by the number

of females in a household Time-invariant characteristics were those that we held constant over the three waves They were characteristics of the community, household head characteristics (gender, and years of completed school of household head,) living region of a household

As can be seen from table 4.3, the international remittance received by one household member is equal to over half of the sum from other income sources Therefore, remittance keeps a vital role in maintaining and enhancing living standards

of not only received but also non received households The amount of remittance received per capita increases along with the increase of other income but the increase

of remittance in the period 2014-2016 is higher than other income per capita There are about 30% of Vietnamese households are located in urban areas and about 24%

of household heads are female Most household heads in Vietnam are getting married but the share of this status is decreasing significantly from 82% in 2012 to 79.1% The number of working people from 15 to 60 years old declines from 2.56 to 2.4 persons along with the decrease of household size from 3.87 to 3.71 on average for each household Most of the household heads possess agriculture business and a large part of household heads are hired by other people Because most households in Vietnam live in rural areas, the household heads have to maintain the agriculture business as their main source of income besides seasonal work