Determinants of domestic savings in asean developing countries in the period 1986 2000, the case of VN

Basing on the integrated theory of savings and previous empirical works, this study also found income growth, demographics, financial liberalization and financial development are main fa

VIET NAM- THE NETHERLANDS PROJECT FOR M.A ON DEVELOPMENT ECONOMICS

DETERMINANTS OF DOMESTIC SAVINGS IN ASEAN DEVELOPING COUNTRIES IN THE PERIOD 1986-2000,

THE CASE OF VIET NAM

'

,

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

MASTER OF ARTS IN ECONOMICS OF DEVELOPMENT

Academic Supervisor: Dr NGUYEN TRQNG HOAI

HOCHIMINH CITY, NOVEMBER 2003

ACKNOWLEDGMENTS

First of all, I would like to express my sincere thanks to all teachers and the st~ff in Viet Nam - the Netherlands Project for M.A on Development Economics, especially Mr Tran Yo Hung Son, Mr Nguyen Huu Dung and Ms Tran Thi Ben They have created an excellent environment for our study Many thanks are also released to Ms Dinh Anh Nguyet, the project secretary and Ms Dang Kim Chi, the project librarian for their assistance during the whole course

I am grateful to Dr Karel Jansen, Dr Youdi Schipper and Dr Joost Buurman for their comments on the thesis proposal Especially, I am indebted to the knowledge got from

Dr Karel Jansen in the field Money and Banking, his useful and interesting lecture as well as works relating to finance dropped a hint for me to follow this research

,I am so fortunate to be guided by Dr Nguyen Trang Hoai His deep understanding on Econometrics and Savings is invaluable for my thesis His enthusiasm and encouragement made me more confident in the difficult times I hope my result is not a disappointment to him

I would like to thank all my classmates, friends and colleagues at Labor and social · Affairs School HCM city for their support in my study process

Last but not least, God bless my parents and lovely relatives There is no word to reveal

my love to them I hope they would consider this thesis as one of worthy gifts from me

an ongoing liberalization of their finance, trade and investment regime However, from the

1997 financial crisis and its consequences for ASEAN countries' economies, one of the lessons withdrawn is that these countries passively depended much on foreign capital Therefore, realizing the determinants of domestic savings in these countries is necessary to mobilize and well use the available resource for high and sustainable economic growth Basing on the integrated theory of savings and previous empirical works, this study also found income growth, demographics, financial liberalization and financial development are main factors explain for the domestic savings in Southeast Asian countries The regression result gave out the positive impacts of income growth on domestic savings in VietNam and Indonesia; the positive effects of financial development on domestic savings in Viet Nam and Malaysia Furthermore, the negative impacts of age dependency ratio and real interest rate on domestic savings are revealed in all five countries

In order to mobilize domestic savings for modernization and industrialization, Vietnamese government should try to stabilize socio-economic problems, control inflation, reform state-owned enterprises and encourage the development of private sectors, continue to improve the effectiveness of financial and banking sectors as well as birth-control and family pla'nning programme

ABBREVIATIONS

ADB: Asian Development Bank

A SEAN: Association of Southeast Asian Nation

DEP: Age Dependency Ratio

FDI: Foreign Direct Investment

GOP: Gross Domestic Product

GDS: Gross Domestic Savings

GSO: General Statistics Office

IMF: International Monetary Fund

M2: Broad Money ;:-;,

ODA: Official Development Assistance

RIR: Real Interest Rate

USD: United State Dollar

VND: VietNam Dong

WDI: World Development Indicators

TABLE OF CONTENTS

ABSTRACT , 1

ABBREVIATIONS 11

LIST OF TABLES V LIST OF FIGURES V CHAPTER 1: INTRODUCTION 1

1.1 PROBLEM STATEMENT 1

1.2 OBJECTIVES OF THE STUDY 2

1.3 DATA AND RESEARCH METHODOLOGY 3

1.4 STRUCTURE OF THESIS 3

CHAPTER 2: LITERATURE REVIEW 5

2.1 DEFINITIONS 5

2.2 SELECTIVE THEORIES OF SAVINGS 6

2.2.1 KEYNESIAN SAVINGS FUNCTION " 6

2.2.2 THE PERMANENT INCOME HYPOTHESIS 7

2.2.3 THE LIFE CYCLE HYPOTHESIS 9

2.2.4 THE McKINNON- SHAW'S FINANCIAL LIBERALIZATION 14

2.2.5 INTEGRATED SAVINGS THEORY 21

2.3 EMPERICAL STUDIES OF DETERMINANTS OF SAVINGS 22

2.3.1 INCOME AND ECONOMIC GROWTH 22 ·

2.3.2 DEMOGRAPHICS 24

2.3 3 REAL INTEREST RATE 25

2.3.4 FINANCIAL DEVELOPMENT 28

2.3.5 THE STUDY OF VIETNAM INSTITUTE OF FINANCE 29

I 2.4 SUMMARY 29

CHAPTER 3: DETERMINANTS OF DOMESTIC SAVINGS IN ASEAN COUNTRIES: AN OVERVIEW · 31

3.1 SIMILARITIES OF THE FIVE ASEAN COUNTRIES 31

3.2 GROWTH, DEMOGRAPHICS AND FINANCIAL LIBERALIZATION 31

3 2.1 A SEAN ECONOMIC GROWTH 31

3.2.2 ASEAN POPULATION BOOM, HIGH DEPENDENCY RA TJO 33

3.2.3 FINANCIAL LIBERALIZATION IN ASEAN COUNTRIES 36

3.3 SITUATION AND POLICIES ENCOURAGING SAVINGS IN VIETNAM 41

3.3.1 OVERVIEW OF ECONOMY AND ACCUMULATION BEFORE 1988 41

3.3.2 STABILIZATION, REFORM AND CONSOLIDATION AFTER 1988 43

l3.3 IMPACTS OF ECONOMIC REFORM ON SA VINGS 45

3.4 SUMMARY 52

CHAPTER 4: DETERMINANTS OF DOMESTIC SAVINGS IN ASEAN DEVELOPING COUNTRIES, THE CASE OF VIET NAM 55

4.1 MODEL SPECIFICATION AND ESTIMATION METHOD ~ 55

4.2 DATA SOURCES 60

4.3 ESTIMATION RESULTS 60

4.4 MAIN FINDINGS 68

CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS 73

5.1 CONCLUSIONS OF THE THESIS 73

5.2 POLICY IMPLICATIONS 74

5.3 SUGGESTIONS FOR FURTHER STUDY 77

APPENDIX 1: DATA SET (0/o) 78

APPENDIX 2: DESCRIPTIONS OF HYPOTHESES TESTING 79

APPENDIX 3: REGRESSION RESUL TS 81

REFERENCES 84

LIST OF TABLES

LIST OF FIGURES

CHAPTER 1: INTRODUCTION

1.1 PROBLEM STATEMENT

A main problem in economic development theory is how to increase the resources available for capital formation, or how to control consumption and increase savings (Jansen 1990, p.8I)

The allocation of resources between present and future consumption (savings) is one of the most fundamental economic choices facing any economy This choice affects not only the rate of economic growth a country can enjoy, but also the standards of living for future generations jet unborn (Gillis, Perkins, Roemer and Snodgrass I 996, p.304)

Traditional development theories have put much emphasis on the importance of savings in determining economic growth rate in developing countries

According to Jansen (1990), the resources available for investment and economic growth in any country can be increased through more domestic savings or through a larger capital inflqw from abroad Although many developing countries relied heavily on foreign savings

as a source of investment finance, the rise in investment ratios was accompanied by roughly commensurate increa~e in the share of gross domestic savings in GOP (Gillis, , Perkins, Roemer and Snodgrass I 996, p.305)

a-The high ratios of domestic savings and investment are one of the striking features of East Asian and Southeast Asian economies ln reality, these high ratios have been reliable foundation for the higher economic growth rate of these countries (Le et all2002, p I 00) In general, there are many evidences showing the same trend in which domestic savings and

economic growth rates go in the same direction Basing on this, many governments in developing countries attempt to raise savings rate in order to achieve high and sustainable

growth rates Therefore, it would be necessary to deeply understand what determine

2000 (WDI, 2002) However, according to Le (1997), it is able to see that VietNam is in the

'Determinants of domestic savings in ASEAN developing countries in the period 1986

2000, the case of VietNam'

1.2 OBJECTIVES OF THE STUDY

Basing on the integrated theory of savings and previous empirical studies on determinants of savings in developing countries, this thesis aims to view the impacts of some main factors such as income growth, age dependency ratio, real interest rate and financial development

on domestic savings in selected ASEAN countries (Indonesia, Malaysia, the Philippines, Thailand and VietNam) in general and especially for the case of VietNam This research is therefore designed to investigate and answer the following key research questions:

• Whether are there positive impacts of income growth, real interest rate and financial

development on domestic savings?

• Is there negative relationship between age dependency ratio and domestic savings?

1.3 DATA AND RESEARCH METHODOLOGY

To investigate the relationship between domestic savings and income growth, age dependency ratio, real interest rate and financial development in selected ASEAN countries

in general and VietNam in particular, this research uses time-series and cross-section data (panel data) for the period 1986 - 2000 In the thesis, all data are secondary and taken from

2002 World Development Indicators CD-ROM of the World Bank, Asian Development Outlook 2002 of the Asian Development Bank, and the Viet Nam Statistical Yearbooks of GSO as well as data from update researches of VietNam Institute of Finance

In this thesis, both qualitative and quantitative methods are applied, however the regression econometric analysis is considered as the main methodology From the collected data, we can examine the relationship between domestic savings and other variables by making the ' regression so that we are able to explore the effects of each factor on domestic savings

1.4 STRUCTURE OF THESIS

The thesis is composed of five chapters Following the introduction, chapter two presents the literature review and empirical studies Because no single theory can completely explain the

determinants of savings in developing countries, the integrated theory of savings is applied

as the analytical framework of the research Chapter three provides a brief overview of determinants of savings in five ASEAN developing countries in general and Viet Nam in

'

particular in order to provide the background for understanding the situation and solutions encouraging domestic savings in these countries Chapter four covers the core study of the thesis It discusses the model specification and estimation method, gives the basic findings

of the regression results and answers the research questions Finally, chapter five summarizes the main findings of the thesis This chapter also presents policy implications drawn from the main findings and offers some suggestions for further study

CHAPTER 2: LITERATURE REVIEW

This chapter will give definition of the key concepts relating to savings, theory review and the empirical evidences The clear definition will help the taxonomy of savings more favorable for testing its determinants The theory review focuses on some main theories used

to explain determinants of savings such as Keynesian savings hypothesis, Permanent income hypothesis, Life cycle hypothesis, McKinnon and Shaw approach relating to financial liberalization that is expected to encourage savings and then the integrated theory of savings The last one will present some empirical studies, which examine the relationship between savings and its determinants

Private domestic savings also arise from two sources: corporate savings and household savings Corporate savings is defined as the retained earnings of corporate ·enterprises (corporate income after taxes minus dividends paid to shareholders) Household savings is

simply that part of household income not consumed Household savings includes savings from unincorporated enterprises (single proprietorships, partnerships, and other

noncorporate forms of business enterprise) (Gillis, Perkins, Roemer and Snodgrass, 1996)

In this thesis, for regression, I base on World Development Indicators (2002) that the gross domestic savings are calculated as GOP less final consumption expenditure.· Final consumption expenditure is the sum of household final consumption expenditure (private consumption) and general government final consumption expenditure (general government consumption) I consider the gross domestic savings in form of percentage of GOP

2.2 SELECTIVE THEORIES OF SAVINGS

2.2.1 KEYNESIAN SAVINGS FUNCTION

Mikesell and Zinser (1973) presented the Keynesian savings (consumption) function is linear with a constant marginal propensity to save This function was formulated in the context of a short-term model in which macroeconomic fluctuations occurred (Jansen 1990, p.82) Hence savings, a residual of income and consumption, is mainly affected by current income But it is observed that the savings ratio is relatively constant in the long run but shows considerable fluctuations in the short run (Jansen 1990, p.83) The research of Mikesell and Zinser (1973) relating savings to current income suggested two general tendencies First, as moving from gross domestic savings to private and personal savings, the relationship between savings and income becomes more proportional for different levels

of per capita income, and over time the gap between marginal and average savings rates tends to disappear Second, the results of cross-sectional analysis tend to show a higher marginal propensity to save and a larger divergence between marginal and average savings propensities than do results derived from time series regressions One explanation of these

findings is that savings behavior is determined not only by current income but also by past income levels, the rate of income growth and the age distribution of households of the community

Determinants of Savings Derived from Keynesian Approach

The Keynesian theories establish a close linkage between the level of savings and the level

of current income As income rise, consumption also rise but less proportionately Additional income part would be saved From this simple perspective, one would expect that there is positive relationship between the level of savings and the level of income (Mikesell and Zinser, 1973)

Rather than linking savings to current income, as in Keynesian approach that can face these paradoxes above, the Permanent income hypothesis relates them to permanent income and tht:( Life cycle hypothesis to life-time income (Jansen 1990, p.83)

2.2.2 THE PERMANENT INCOME HYPOTHESIS

Friedman's 'Permanent income hypothesis' is the starting point for a variety of specifications of saving-income relationship (Mikesell and Zinser, 1973) In its most simple form, the linear equation is:

(Gillis, Perkins, Roemer and Snodgrass 1996, p.313)

Where Yt is transitory income and Yp is permanent income Permanent income is defined in terms of a long-run expectation over a planning period, and transitory income is the

7

difference between current income (Y) and permanent income Friedman's hypothesis is that individuals consume virtually no transitory income This implies a heavy reliance on past behavior as a determinant of consumption spending; but changes in transitory income will

I immediately leads to changes in the level of saving

Because savmgs m this theory is defined as a difference between current income and permanent income so it equals transitory income This means that any increase in transitory income will lead to an increase in household savings by the same amount As transitory (unexpected) income can be positive or negative depend on the values of current and permanent income, hence, so does savings

An interesting aspect of the Permanent income hypothesis is that it may help to explain the saving behavior of difference income groups Self-employed persons generally have a greater transitory component in their income than wage and salary workers, and hence they are likely to show a higher ratio of savings to observed income (Jansen 1990, p.85)

Determinants of Savings Derived from the Permanent Income Hypothesis

In Permanent income hypothesis, income variable is introduced as a key determinant of savings Consumption decisions and thus savings decisions are made on a basic of, permanent income which is viewed as a stream of income derived from total wealth

This theory assumes that household's consumption decisions depend on permanent income while keeping their total wealth intact Hence, if the interest rate increases, the present value

of total wealth will decline, thus permanent income, therefore, transitory income or savings will increase

Among theories of savings, the Permanent income hypothesis helps us to explain the positive impact of exports on savings in developing countries These countries mainly expot1 primary goods These export earnings suffer from great instability and fluctuation, however

As a result, export earnings have a larger transitory component and thus most of them are saved This leads one to expect that countries with higher share of export sector would have higher saving ratios

'

The Permanent income hypothesis has in common with the Life cycle hypothesis the property that long-term income is assumed to be the primary determinant of consumption and savings However, the main different between the two theories is that the Permanent income hypothesis focuses on income fluctuation, while the Life cycle hypothesis on the · fitness of life and the plan for retirement (Wai, 1972)

2.2.3 THE LIFE CYCLE HYPOTHESIS

The Life cycle hypothesis associated with the writings of Modigliani, Brumberg and Ando · postulates that individuals adopt a planning horizon for their life-time consumption It is , assumed that individuals plan no net life-time saving but attempt to spread their life-time consumption evenly over their lives by accumulating enough savings during their earning years to maintain the consumption standard during retirement (Mikesell and Zinser, 1973, p.ll ) It also means that they set their life-time patterns of consumption and savings so as to maximize utility subject to life-time budget constraint

According to the Life cycle hypothesis, households' income fluctuates systematically during their life-time In the period of nonworking age, because of the individuals' low earnings, most of expenditures are financed by available wealth or borrowing lead to negative savings rate for this age group In other hand, thanks to the relatively high earnings during the working age (15 - 64) compared with other stages, savings are indispensable in o~der to maintain their consumption standard during the life-time

This model, therefore, shows that age composition of households and their current incomes have important effects on savings For more detail, in a society with a growing population and/or growing per capita income, aggregate net personal savings is positive because the working population tends to be larger than the retired one, and the higher the level of current capita income the larger will be the amount of saving necessary to maintain an individual's consumption level in retirement (Mikesell, and Zinser 1973, p.ll )

Briefly, the level of savings will depend on the structure of population in each stage If population rapidly grows, savings tend to decrease because the amount of minors rises faster than the increase of population in working age The decrease of savings in this case is relatively large because the earnings of infants are often zero while the expenditures for them are not small However, if population growth rate decreases, savings are able to increase strongly because the population in the minor age rises slower than the increase of population in working age and despite the raise of population in retirement age, the decrease ,

of savings in this age group is not too large owing to the demand for consumption tends to

'

reduce when income decreases and individuals face the last stage of their life-time Furthermore, this age group often has income from accumulated wealth during the working age

Figure 2.1 Schematic Life-Cycle Profiles of Earnings and Consumption

Source: Figure 3.1, Deaton, A (1999)

Determinants of savings de1·ived from the Life cycle hypothesis

Jansen ( 1990) reviewed the researches of Mikesell, Zinser and Modigliani that strongly gave support to the rate Of growth as a determinant of savings ratio In short, variables such as GOP growth rate can be used as a proxy for the income variable which is of a positive impact on the level of saving in an economy

The Life cycle hypothesis predicts a relationship between the age structure and the level of savings In other words, it establishes a close link between patterns of savings and pattern of age composition For simplicity, the most widely used variable to represent demographic

variables is the dependency ratio The increase in dependency ratio will lead to decrease in savings ratio

It is argued that interest rates work in ambiguous directions with respect to savings An increase in interest rate may either raise savings This is because higher interest rates simultaneously have both substitution effect and income effect (Jansen 1990, p.95) Firstly,

it increases the present value of wealth, which has a positive effect on current consumption Secondly, it leads to substitution between current and future consumption, which· has a negative impact on current consumption These effects work against each other That is exactly the reason why the impact of higher interest rates on saving is unclear

Summary, we can assess these above saving theories (Keynesian savmgs function, Permanent income hypothesis and Life cycle hypothesis) like that, to some extents, they are developed in and for industrialized countries (Jansen 1990, p 91 ) In applying these theories

to developing countries at least three problems affecting on saving decisions could be faced The first of these is that developing countries are going through a process of rapid structural change Secondly, most households are primarily production units and not only consumption units The third problem is that the assumption of perfect markets, that underlines most of the saving theories, is not necessary fulfilled (Jansen 1990, p.91 )

Structural change: The simple Keynesian saving function linking the level of saving to the

level of income is not very meaningful According to Jansen (1990), the increase in the savings ratio may be not so much determined by the increase in income as by the long-term structural changes in the economy that occur with economic development Such changes may include the modernization of agriculture, the growing share of the corporate sector and

the changing role of government The impact of these processes of structural changes on the overall savings ratio will be felt only in the long-term and they may dominate any effect that a higher per capita income may have The effects of structural change are nor easily

'

predictable in reality

Production units: The saving theories above are derived from consumption behavior But

many, if not most, households in developing countries are simultaneously consumer and producer (Jansen 1990, p.92) Hence saving decisions and the acquisition of assets are determined not only by a desired consumption pattern over time but also by the conditions

of production in the household enterprise

Imperfect markets: The Keynesian saving function, the Permanent income hypothesis and

Life cycle hypothesis assume utility-maximizing households whose behavior is not constrained by market imperfection It means that the household can decide on an optimal saving-consumption pattern over time However, it seems to be not the case in developing countries where capital markets and employment condition are quite different from developed ones There is no retirement, people work as long as they can; children contribute

to household income after a certain age and often look after their parents in old age This makes the pattern of age-earning profile of households unclear Moreover, capital markets are far from perfect and often hardly accessible so they make households more difficult iiT smoothing consumption over their lifetimes

Generally speaking, they contribute not enough understanding of saving behaviors in developing countries Therefore, in order to explain the determinants of domestic savings in these countries we should apply one more theory relating to savings

2.2.4 THE McKINNON- SHAW'S FINANCIAL LIBERALIZATION

The case for financial liberalization does touch on most aspects of developing countries

'

financial markets The primary policy prescription arising from McKinnon and Shaw's work

is the general freeing and raising of institutional interest rates and/or a reduction in the rate

of inflation It would appear the two policies are treated as identical, both policies increasing the real rate of interest and thereby increasing the savings rate (Sikorski 1996, p.66)

• Theoretical Foundation: McKinnon (1973)

The first theoretical pillar in the house of financial liberalization is the premise that interest rates have a positive relationship with economic growth The formal analysis of McKinnon adopts the assumption that all economic units are confined to self-finance and that in these countries, indivisibilities in investment have an important role to play The first assumption

I

drives McKinnon's model by implying that any potential investor must have accumulated all the money balances necessary for investment prior to undertaking the project A higher real deposit rate of interest makes the opportunity cost of saving real balances to invest lower, and hence acts as an impetus for firms wanting to finance investment projects

• Theoretical Foundation: Shaw (1973)

Shaw's theoretical contribution is on the role financial intermediaries have to play in , development A case for financial liberalization, in terms of increasing interest rates, arises mainly from expanding the amount of financial intermediations occurring between savers and investors By increasing the returns offered to savers, financial intermediaries' capacity

to lend is increased and the banks are able to allocate this larger volume of investment funds

Shaw's model relies on neo-classical market-clearing assumptions, specifically the assumption that the interest rate functions in an equilibrating manner to ensure equality

between the supply and demand of loanable funds in the financial markets Given efficient

financial markets in Shaw's world, the decision to reduce consumption and increase savings does not reduce demand but, rather, alters its composition from consumption towards investment expenditure

In general, one of the main content of McKinnon and Shaw's model is that the real interest rates have positive effects on the level of savings Hence financial liberalization that makes increase in real rate of interest is an effective way to escape from the context of poor capital accumulation

Because currency is a central field in the market economy so the government often meddles

in this one These interventions take various forms such as interest ceilings, reserve requirements, selective credit programs with subsidies interest rates, of which the most widely financial instrument used in a financially repressed economy is interest rate that is usually set below its equilibrium level in order to stimulate economic growth and improve employment condition McKinnon and Shaw criticize this way According to them, the low real interest rates will lead to the poor level of savings Hence the level of investment is limited and the low efficiency of investment is a next consequence of this mechanism.·· Therefore, it should be financial liberalization in order to gradually make interest rate to its equilibrium level We can present their assumptions for the model like that:

The supply of deposits equals the supply of savings in the financial market because all kinds

of saving are concentrated on banks in form of bank deposits that are provided to inve~tment

funds by financial intermediaries

All expenditures for investment are financed only by bank deposits It means there ts equality between the demand for investment and demand for deposit

'

Both saving and investment are functions of real interest rate However, there is a positive relationship between savings and real interest rate meanwhile the relationship between investment and real interest rate is vice versa With the expectation that individuals maximize their utility of consumption, once the real interest rate increases/decreases, the current expenditure will be relatively expensive/cheap compared with the future one Hence current consumption will be adjusted to reduce/rise, in turn, saving tends to increase/decrease In other hand, investment will be harmed/encouraged when the real rate

of interest increases/decreases owing to the profit maximization

With the assumptions above, the relationship between saving, investment and interest rate can be graphically illustrated as the figure below:

If institutional interest rates (both deposit and lending rates) are imposed at r1, then the real level of savings is at S1, lower than the possible level of savings (So) at the equilibrium rate

of interest, hence the volume of investment is restricted at this level (S1) It reveals that the purpose of maintaining low rates of interest in order to promote investment does not meet demand Furthermore, because of the excess demand for credit at r1 leads to the non-price rationing of credit occur in the economy Consequently, the volume of capital supply is low; the production process is distorted and the financial system is eroded

In contrast, if the currency management is liberalized (or expanded), the rate of interest will rise and get the equilibrium level The volume of savings will shift along the supply curve to the equilibrium rate, of which lead to the increase of real investment Hence going with the increase in quantity, the efficiency of investment is also improved because the low profitable projects may no longer have change to receive sponsors and the credit allocations be also constrained to some extent As a result, there is improvement of economic growth, in turn, shifts the savings supply curve to the right [S(Yo) => S(Y1)] In other word, the level of saving will increase with unchanged rate of interest Investment will be better not only in quantity but in quality with this interest rate The economic growth in turn will be strengthened and this process be continued

Therefore, increasing the real rate of interest will affect on savings through the two

ch~nnels-• ch~nnels-• Direct way: rising the real interest rate will raise the level of savings

• Indirect way: raising the real interest rate will improve the efficiency of investment, the economic growth and thereby, the level of savings will be better

McKinnon ·and Shaw believed that there are at least two beneficial effects of financial liberalization These effects are known as saving effect and efficiency effect

• Saving effect: increasing the real rate of interest will raise the level of total savings, and thus the level of investment and growth rate

• Efficiency effect: an increasing in real interest rate will improve the efficiency of investment by appropriately allocating resources in the economy, thereby stimulating the economic growth rate

Determinants of Savings Derived from the McKinnon and Shaw Model

It argues that real interest rate has positive impact on savings Increasing the real rate of interest will affect on savings through the two channels: direct and indirect way A World Bank Policy Research Report (1995) revealed that several studies had documented the positive association between real interest rates and the growth of savings deposits

In this model, a case for financial liberalization, in terms of increasing interest rates, arises mainly from expanding the amount of financial intermediations occurring between savers and investors These two kinds of customers will have more choices to save and invest Wai (1972) and Gupta (1984) found a positive relationship between ·savings and financial development in some countries and regions Given data problems encountered in many countries, there has been a tendency to use rather simple indicator, such as the number of bank offices per head of population or total bank assets as percentage of national income -

or even simpler the M2/GDP and M3/GDP ratios - as indicators of financial development (Jansen, 1990)

The improvement of economic growth increases the ability to save in the economy Nguyen, C.N and Nguyen, H.H (2000) used variable so-called GOP growth rate as a proxy for the income variable and then found its positive impact on the level of saving in countries studied

However, with the assumptions relating to perfect capital market and the efficiency of financial intermediaries above, the approach of McKinnon and Shaw (Financial liberalization) applied in developing countries also cannot itself explain sufficiently the determinants of domestic savings

To some extents, the derivation of savings function basing on these theories above may be Jess relevant in the.case of developing countries But they are useful for us to withdraw some main determinants of domestic savings in developing countries, for example, the Keynesian savings function, the Permanent income and Life cycle hypotheses consider the impacts on savings of income, demographics variable and interest rate; the McKinnon and Shaw financial liberalization explains the effects of real interest rate, financial development (M2/GDP) and income on savings Besides these variables, basing on empirical studies we can find other determinants of domestic savings that are not derived from the above standard theories of savings such as inflation, capital inflows and taxation but enable us to fut1her understand the aggregate saving behaviors

The post-Keynesian approach suggested two mechanisms through which inflation can stimulate savings The first is the income redistribution towards profits; the second is the inflation tax on money holdings

In empirical studies of aggregate savings, the inflation rate is often introduced together with the nominal rate of interest to establish the effect of the real interest rate on aggregate savings But the direction of the impact of inflation rate on savings is still arguments Empirical findings have not been able to establish a clear relationship between inflation and savfng Thirlwall (1974) failed to find significant results of the relationship between savings and inflation He argued that the reasons for this may be that different countries may react differently to inflation - so that cross-countries comparisons fail to find significant results -and that inflation may have different causes

Jansen (1990) reviewed the studies of Griffin (1970), Weisskopf (1972), Papanek (1973) found that there are number of mechanisms that capital inflows can impact on domestic savings as follow:

The inflow of capital might have a negative effect on government savings Tax revenue can be used to pay for more consumption expenditure now that aid finances investment

Corporate savings might fail if access to international capital reduces the need to retain profits for self-financing

Household savings might fail if increased availability of foreign exchange that follows a capital inflow leads to a lowering of import controls and increased consumption of imported consumer goods

However, capital inflows could also stimulate domestic savings efforts

In countries with a tight foreign-exchange constraint, capital inflows release forex budget constraints, increasing investment oppotiunities This leads to increased incentive to' save

Capital inflows go straight into higher income groups who have high propensity to save, therefore increasing domestic savings

Most capital inflows initially flow to investment projects which increase income, and therefore consumption and savings The ultimate effect on saving could be manipulated by government

According to Heller ( 1967), development economists have usually regarded the level and structure of taxation as an extremely important variable in the determination of domestic saving, and fiscal policy has been regarded as a means of mobilizing savings to promote growth Although several studies have questioned the relationship between taxation and saving, most of them show that government saving increases with increased revenue (Mikesell, and Zinser 1973, p.15)

2.2.5 INTEGRATED SAVINGS THEORY

With the savings theories above (the Keynesian savings function, the Permanent income and , Life cycle hypotheses; the McKinnon and Shaw financial liberalization), although each of them has each strength in explaining the determinants of domestic savings, no single one can sufficiently explain It is essential that the determinants withdrew from these theories should

be simultaneously survey Therefore, Wai (1972) and other economists have developed an integrated savings theory that includes economic and non-economic factors affecting on

savings and it can be revealed by the function S = f (A, W, 0) where: S is savings, A is

ability to save, W is willingness to save and 0 is opportunities to save

In this theory, decisions to save are determined by three main factors such as the ability to save, the willingness to save and the opportunity to save Each factor is a function of socioeconomic variables Where: ability to save is a function of income (proxied by GOP) and dependency ratio The willingness to save is a function of real interest rate and the opportunity to save is a function of financial development (proxied by M2/GOP) From these descriptions, the savings function suitable for our study maybe like that:

where: GDS is Gross domestic savings (% of GOP)

GOP is Gross domestic product growth rate OEP is Age dependency ratio

RIR is Real interest rate M2 is Broad money (%of GOP)

2.3 EMPERICAL STUDIES OF DETERMINANTS OF SAVINGS

Basing on the integrated theory of savings, and empirical studies about the determinants of savings, following the objective of the thesis, I just present some researches related for, explaining determinants of savings in developing countries

2.3.1 INCOME AND ECONOMIC GROWTH

A natural theoretical framework that is used to think about the correlation between savings ana growth is the life cycle model (Attanasio, Picci, and Scorcu 1999, p.l ) Such a model

might imply both long run· relationship between past growth and current saving rates and between future expected growth and current savings Income growth can be considered as

having an important positive effect on savings since saving and growth have been highly ·

I correlated in the long run

Modigliani (1990) studied the income-savings relation for eighty five developing countries

in the period 1982-88 and showed that the growth variable enters as predicted and is again both well determined and in the range supported by the theory of life cycle hypothesis

Deaton (1990) reviewed the research of Gersovitz (1988) that the cross-country empirical evidence generally supports a positive effect of per capita income growth on saving rates, however, the results are rarely well-determined and ar~ uncomfortably reliant on the treatment of the simultaneity between saving and growth

Fry (1995) estimated the savings functions for eleven Asian developing countries in the period 1961 - 1988 basing mainly on the life cycle hypothesis His results revealed the positive effect of income variable on saving ratio; dependency ratios exert a negative impact

on savmgs

Paxson and Deaton (1998) studied the savings-growth by looking at the cohort evidence In their research, I put out the case of Taiwan and Thailand because there is something interesting Here the time range is different from each country For Taiwan, they use 20' years of data from 1976-95; for Thailand, they use the 1976, 1981, 1986, 1988, 1990 and

1992 rounds of the Socio-economic Survey The result was that the size and sign of these effects differ across countries For Taiwan, the increases in growth can potentially result in large increase in saving rate The result for Thailand indicated the opposite: for most combination of rates of economic and population growth, increases in economic growth

raise the wealth of the very youngest individuals, who are dissavers, causing a reduction in the aggregate savings

2.3.2 DEMOGRAPHICS

One of the most celebrated and most investigated predictions of the life cycle hypothesis is that there should be a relation between aggregate saving and the rates of population and income growth (Deaton 1990, p.76)

Leff (1969), on the basic of a multiple regression analysis of data for seventy-four countries, concluded, 'dependency ratios are a statistically significant and quantitatively important influent on aggregate savings ratios High dependency ratios - and ultimately high birth rates - are among the important factors that account for the great disparity between developed and underdeveloped countries'

Mikesell and Zinser (1973) reported the work of Gupta in 1971 that he partially supported Leffs results but found that they do not hold for all less developed countries Gupta repeated Leffs regressions but disaggregated less developed countries into three groups according to the level of per capita income He found that in the two lowest income categories, dependency ratios were statistically insignificant as determinants of either the average saving rate or saving per capita Only in the highest income category did Leffs fin,dings hold

The study of Ram (1982) estimated savings functions on the basis of country-level data for a total of 121 countries, pertaining mostly to the period 1970-77, by using different specifications and samples The main finding is that, contrary to the conclusions reached in Leffs original work and the fairly extensive literature cited by him as supportihg his

inferences, there is little evidence of a significantly adverse effect of high dependency rate

on aggregate savings in the less developed countries where the issue has the g'reatest relevance The answer for this difference seems to be: partly on account of the sample

I coverage; partly because of the period studied; and partly due to the specifications adopted

However, as he said, the closer one moves to the Leff's sample and specification, the greater

Real interest rate plays an important part in determining savings behavior As an increase or

a decline in real rate of interest will impact on the value of assets and therefore, savings However, it is necessary to make a clear distinction between total savings and financial savings when searching the impacts of real interest rate on domestic savings Here total savings is savings of all sectors in the economy and proxied by ODS while financial savings measure the amount of total savings that is channeled via ·financial assets (Warman and Thirlwall, 1994) It is measured by the change in the stock of monetary assets

Galbis (1977) studying the efficiency of investment resulting from financial liberalization argued that invested resources would transfer from low return sector to high return via the banking system It means that financial savings increases as deposit rates raise while total

savings seems only to be changed in term of structure

The study of Warman and Thirlwall (1994) for the case of Mexico in the period 1960-80 did not support for McKinnon and Shaw approach They stated that most researches testing McKinnon and Shaw hypothesis viewed total savings and financial savings synonymously According to them financial savings is only one type of savings, and as interest rates is raised there may simply be a substitution between financial assets and other assets leaving total savings unchanged

A World Bank Policy Research Report (1995) revealed that several studies had documented the positive association between real interest rates and the growth of savings deposits and broad money aggregate This is especially evident for changes from highly negative to positive real rates of interest For example, when Taiwan, China, raised real interest rates on bank deposits from a negative 300 percent in 1949 to about 8.5 percent in 1953, the ratio of time and savings deposits to money stock rose from 2 to 34 percent in three years Indonesia and Korea achieved similarity dramatic increases in financial savings after stabilizing inflation and shifting from negative to positive real interest rates

Nguyen, C.N and Nguyen, H.H, (2000) reported the study of Fry ( 1995) on the relationship between real interest rate and savings that his research based on the panel data from seven Asian countries (Myanmar, India, Korea, Taiwan, Malaysia, Singapore and the Philippines) during the period 1962-73 The estimated result indicated that a one-percentage point increase in real interest rate is associated with an increase in the domestic savings rate of from 0.16 to 0.21 percentage point After the study of Fry, Balassa ( 1983) also affirmed the

right of financial liberalization model by empirical research for less-developing countries in

Deaton's (1992) view of empirical evidence IS that savings is not much influenced by interest rates Given the theoretical ambiguities, he did not find it surprising that clear results have not emerged from the analysis of the data, nor that different results can be obtained looking at different data sets in different circumstances

Bandiera, Caprio Jr, Honohan and Schiantarelli (1998) using principle components, constructed a twenty five-year time series index of financial liberalization for each of eight developing countries: Chile, Ghana, Indonesia, Korea, Malaysia, Mexico, Turkey and Zimbabwe They found that the pattern of effects differ across countries In summary, liberalization appears to have had a significant positive direct effect on saving in Ghana and Turkey, and negative effect on Korea and Mexico No clear effect is discernible in the other countries There is no evidence of significant, positive and sizeable interest rate effects There results indicate that there is no firm evidence that financial liberalization will increase saving Finally, they concluded the effect of financial liberalization on private saving is theoretically ambiguous, not only because the link between interest rate levels and saving is itself ambiguous, but also because financial liberalization is a multi-dimensional and phased process, sometimes involving reversals

From these evidences above, it is still ambiguous explanation for the relationship between the real rate of interest and the level of savings But one could expect that financial liberalization would improve the efficiency of investment by more appropriate resource

I allocation In other words, financial liberalization could increase financial savings and thus change the savings structure

2.3.4 FINANCIAL DEVELOPMENT

The relationship between financial development and saving ratio is still ongoing debate There is no general agreement on the direction of the impact In theory, financial development offers a wide range of financial instruments that help to bring higher returns on financial assets, lower risks Hence this creates incentive and more willingness to save However, higher returns on assets can harm savings mobilization owing to the income effect because higher returns allow people to increase consumption while they can get the same

I

targeted income, therefore reducing savings

According to Jansen ( 1990), some researches were done on the relationship between financial development and savings ratios Wai (1972) found a positive relationship in some countries or region but not at all Gupta (1984) estimated a saving function for nine Asian countries in which the level of financial development was one of the explanatory variables

In only four of these cases were the coefficient of this variable significant; in two cases it had a negative sign and in two cases a positive sign

In sum, the expected positive relationship between financial development (that proxied by M2/GDP) and savings ratios exists in some cases, but no general and clear conclusion emerged

2.3.5 THE STUDY OF VIET NAM INSTITUTE OF FINANCE

Basing on the Life cycle hypothesis and Financial liberalization approach with data from

'

The World Bank, Nguyen, C.N and Nguyen, H.H (2000) estimated the savings functions for thirty-five developing countries; The Philippines and VietNam in the period 1985 - 1995 They found that:

For thirty-five developing countries, the interest elasticities of savings are not statistically significant However, the positive impact of economic growth and the negative effect of the dependency are confirmed

The regression for the Philippines also reveals the positive impact of economic growth and the negative effect of the dependency on the levels of savings But the real interest rate is seen as constrained determinant Furthermore, if respectively test, rate of interest has no · relation with the level of savings in the specific model

And for the case of VietNam, they estimated the saving functions in the linear form with some main explanatory variables such as real interest rate, income per capita growth rate, foreign direct investment growth rate and M2/GDP The regression result presents all factors above are important variables and they explain 99.2% of saving actions in over years In those, the real interest rate has negative impact on saving rates; foreign direct investment and especially income per capita growth rate have positive effect on savings

In short, we spent some main theories that are used to explain the determinants of domestic savings So as to avoid the paradoxes that Keynesian savings theory can create because of basing on current income, the Permanent income and Life cycle hypotheses relate them to

29

permanent income and life-time income to explain savings behaviors However, exactly calculating permanent or life-time income in the case of developing countries seems to be out of reach The development of labor, capital and financial markets in these cou~tries have not fully satisfied the requirements of these theories above Hence, in order to solve the above limitations to some extents, some researches on savings in developing countries applied the McKinnon and Shaw Financial liberalization to explain the determinants of savings But in reality, unlike the theory, the increase of real interest rate as well as the development of financial system has unclear impacts on savings Therefore, a suitable approach is to integrate theories of savings together From this, savings is determined by ability to save (proxied by GOP and Dependency ratio); willingness to save (proxied by Real interest rate) and opportunity to save (proxied by Broad money/GOP) With this model, the empirical studies show the positive impact of GOP on domestic savings, the negative relationship between dependency ratio and domestic savings meanwhile ambiguous effects

of real interest rate and broad money on domestic savings

CHAPTER 3: DETERMINANTS OF DOMESTIC SAVINGS IN ASEAN

COUNTRIES: AN OVERVIEW

This chapter provides a brief review of some main factors found in ASEAN developing countries such as economic growth, demographics and financial liberalization that may have impacts on savings in order to understand the situation and trend of savings in ASEAN developing countries in general and especially the case of VietNam Before presenting these factors, the similarities in geography, culture and economic process will be mentioned

3.1 SIMILARITIES OF THE FIVE ASEAN COUNTRIES

The five ASEAN economies analyzed in this thesis are Thailand, Malaysia, Indonesia, the Philippines and VietNam These countries were selected not only to boost the model sample size so as to get a reliable econometric result but as these economies have similar characteristics to VietNam First, these countries have a geographic proximity to each other (Lim Chong Yah 2002, p.l4), they are all located in the region of South East Asia and to some extents, have the Chinese culture Second, the economic background of these countries

is also similar to VietNam's economic background They are all in the process of regional integration with greater overall openness of their economies and an ongoing liberalization of their finance, trade and investment regime (Lim Chong Yah, 2002)

3.2.1 ASEAN ECONOMIC GROWTH

For the past decades, ASEAN developing countries have spent the vigorous and dynamic period of economic development, gained the relatively important position in East Asia and

the world This area annually created over 750 billion dollars of goods and services (Pham et

rate in the world and has contributed adequately to the miracle of East Asia (Pham et all

Table 3.1 GDP Growth Rates in Selected ASEAN Developing Countries

86-90 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Thailand 11.5 8.56 8.08 8.38 8.95 9.31 5.88 -1.45 -10.77 4.22 4.31 Indonesia 7.45 8.93 7.22 7.25 7.54 8.39 7.64 4.69 -13.12 0.85 4.77 Malaysia 8.32 9.55 8.89 9.89 9.21 9.83 10 7.32 -7.36 6.08 8.3 Philippines 4.85 -0.58 0.34 2.12 4.39 4.68 5.85 5.19 -0.58 3.39 4.01 VietNam 5.34 5.96 8.65 8.07 8.84 9.54 9.34 8.15 5.8 4.8 5.5

Source: World Bank Indicators 2002

However, in the last years of 1990s, some countries faced serious financial crisis and leaded

to the socio-economic difficulties in all countries studies except VietNam, to some extents

In 1998, these countries had negative rates of economic growth

This financial crisis has destroyed ASEAN developing countries' economies and showed the instability and limitation on the way of development of these countries One of the lessons is that these countries depended a lot on foreign capital (Pham et all 2002, p.l6) Domestic savings did not play adequate part in the process of economic growth