Remittances domestic investment in the east and southeast asia

LIST OF ABBREVIATIONS FD Financial Development FDI Foreign Direct Investment FE Fixed Effect GOP Gross Domestic Product GLS Generalized Least Square GMM Generalized Method of Moment IMF

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

' ' ' ,

REMITTANCES- DOMESTIC INVESTMENT

IN THE EAST AND SOUTHEAST ASIA

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By NGUYEN THI HAl YEN

Academic Supervisor:

Dr NGUYEN HUU DUNG

ABSTRACT

Over the past decades, remittances have grown to become one of the largest sources of external financial flows to developing countries, left ODA far behind and even exceeded FDI in some countries While it is undeniable that remittances have poverty-alleviating and consumption-smoothing effects on households in receiving economies, empirical questions are whether they directly impact investment and if not through which channels they have positive effects on investment This thesis goes in search of the answers to these questions by employing different methodologies including Ordinary Least Square, Fixed Effect and Generalized Method of Moments However, research scope is limited within some economies in East and South-east regions The results show that remittances have no direct impact on investment but do when being combined other channels such as financial development and institution Remittances are complementary to financial development in stimulating investment Meanwhile, relationship between remittances and institution quality is not clear Some indicators indicate remittances are complementary to institution quality in boosting investment; whereas the others are substitutionary

Key words: Remittances, Investment, East & Southeast Asia, Financial Development, Institution Quality

CONTENT

CERTIFICATION i

ACKNOWLEDGMENTS ii

ABSTRACT iii

CONTENT iv

LIST OF TABLES i

LIST OF ABBREVIATIONS ii

CHAPTER 1: INTRODUCTION 1

1.1 Problem Statement 1

1.2 Research objectives 3

1.3 Research Questions 4

1.4 Research Methodology 4

1.5 Structure of the thesis 5

CHAPTER II: LITERATURE REVIEW 6

2.1 The concepts of remittance 6

2.2 Motivation and economic effects of remittances 7

2.3 Relationship between remittances and investment 10

2.4 Empirical studies 14

CHAPTER III: RESEARCH METHODOLOGY 18

3.1 Data and variable description 18

3.1.1 Data 18

3.1.2 Description of variables ofinterest 19

3.2 Model specification & analytical steps 23

3.3 Methods of parameter estimations 24

~ 3.3.1 Ordinary Least Squares (OLS) estimator 24

3.3.2 Fixed Effect (FE) estimator 25

3.3.3 Generalized Method of Moments (GMM) 27

3.3.4 Application of GMM estimator in the thesis corresponding to each specification 28

CHAPTER IV: EMPRICAL RESULTS 31

4.1 Descriptive statistics 31

4.1.1 Remittances in East and Southeast Asia 31

4.1.2 Investment and economic growth in East and Southeast Asia 32

4.1.3 Financial development, openness and lending interest rate in East and Southeast Asia 32

4.1.4 Institution quality in East and Southeast Asia 33

4.2 Bivariate correlation of the variables of interest 34

4.2.1 The relationship between remittances and variables of interest 34

• 4.2.2 The relationship between investment and the other variables 34 4.2.3 The relationship between financial development and other variables 36

4.3 Empirical Results 37

4.3 1 Examining the impact of remittances on domestic investment using OLS estimation technique 3 7 4.3.2 Examining the impact of remittances on domestic investment using fixed effect (FE) estimation technique 50

4.3.3 Examining the impact of remittances on domestic investment using Generalized Method of Moment (GMM) estimation technique 61

4.4 Brief discussion about the best model 73

CHAPTER V: CONCLUSION, RECOMMENDATION & LIMITATION 75

5.1 Conclusion 75

5.2 Policy recommendation 78

5.3 Limitation and suggestion for further study 80

REFERENCE 82

APPENDIX 88

LIST OF TABLES

Table 1: Variable description, Expected sign & Data Source 22

Table 2: OLS- REM 37

Table 3: OLS- REM- FD 39

Table 4: OLS - REM - FD (M2) - INST 43

Table 5: OLS- REM- FD (BANCRE)- INST 45

Table 6: OLS- REM- FD (PRlCRE)- INST 47

Table 7: FE -REM 50

Table 8: FE - REM - FD 52

Table 9: FE - REM - FD (M2) - INST 54

Table 10: FE - REM - FD (BANCRE) - INST 56

Table 11: FE -REM - FD (PRJCRE) - INST 58

Table 12:.GMM- REM 61

Table 13: GMM- REM- FD 63

Table 14: GMM- REM- FD (M2)- INST 66

Table 15: GMM- REM- FD (BANCRE)- INST 69

Table 16: GMM- REM- FD (PRICRE)- INST 70

Table A: Descriptive statistics on interested variables without institution quality 88

Table B: Descriptive statistics on institution quality indicators 88

Table C: Bivariate correlation of the variables of interest 89

LIST OF ABBREVIATIONS

FD Financial Development

FDI Foreign Direct Investment

FE Fixed Effect

GOP Gross Domestic Product

GLS Generalized Least Square

GMM Generalized Method of Moment IMF International Monetary Fund INST Institution

ODA Official Development Assistance OLS Ordinary Least Square

REM Remittances

WB World Bank

- - - - - - - - - - -

-CHAPTER I: INTRODUCTION

This chapter will explain the reason for carrying out the thesis, its objectives and research questions In addition, a brief of methodology is also mentioned in this part Finally, the structure of the thesis will be presented

1.1 Problem Statement:

Rapid economic growth and sustained development are goals pursued by almost governments and policy-makers Investment is a crucial motive for growth and development process Investment, of course, requires capital Most of developing countries, unfortunately, usually encounter shortage of capital Capital source from domestic savings is rather limited due to living standard of residents generally is still low Therefore, attracting foreign capital flows is one of the leading national policies for development Regarding foreign capital flows, portfolio investment is high volatile and may cause negative effects for the economy without appropriate and active management (Silva et al., 2009) Foreign Direct Investment (FDI) with relatively stable characteristics compared to portfolio investment usually attracts lots of attention and interest thank to its big benefits such as capital supply, job creation, technology transfer, and managerial know-how FDI, however, typically requires a physical investment as well as transparency, stability and soundness in political environment and macroeconomic policies in the country

A foreign capital flow which has recently received much interest from economists is remittances Remittances, the amount of money transferred by migrants to their families in home countries, have steadily increased since the mid-1980s Statistically, recorded remittances reached an estimated volume of US$206 billion in 2006, compared to US$19.6 billion in 1985 (Ambrosius et al., 2008) The latest available data shows that remittance flows to developing countries reach US$328 billion, accounting for over 70% total global remittances in 2008 (see Key Indicators for Asia and the Pacific 2009) The volume of them is nearly the same

with foreign direct investment and twice higher than official development aid that help them rank at the second position in term of external finance source for developing countries Moreover, out of non-trade foreign currency inflows, remittances are in general the most stable, least volatile (Chami et al., 2008; Global Development Finance 2003) Thanks to the characteristic of stability, these inflows may counter negative effects caused by the falling in capital flows such as FDI, debt, and equity flows in recipient countries during an economic downturn (Silva et al

2009) and hence consolidate macroeconomic stability (Bugamelli and Paterno, 2008)

Furthermore, m developing countries, the financial system is commonly underdeveloped and the economy has a low degree of monetization since they usually suffer a low ratio of credit to GDP and a low ratio of broad money supply to GDP (Ambrosius et al., 2008) The authors insist: "A large proportion of the population and typic~lly the small and micro enterprises of the informal sector find difficult to access bank credit and thus operate outside the financial sector, with low capital intensity and low productivity" A recent study by Aggarwal et al (2005) shows that remittances promote financial development; whereas the finding of Bettin and Zazzaro (2009) holds that these inflows play substitute for underdeveloped financial sector in developing countries in stimulating investment and growth

Remittances' impacts on poverty, education, inequality are extensively exploited; whereas research on remittances-investment nexus is still limited, leaving

a big gap that needs further exploration This gap derives from a common perception that remittances mainly serve for consumption purposes, not for investment That is why this thesis would like to exploit this aspect The thesis seeks to find out the impact of remittances on investment in East and Southeast Asian developing countries There are some reasons for choosing these regions in the thesis Firstly, they are one of the largest recipients of international remittances among all developing regions in which some are in the top remittance-receiving

countries in the world in absolute terms; namely, China, Philippines, Thailand and VietNam (see Key Indicators for Asia and the Pacific 2009) Next, research on the impact of remittances focus on other regions like Latin America (Ramirez et al.,

2008), Africa (Anyanwu et al., 2010), Sub-Saharan Africa (Gupta et al., 2007; Singh et al., 2009), East and West Africa (Black et al 2004), Europe and Central Asia (Black et al 2007), South Asia (Khatri, 2007) whereas these regions seem to

be ignored; if any just at country level (Yang, 2004) Finally, right after the financial crisis in 1997-1998 East & Southeast Asian saw a dramatic withdrawal of capital from the regions and limited decreased banking lending by banks in other regions; then flows have picked up gradually but not so high as in the period before crisis Strengthening financial system with the aim of ensuring domestic funds are allocated efficiently while limiting economies' vulnerability to future fluctuations in capital flows, stimulating investment and hence economic growth is an important objective that governments of these countries have been making effort to obtain (see East Asian capital flows in Economic Roundup Spring 2003) However, once domestic financial system has not been developed completely yet, whether some factors, remittances for instance, could be a complementary or substitutionary element to financial sector in fostering investment in order to achieve the final objective: economic development The answer will be founded in this thesis

1 To examine the impact of remittances on domestic investment

- -

-2 To consider the interaction between remittances and vanous channels including financial development and institution quality in boosting investment

3 To submit recommendations to governments on the policies that might best promote the effective use of remittances in investment activities

Based upon the results to these questions, the following research hypotheses will be tested using the dynamic panel data model with different approaches: static and dynamic

1 Do international remittances directly affect investment?

2 What are roles of financial development and institution quality m remittance-investment nexus, complementary or substitute?

1.4 Research Methodology:

This paper uses annual panel data of 9 economies in East and Southeast Asian regions including Cambodia, China, Hong Kong, Indonesia, Korea, Malaysia, Philippines, Thailand, Vietnam in which remittance flows are rather significant and other data resources are available for the period 1995-2008

To estimate the impact of remittances as well as other channels on domestic investment, both static & dynamic approaches are employed Static approach includes Ordinary Least Square (OLS) and Fixed Effect (FE); whereas Generalized Method of Moment (GMM) is applied in the spirit of Arellano and Bover (1991)

- - - - - - - ,

1.5 Structure of the thesis:

This thesis is structured in five chapters

Chapter I: explains the reason for carrying out the thesis, its objectives and

research questions and briefly state employed methodology to examine hypothesis

questions

Chapter II: demonstrates the literature review It starts with the concept of

remittances Following is summarized presentation on motivation of remittances

and their economic impacts on receiving economies Theoretical and empirical

literature on relationship between remittances and investment is the final part of the

chapter

Chapter III: introduces research scope, data source, variables of interest and

analytical steps associated with specific models Different estimation techniques

together with null hypothesis testes are also presented in detail in this chapter

Chapter IV: indicates empirical results corresponding to each estimation

technique as well as results of hypothesis testes Yet, to clearly catch up analyzed

contents, let star with overview of East and Southeast Asia via descriptive statistics

as well as bivariate correlation between variables

Chapter V: will give conclusions of empirical results, limitation of the thesis and

suggest recommendations

CHAPTER II: LITERATURE REVIEW

The chapter firstly presents the concept of remittances Next, a brief description on motivations of remittances together with economic impacts of remittances on remittance-receiving economies is presented Finally, theoretical and empirical literature on remittance-investment nexus is mentioned

2.1 The concepts of remittance:

Appendix 5 on remittances to the "Balance of Payments and International Investment Position Manual" (BPM6), International Monetary

Fund introduced the economic concept of remittances as: "Remittances represent household income from foreign economies arising mainly from the temporary or permanent movement of people to those economies Remittances include cash and noncash items that flow through formal channels, such as via electronic wire, or through informal channels, such as money or goods carried across borders They largely consist of funds and noncash items sent or given by individuals who have migrated to a new economy and become residents there, and the net compensation of border, seasonal, or other short-term workers who are employed in an economy in which they are not resident "

There are many components which are specified to constitute remittances A common practice is to sum the three categories in the IMF's Balance of Payment Statistics Yearbook (BOPSY) when compiling statistics on remittances They are

"Workers' Remittances", "Compensation of Employees" and "Migrant Transfers" According to a statistic report about development indicators by World Bank, the organization gives a summarized definition on these items based on the IMF's Balance of Payments Manual (the fifth edition) as follows:

• Workers' remittances are "current private transfers from migrant workers

resident in the host country for more than a year, irrespective of their immigration status, to recipients in their country of origin."

• Compensation of employees is "the income of migrants who have lived in

the host country for less than a year."

• Migrant transfers are "the net worth of migrants who are expected to

remain in the host country for more than one year that is transferred from one country to another at the time of migration."

2.2 Motivation and economic effects of remittances:

Last decades have witnessed a substantial increase in volume of remittances, and it draws lots of attention from researchers and policy-makers and then leads to a vast of literature on these flows Generally, theoretical and empirical econom1c literature on migrants' remittances IS divided into two main sections: microeconomic section on remittance determination and macroeconomic on remittances' effects on recipient economies

With regard to the theory of factors driving remittances, the findings at micro level suggest that migrants' behavior of transferring money home back is explained

by various motives such as altruism, exchange, inheritance, strategic motive,

insurance, loan repayment (see Chami et a/., 2008) Despite various motives

suggested by economists, the literature generally identifies two basic motivations for remitting: altruism and self-interested exchange (or the exchange motivation) Altruistic motivation is referred to the fact that one migrant altruistically motivated

to insure his/her family, so he/she remits funds Under this motivation, remittances are compensatory transfers as the care of the migrant for those left behind and are usually associated with economic shock in the home country Meanwhile, exchange motivation is referred to remitting with a main purpose of increasing welfare, gains for remitters while assuring benefits for recipients Interestingly, researches find that both altruistic and exchange motivations for remittances are not mutually exclusive at all but may be highly operative at the same time

Conclusions drawn from literature on remittance determination do not point out clear implication for the economic impacts of remittances in recipient

economies, but give useful insights into these flows' economic impacts Impacts caused by remittances for recipient countries are positive or negative are much dependent on how these flows give response to economic conditions in both the home and host countries If they are mainly transferred with the aim to take advantage of favorable economic conditions in the receiving country, then they are similar to other capital flows with the final objective of profits and they should be considered as such inflows Otherwise, they are primarily compensatory transfers; their economic impacts will also be different to those of capital flows (Chami eta/.,

2008)

Regarding effects of international remittances on the recipient economies, the theoretical and empirical literature state two contrasting views: remittances may have both favorable and unfavorable consequences The optimistic view considers remittances as mechanisms for economic development (Taylor, 1992; Faini, 2002); whereas the pessimistic suggests that remittances have a negative impact that deteriorates the economy (Cattaneo, 2005, Chami eta/ 2005)

In the former view, there is a strong consensus that remittances motivated by altruism play a role as addition income in poor regions, help smooth consumption constraints, leading to poverty reduction and inequality that contribute to economic development [Black el a/., 2007) on Moldova, Tajikistan and Kosovo; Anyanwu et a/ (2010) on Africa and Portes (2009), Acosta (2007) on Latin America] Moreover, remittances promote human capital development by increasing the capacity of households to spend on education, health, and nutrition [Acosta et al., 2007) on Latin America, Yang (2004) on Philippines; Adenutsi (2009) on Sub-Sahara Africa] Remittances can foster economic growth by spurring entrepreneurial activity, improving labor productivity [Massey & Parrado (1998) on West-Central Mexico, Woodruff & Zenteno (2001) on Mexico] They may also contribute to macro stability of recipient economies by providing them with foreign exchange and improving their creditworthiness [Mongardini and Rayner (2009) on Sub-Saharan Africa, Coronado (2009) on Mexico, EI Salvador and Turkey] Moreover, by easing

This leads to a change in demand for and supply of these goods and the ratio between them To a certain extent, the share of non-traded good consumption gets larger that make the elasticity of substitution between traded and non-traded goods smaller To ensure a balance in the economy, it requires the real exchange rate to be appreciated So, under this context, remittances cause appreciation of exchange rate, thereby dampening trade competitiveness, creating a phenomenon called Dutch disease (Chami eta!., 2008; Acosta eta!., 2009) Also, remittances also may lead to

high inflation which causes bad effects for many economic activities (Coulibaly, 2009) Besides, remittances are considered as a factor causing disincentives to work among recipients, dampening economic growth The explanation lies in the fact that

if remittances rise up, receiving households are likely to increase expenditure for consumption and reduce labor supply It means that they will consume more leisure

as well as more goods Consequently, the real output of the economy will fall down due to a decrease in the aggregate supply of labor (Chami eta!., 2008) Moreover,

Karunaratne (2008) shows that remittances just give benefits for some group in Lanka, not for the poor, so increase gap of inequality

Sri-There are also some efforts seeking impact of remittances on economic growth

The results, however, are mixed and inconclusive The empirical evidences by Solimano (2003) for Andean countries and by Giuliano et a! (2006) for 100

developing countries show the positive impact of remittances on growth The opposite result is found in a study by Chami et a! (2005) using a panel of 113

countries Meanwhile, World Economic Outlook (2005) indicates that there is no

sufficient evidence about correlation between these variables, at least at the country level

2.3 Relationship between remittances and investment:

As commonly conceived as resources served consumption purposes rather than for investment, the impact of remittances on investment has not yet been explored deservedly Yet, economists make efforts in analyzing this relationship and suggest that remittances may affect the rate of investment in recipient economies, depending how remittance receipts are used as well as motives driving remittance flows

(Chami et a!., 2008) This relationship is considered in different cases in which

remittances are in turn regarded as pure income transfers, transfers with a merit good component and as a capital flow In the first case of pure income transfers, a framework to analyze remittance-receiving households' consuming and saving behavior in which a one-good production structure is assumed and the representative household is allowed to consume goods and leizure and to accumulate capital is employed The current output is supposed to be costlessly converted into physical capital and hence capital stock in the economy is always at its desired level

In a frictionless world, these households tend to consume rather than to save

an additional dollar of remittances if these flows are permanent The additional saving does not lead to a higher level of investment since it does not affect domestic cost of capital Accordingly, remittances have no effect on the level of domestic investment However, with an assumption that perfect capital controls impose a condition of financial autarky on the domestic economy, remittances really affect the level of domestic investment through saving channel In this situation, to the extent that remittance flows will be saved and these savings must be completely devoted to finance domestic investment since that is the only way household sector can accumulate wealth

In the friction world with types of productive assets, assummg that all households are rational, they have many choices to accumulate and increase their wealth To maximize their wealth, these households will allocate resources among such assets so that marginal rates of return they can gain are equal The main point

of the issue lies in the fact that remittance inflows may cause impacts on domestic investment only when external capital mobility in receiving countries is less perfect

In other words, the state of financial system in recipient countries decides remittances' effects on investment rate It is interpreted that lots of households in receiving countries, financial sector of which are weakly developed financial sectors, find difficult in approaching formal credit sources since they have to pay high costs to get such sources Therefore, they are limited in investing potentially productive projects Remittance inflows will ease these households' capital pressure, helping them undertake these projects and thus boost the level of investment and enhance economic growth (Chami et al., 2008)

In the case that remittances are transferred as a merit good component, the remitter sends money to his/her home country with specific purposes Transfers like that affect the recipient's decision in allocating remittances between consumption and investment Suppose that migrants here are parents who send remittances to the home country to pay tuition fees for their children The merit good here takes a form of investment in human capital

As presented above, one of determinations of remittances IS exchange motivation which is referred to the recipient uses transfers on the migrant's behalf

in order to increase the remitter's welfare Lucas and Stark (1988) set up a model based on the economic theory of the family indicating relationship interaction between the migrant and his family back home via remittances transferred The migrant through his family member· as agents do investment in his home country and asks the family member to look after the investment project In return, the migrant offers the family member some monetary transfer In this circumstance, remittances play a role as capital flows which are sent primarily to take advantage

of high return or other favorable conditions in the home country Therefore, remittances may contribute to stimulation of domestic investment

Furthermore, if remittances are regarded as capital flows, their impacts on investment are once again clearly observed through an analysis of relationship between capital flows, financial development and economic growth by Bailliu (2000) She applied a framework based on the endogenous growth model called AK model to point out the role of financial intermediation in increasing investment rate

by efficiently allocating savings mobilized from residents and foreign capital flows into highly productive projects This framework can be explained as follows: given the aggregate production function of the economy

Yt = AKt (1)

Where Y is aggregate output, K is aggregate capital stock and A represents the overall productivity of capital Gross investment equation has a form as follows: It= Kt+ 1- (1-x)Kt (2)

Where It and Kt refer to gross investment and capital stock in the economy at t period, x is capital depreciate rate

As financial intermediaries play a role in collecting free-resources in residents and allocating them in high-yield projects, the intermediaries will retain some as reward for supplied service Assuming a fraction of each dollar saved is available for investment is z, the intermediaries will retain 1-z This transaction cost is regarded as the spread between lending and borrowing rates charged by banks In the closed economy, the equilibrium requires:

zSt =It (3)

Where St is gross savings mobilized from residents

Combining three equations from (1) to (3) and dropping the time indices, the growth rate of output, g, can be written as follows

g = A(IIY)- x = Azs- x ( 4)

Where s denotes the gross saving rate The equation ( 4) reveals two main channels through which financial development can affect economic growth The first channel

involves the efficiency with which saving are located to investment As financial system is developed, financial institution will be more efficient and the spread between lending and borrowing rates falls, leading an increase in z As a result, the proportion of savings channeled to investment will be higher The second channel is

an increase in the overall productivity of capital A The interpretation lies in the fact that a developed financial system will be more efficient in evaluating alternative investment projects, selecting and allocating savings to the most productive ones Now, the context of open-economy with the presence of international capital flows is considered For simplicity, the capital flows is supposed to go through the financial intermediaries The capital market equilibrium becomes:

From the framework above, capital flows can promote economic growth through two channels The first channel is through an increased investment rate thanks to higher savings (s*>s) The second one is through an improved banking sector which is more effective due to less costs of transaction (z*>z) and better at allocation of capital to highly yielding investment projects (A *>A) This framework stresses the importance of domestic financial development in enhancing capital flows' positive effects on the economy It implies that between two countries receiving the same amount of capital, the country which has more developed financial sector is expected to gain higher economic growth

by Dustmann and Kirchamp (200 1) indicates that the savings from returning migrants may play an important capital source to start up micro-enterprises Similarly, employing state-level data in order to investigate the relationship between migration and small investment in Mexico, Woodruff and Zenteno (200 1) also find out a positive correlation between the amount of migration and business investment in the state and almost 20% of the capital invested in micro-enterprises throughout urban Mexico comes from transferred remittances

Furthermore, the finding by Brown (1994) indicates households in Tonga and Samoa those receive remittances tend to save more or do more investment Alderman (1996) and Adams (1998) find that remittance-receiving households in Pakistan tend to use their addition income from transferred remittances to acquire land and buildings Mesnard (2000) carries out an empirical research in Tunisian Republic and discovers that a large of workers who face liquidity constrains in their home country will overcome those pressures by using accumulated savings obtained

in ·a foreign country for their investing activities

The uses of remittances in human development also increase physical investment level because they create demand for social projects including school, hospital buildings, clinics and other infrastructure (Yang, 2004)

Remittances may affect the efficiency of investment m the recetvmg economies through their impact on the development of the domestic financial sector

As extensively proved in theoretic and empirical evidence, the financial development positively impact economic growth and development (King & Levine,

1993, Levine & Zervos, 1998; Beck, and Beck, Levine & Loayza, 2000) Lots of researches which are conducted across countries suggest the higher financial development the greater economic growth These literature holds that financial institutions with huge power of knowledge and information collection about markets and abundant resources can easily mobilize "free" savings from residents, efficiently re-allocate these financial resources to other profitable projects which are high return but require a lot of capital, and monitor and control operation of enterprises to which they make loans to assure mechanisms run well Hence, investment activities are encouraged and eventually enhance growth A notable

study conducted by Aggarwal et al (2005) closely examines the link between

remittances and financial sector development at the country level Using a panel data of 99 developing countries over the period 1970-2002 the authors provide strong support for the notion that remittances promote financial development in term of the ration of deposits to GDP Complementary to this result is another study

by Gupta et al (2007) They also find out the positive impact of remittances on

financial development in Sub-Saharan Africa More interestingly, some papers study about the interaction between remittances and financial development to investment and economic growth as well Typically, Mundaca (2005) and Giuliano

et al (2006) differ in their approach to measuring the link between remittances,

financial development, and growth Using a panel data set in Central America, Mexico, and the Dominican Republic over 1970-2003 Mundaca finds that remittances have stronger effect on growth where financial system is rather developed since financial development potentially leads to better use of remittances, hence boosting growth Contrast to this result, using a panel of more than 1 00

countries for the period 1975-2003, Giuliano et al (2006) show that remittances

are substitutes for financial development in fostering investment and economic growth Remittances are significantly positive with growth in economies where financial sectors are underdeveloped

Different to methodology of other exiting studies, Diaz (20 1 0) uses path analysis to research the relations between remittances and investment and economic growth The method allows Diaz to simultaneously analyze multiple causal relationships among variables, both directly and indirectly The result shows that remittances have positive impact on growth but not significant He finds, however, that remittances have a positive and significant impact on investment and remittances indirectly positively affect economic growth through the channel of investment

Most recently, Gheeraert et al (2010) consider the impact of remittances on

banking deposit and formal investment through examining the interaction between remittances and different transaction costs, namely Cost of Bank Depositing (CDEP) and Cost of External Finance (CEXF) that were associated with financial development Using a sample of 100 developing countries over the period 1975-

2004, they model the loanable fund market which unfolds two stages: depositing and investing, to highlight the effects of remittances on investment The result indicates that remittances have significantly positive role of stimulating both investment and bank deposit, for all levels of the two transaction costs considered However, remittances have a positive effect on informal investment, which increases with CDEP; whereas the marginal impact on formal investment declines with the CDEP The research is complementary to previous studies which claim that financial sector development is crucial as it allows remittances to better fuel domestic investment

Another interesting research conducted by Bjuggren et al (2010) The authors

use dynamic panel data approach with remittances data covering 79 developing countries during 1995-2005 to address the potential endogeneity of remittance

inflows Relatively closed the model by Bettin et al (2009) that studies the impact

of remittances on growth, the authors seek to find out remittance-investment nexus through simultaneously examining the roles of financial development and institution quality in stimulating the impact of remittances on investment However,

rather different to several previous studies, the study keeps annual subsequence data instead of averaging observations over 5-year periods to get around the threat of autocorrelation The result reveals that remittances, high quality institutional framework and well-developed credit market have positive roles in stimulating investment However, the marginal effect of remittances will go down in countries where institutional framework is improved and credit market is well-developed

In conclusion, the chapter presents three main points Firstly, remittances are driven by two main motives: altruism and self-exchange Next, these inflows have both positive and negative impacts on receiving economies dependent on how they are used and their motives Finally, remittances may stimulate investment directly and/or indirectly through channels, namely financial development

CHAPTER III: RESEARCH METHODOLOGY

This chapter is separated into three main sections The first is the introduction

of research scope, data source as well as description of variables of interest, which are employed in empirical models The next part states analytical steps associated with specific models to consider the direct impact of remittances as well as the interaction between these money inflows with channels, including financial development and institution quality, on domestic investment to examine hypothesis questions._ The final is description of different estimation techniques including: Ordinary Least Square (OLS), Fixed Effect (FE) and Generalized Method of Moment (GMM) together with how to test significance of the employed methods

3.1 Data and variable description:

3.1.1 Data:

This thesis focuses on the impact of remittances on investment in only some economies in East and Southeast areas since 1995 up to 2008 Due to the lacks of data and continuous observations for some countries in the study period of 1995-

2008 made not all countries in the two regions included in the estimation This thesis uses panel data of 9 developed and developing economies in which remittance flows are rather significant and other data resources are available for the period 1995-2008 Namely, the Southeast Asia includes Vietnam, Thailand, Philippine, Malaysia, Indonesia The East Asia includes China, Korea and Hong Kong Most of data is obtained from Key-Indicators 2009 of Asian Development Bank and World Development Indicator of World Bank As for indicators related to Institution Quality, data are obtained from Governance Matters 2009 Worldwide Governance Indicators 1996-2008 These indicators are available only after 1996 In summary, the empirical study is carried out based on a data panel of 9 cross-section representing 9 economies and 14 periods from 1995-2008 with annual observations

3.1.2 Description of variables of interest:

3.1.2.A Dependent variable:

Dependent variable Investment (INV) employed here is the ratio of gross capital formation to gross domestic product from World Bank as the principal measure of investment's share of aggregate output As defined by WB, "gross capital formation (formerly gross domestic investment) consists of outlays on additions to the fixed assets of the economy plus net changes in the level of inventories Fixed assets include land improvements, plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings Inventories are stocks of goods held by firms to meet temporary or unexpected fluctuations in production or sales, and 'work in progress'."

3.1.2.B Independent variables:

countries of origin Remittances are constructed as the sum of three items Workers' Remittances, Compensation of Employees and Migrant Transfers over GDP

to measure In this thesis, three traditional quantity-based indicators are utilized to proxy financial development and these indicators are measured based on ratios of each proxy to gross domestic product (GDP) To measure development levels of financial system, we can utilize proxies in absolute dollar amount or as a ratio of GDP If the first application, identifying the absolute dollar amount of proxies, is informative; the second "facilitates benchmarking of the state of financial development and allows comparison across countries at different stages of development" (at page 16, Chapter 2 of Indicators of Financial Structure, Development, and Soundness in a handbook named Financial Sector Assessment issued by cooperation of the World Bank and IMF)

Broad money supply (M2): is measured by the ratio of liquid liabilities of the

financial system (currency plus demand and interest-bearing liabilities of banks and non-bank financial intermediaries) to GDP This is an important indicator reflecting the degree of the currency of a country's economy Overall, the high ratio indicates the high degree of monetization of the economy & implies the developed finanCial sector Otherwise, the lower of monetization of the economy points out the backward financial sector

Domestic credit provided by the banking sector (BANCRE): It includes

"all credit to various sectors on a gross basis, with the exception of credit to the central government, which is net The banking sector includes monetary authorities and deposit money banks, as well as other banking institutions where data are available (including institutions that do not accept transferable deposits but do incur such liabilities as time and savings deposits) Examples of other banking institutions are savings and mortgage loan institutions and building and loan associations" under the concept given by World Bank In the current modem world,

there is cocurrent existence of two types of financial structures including based financial system & market-based financial system In the former, banks play a leading role in mobilizing savings, allocating capital, overseeing the investment decisions of corporate managers, and providing risk management vehicles; whereas

bank-in the latter securities markets share center stage with banks bank-in gettbank-ing society's savings to firms, exerting corporate control, and easing risk management In any systems, the huge importance of banks for economies is undeniable Accordingly, it

is suitable to employ this indicator, which measures how much intermediation is performed by the banking system, including credit to the public and private sectors,

to proxy financial development

Domestic credit to private sector (PRICRE): This refers to "financial

resources provided to the private sector, such as through loans, purchases of equity securities, and trade credits and other accounts receivable, that establish a claim for repayment" under the concept given by World Bank On the website of

non-World Bank (Working for a non-World Free of Poverty), this organization states its point of view about private sector as follows: "private sector development & investment are critical for poverty reduction"; especially in competitive markets, private investment has tremendous potential to economic growth & development

"Private markets are the engine of productivity growth, creating productive jobs &

higher incomes And with government playing a complementary role of regulation, funding, and service provision, private initiative and investment can help provide the basic services and conditions that empower poor people - by improving health, education, and infrastructure" Unfortunately, in less & developing economies where financial systems are weakly developed, governments almost focus on key industries and often intervene domestic financial markets, causing distortion, a phenomenon called financial repression; private sectors find difficult accessing credit sources Therefore, this indicator combines quality and depth and is a better measure compared to other common indicators proxied for financial development because it only includes credit issued by the private sector to the private sector (Levine eta!., 2000)

Institution quality (INST): m this thesis, SIX indicators developed by Kaufman et a! (2006) are employed to proxy institution quality They includes

Control Of Corruption (CoCor) measures people's perceptions of corruption;

Regulatory Quality (ReQua) measures the incidence of market-unfriendly

regulation like price controls or inadequate bank supervision; Rule Of Law

(Rulaw) quantifies the confidence of the agents in the rules of society and in the

protection of property rights; Voice & Accountability (V oiAc) measures the

citizens' participation in politics; Political Stability (PoSta)represents the risk of sudden and violent changes in government; Government Effectiveness (GovEf)

captures the ability of the government to implement good policies All indicators vary between -2.5 and 2.5, with higher scores corresponding to better institutions

GDP Growth (GROW): is the growth rate of gross domestic products which presents expension of productive capacity of a country This variable is employed to

control for the accelerator (Bjuggren, 201 0) and expected to be positively associated with investment

Lending rate (LEND): refers to the rate charged by banks on loans to prime customers Economic theory holds that resources for economic development are scarce and costly This is also not an exception for financial resources which are required to undertake the necessary investment As a lending rate is likely to constrain the optimum investment it is employed to control for this effect and expected to be negative to investment

Openness (OPEN): is measured by the share of trade (exports plus imports) in GDP and is used to control for the access to the international market This variable

is an important indicator to measure a country's capacity to approach the international market This variable is expected to have positive relationship with investment

Table 1: Variable description, Expected sign & Data Source

Sign Dependent variable

INV Gross Fixed Capital Formation I World Development

PRICRE Domestic credit provided to the (+) World Development

private sector/GDP Indicators The SlX governance indicators

Governance Indicators INST developed by Kaufman et al (+)

for 1996-2008 (2006)

The growth rate of gross World Development

Indicators domestic product

World Development OPEN (Export+ Import)/GDP (+)

Indicators The rate charged by banks on World Development

Indicators loans to prime customers

3.2 Model specification & analytical steps:

The impact of remittances on investment 1s m tum examined through varied specifications following each step below:

Step 1: Firstly, the question of whether remittances significantly stimulate domestic investment will be examined through the specification below without consideration of effect of financial development and institutional quality

INVi,t = ao + ~~REMi,t + ~2,Yi,t+ ui,t(la)

Where i refers to the country and t refers to the time period from 1995 to 2008 INV

& REM refer to the ratio of investment to GDP and the ratio of remittances to total output respectively Y is a set of variables that 'lffect investment outlays, including economic growth, lending rate and openn:-:s;; as r.r.entioned above Meanwhile, u represents error term

Step 2: Next, financial development will be added in the model to examine how remittances affect investment as the role of financial development is considered Simultaneously, which relationship between remittances and financial development, complementary or substitutionary, on investment outlays is also explored via the interaction of them in the following specification:

INVi,t = ao + ~~REMi,t + Y1FDi,t + Y2REMi,t*FDi,t + ~2,Yi,t+ ui,t (lb)

As in the model, the interaction between remittances and financial development stimulate investment is indicated through coefficient y2 The positive sign of y2 implies that remittances are complementary to financial development in inducing investment Otherwise, the negative sign implies that remittances are substitutionary to financial development in boosting investment

Step 3: Finally, one more variable, institution quality, is included in the model

m order to examine the effect of remittances on investment through different channels including financial development and institutional quality This specification is constructed based on the model by Bjuggren eta! (20 1 0):

INVi,t = ao + ~~REMi,t + Y1FDi,t + Y2REMi,t*FDi,t + y3REMi,t*INSTi,t+ ~2,Yi,t

+ ui,t(lc)

Here coefficient y3 shows the interaction between remittances and institution quality, and the sign of y3 points out which role of institution, substitutionary or complementary, in remittances-investment nexus

3.3 Methods of parameter estimations:

The relationship between remittances and investment will be examined using different empirical techniques Ordinary Least Squares (OLS) is employed and then Fixed Effects (FE) estimator is employed to run regression models in static forms Yet, this approach encounters some issue such as autocorrelation and endogeneity that may lead to biased coefficients, so the solution to these problems is Generalized Method of Moments employment of (GMM) estimator in dynamic models Each estimator is presented in more detail as belows:

3.3.1 Ordinary Least Squares (OLS) estimator:

Given a model as follows:

Yit = ao + X\t~ + ui,t (*)

Where Yit is dependent variable, Xi,t is a (K-l)xl vector of exogenous regressors and ui,t - N(O,cr£2) is random disturbance The error term ui,t in OLS

3.3.2 Fixed Effect (FE) estimator:

As running regression with fixed effect method, the equation (*) will be written as follows:

re-Yit = ao + X\t~ + <>i + Eit (**) Now, in the FE model, ui,t is the sum of ()i and Eit Where ()i represents individual specific fixed parameter to be estimated; and Eit is independent and identically distributed errors with zero mean and constant variance It means

a£2~0

E(Ei,t.Ej,s) = 0, i:;t:j, t:;z!:s, E(Oi.Ej,t) = 0 V i,j,t, E(Xi,t.Ej,s) = 0 V i,j,t,s

In fixed effects model, slopes are constant whereas intercepts that differ according to the cross-sectional unit, namely here the country In this type of model, there are no significant temporal effects; yet there are significant differences among countries However, heteroskedasticity problem could easily arise from country-specific differences in this type of model that would further plague estimation Solution to this problem is the choice of cross-section weights in GLS weights and the use of White coefficient covariance estimator The former (cross-section weights) would allow for the presence of cross-section heteroskedasticity; whereas the latter (White coefficient covariance method) facilitates acquistion of robust coefficient standard errors (Yaffee, 2003)

in financial development enables remitance to flow in more or because when those flows through formal financial institutions increase that leads to a larger measured percentage of remittances In addition, financial development might reduce cost of

transmitting remittances, leading to an increase in such flows (Aggarwal et a/

2005)

Furthermore, sometimes autocorrelation inheres within panels from one time period to another, or across observations in panel models, especially with dynamic ones Arellano and Bond ( 1991) suggest a solution to that problem by adding more lags of dependent variables into models with a condition that temporal observations are larger than regressors of the models to account for dynamic effects and by taking first difference to deal with autocorrelation They can handle the problem of

missing data but still ensure efficient gains thanks to the greater number of moment conditions GMM models may overcome issues like to heteroskedasticity and nonnormality, so it is recommended to employ generalized method of moments and combine White and Newey-West estimators to yield robust standard errors in cases

of estimation suffering heteroskedasticity, autocorrelation problems (Y affee, 2003 )

3.3.3 Generalized Method of Moments (GMM):

As mentioned above, the big drawback of the static approach including Ordinary Least Square and the fixed effect estimation encounters several issues such as autocorrelation and endogeneity problems of remittances, financial development and institutional quality to investment that lead to biased coefficients Therefore, GMM estimator in the spirit of Arellano and Bond (1991) is an alternative to solve those problems GMM technique is explained based on a linear dynamic panel data model as follows:

Yit = LpjYit-j + X'itf3 + oi + Eit (a)

Where Yit-j represent lagged dependent variables, X is a set of control variables,

oi is unobserved fixed effect and Eit is error term The dynamic aspect lies in including lags of the dependent variable in the model, so this is called dynamic approach After taking first-differencing this specification to eliminate the individual effect, GMM technique is applied to estimate

~Yit = Lpj~Yit-j + ~\tf3 +~ Eit (b)

Under GMM estimator, the equation (b) will be estimated by employing a different number of instruments for each period in which each period-specific instrument includes corresponding lagged dependent and predetermined variables available at a given period Thus, period-specific sets of instruments are established from corresponding Jagged values of the dependent and other predetermined variables As for the equation (b), if the innovations in the original equation are independent and identically distributed (i.i.d.), then Yi,I is a valid instrument in t = 3 since it is correlated with ~ Yi,2, but uncorrelated with ~Ei,3• Similarly, both Yi,2 and

Yi,I and are potential instruments in t = 4 As a result, a set of predetermined instruments for individual i using lags of the dependent variable will be established Similar sets of instruments may be formed for each predetermined variables (please refer to Panel Estimation in Eview 6 User Guide II)

In the context of GMM estimation, some assumptions are made in order to generate consistent estimates of the parameters in the equation Typically, the error term is assumed to be serially uncorrelated and the explanatory variables are weakly exogenous

Nul hypothesis Testing:

The consistency of the GMM estimates depends on the soundness of the instruments The Sargan test is employed to verify independence between the instruments and the error term The null hypothesis in this case is that the instruments and the error term are independent or over-identifying restrictions are valid Thus, a failure to reject the null hypothesis means that the instruments are indeed valid The Sargan test is didtributed as chi-square under the null hypothesis

As test results for null hypothesis are not stated directly in Eview, but through results which are calculated based on reported J-statistic and the instrument rank (Please refer to method of calculation of p-value for null hypothesis test in the appendix)

3.3.4 Application of GMM estimator in the thesis corresponding to each specification:

Following basic principles of GMM estimation presented above together with each analytical step, here after specifications will be re-written in dynamic forms and taken first difference to eliminated unobserved country-specific fixed effects before applying GMM technique as follows:

Step 1: Considering the impact of remittances on investment without financial development and institution quality

INVi,t = ao + LDslNVi,t-s + ~tREMi,t + ~2,Yi,t + f.li + Vi,t (2a)

After taking first difference

~INVi,t = ~siNVi,t-s + ~~dREMi,t + ~2·~ Yi,t + ~vi,t (3a)

Step 2: Considering the impact of remittances on investment inclusive of financial development

INVi,t = Uo + LDsiNVi,t-s+ ~1REMi,t + Y1FDi,t + Y2REMi,t*FDi,t + ~2,Yi,t + J.li + Vi,t (2b) After taking first difference

~INVi,t = ~siNVi,t-s + ~1dREMi,t + Y1MDi,t + Y2dREMi,t*FDi,t + ~2·~Yi,t + ~Vi,t

(3b)

Step 3: Considering the impact of remittances on investment inclusive of financial development and institution quality

INVi,t = Uo + LDsiNVi,t-s + ~1REMi,t + Y1FDi,t + Y2REMi,t*FDi,t + y3REMi,t*INSTi,t+

~2·Yi,t + J.li + vi,t (2c)

After taking first difference

Y3~REMi,t*INSTi,t+ ~2·~Yi,t+ ~vi,t (3c)

Where INVi,h INVi,t-s are investments at survey & previous periods; Yi,t is a set

of control variables which are presented in variable description; REMi,t is the ratio

of remittances to total output J.li is an unobserved country-specific fixed effect such

as the state of technology, geography and the like; vi,t is the error term Assuming that the error term vit is not serially correlated and that the independent variables are weakly exogenous

An important point should be paid attention that lots of literature gets around autocorrelation issue in models by averaging observations over 5-year periods (see for example: Giuliano and Ruiz-Arranz, 2006) It is explained that five years is thought to be long enough to eliminate business cycle effects, but short enough to capture important changes that occurs over time for a particular country However, that approach seems to be arbitrary and leads to excessive loss of information (Bjuggren et al., 2010) They propose the alternative by adding lagged dependent

variables on the right hand side of the equation and regresses annual observations

- - ·

-Specifications in the thesis are run regression with annual observations in the spirit

of Bjuggren et al., (20 1 0)

In short, the chapter introduces research scope, data source as well as variables

of interest employed in empirical models It also presents analytical steps, going from a simple model to advanced ones in order to clearly state the impact of remittances on domestic investment in cases of they go alone and of they are combined with channels including financial development and institution quality Furthermore, different empirical techniques are also described together with how to test significance of the employed methods Each method has pros and cons in which OLS is rather simple but encounters issues such as heteroskedasticity, unobserved country effect; FE may face autocorrelation and endogeneity GMM seems to be the most feasible solution to these problems

- - - - - -

-CHAPTER IV: EMPRICAL RESULTS

In this chapter, an overview of East and Southeast Asia is stated based on collected data, descriptive statistics as well as bivariate correlation between variables of interest in the appendix After all, empirical results are stated according

to each empirical technique In each technique, data tables and analysis of each variable are once again presented separately into each step as stated in chapter III for convenient follow up With regard to the step 3, both financial development and institution are added in the model in which there are 3 indicators proxied for the former and 6 ones for the latter, so for simple presentation and saved space, regression-result-tables and analysis are presented in 3 parts based on 3 indicators

of financial development Going along with estimated results, results of null hypothesis tests are also indicated

4.1 Descriptive statistics:

4.1.1 Remittances in East and Southeast Asia:

Remittance inflows stand at average level of 2.55% and constitute large shares

of GDP in some of the poor countries such as Philippines (13 72% ), Vietnam (8%) and Cambodia (4.1%) Meanwhile, these flows account for a small part in GDP in more relatively developed countries Typically, Hong Kong is the country receiving least this source at 0.07% of total output in term of remittances over GDP and standing right after Hong Kong is Korea with the figure of 0.1 % This indicates that remittances play an important role in less developed and developing countries as they contribute to lift up earnings of residents, reducing economic pressure, namely pressure on consumption This also partly shows outstanding characteristic of remittance inflows: altruism It means that these money sources are transferred by migrants to their home country in the aim to support their relative, especially in the context of poor economy or economic downturn That is why lots of research

sought the impact of remittances on poverty m these countries, especially Philippines (Yang, 2004)

4.1.2 Investment and economic growth in East and Southeast Asia:

The representative level of investment over GDP for the regions is on average 27.3% over the period Yet, there is a considerable spread over the period from

1995 to 2008 For instance, the lowest share of investment over GDP belonged to Indonesia with 11.3 7% whereas China invests close to 45% of GDP, ranking first in the regions However, considering totally, Cambodia's investment share out of total output is modest compared to those of other economies but tends to increase As for Vietnam, this index gets increasingly higher over the period and reached the climax

at 43.13% in 2007

Investment is the key motive to economic growth and that relationship is clearly reflected through growth rate in China and in the other countries in the regions China has relatively high and stable investment-GDP ratio and also achieves the highest economic growth at 14.2% in comparison to the other Noticeably, almost economies in East and Southeast regions, except for 3 countries: China, Cambodia and Vietnam, saw a decrease in investment share out of GDP that led to negative economic growth in 1998 right after monetary crisis 1997 Thailand

is the first country in which the crisis broken out had to suffer negative growth in two consecutive years 1997 and 1998

4.1.3 Financial development, openness and lending interest rate in East and Southeast Asia:

Generally, Cambodia has the least developed financial system in term of measures of financial development Meanwhile, Hong Kong and Malaysia share first position corresponding to each measure For instance, Hong Kong leads in expansion of money supply (M2) over GDP with maximum level of 378% Malaysia is in the first position in two indicators of banking credit over GDP and private credit over GDP with 221.80% and 210.4% respectively Noticeably, Hong