Relationship between external debt and economic growth in selected developing countries

Besides, using export to service the debt of developing countries maymake a negative impact on economic growth through drawing the income from theservice activities.. All will lead to th

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY

VIETNAM

THE HAGUE THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

RELATIONSHIP BETWEEN EXTERNAL DEBT AND ECONOMIC GROWTH IN SELECTED

DEVELOPING COUNTRIES

BY

NGUYEN THANH THAI CHAN

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, December 2014

UNIVERSITY OF ECONOMICS

INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY

VIETNAM

THE HAGUE THE NETHERLANDS

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

RELATIONSHIP BETWEEN EXTERNAL DEBT AND ECONOMIC GROWTH IN SELECTED DEVELOPING COUNTRIES

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

NGUYEN THANH THAI CHAN

Academic Supervisor:

Ph.D Pham Khanh Nam

HO CHI MINH CITY, December, 2014

TABLE OF CONTENT

Chapter 1: INTRODUCTION

1.Problems statement

2.Research objectives

3.Scope of study

4.Structure of thesis

Chapter 2: LITERATURE REVIEWS

1.Theorectical reviews

1.1.External debt’s concept

1.2.Economic Growth Theories and Models

1.3.Theories and hypothesis on the relationship between external debt and economic growth

2.Empirical Literature Reviews

Chapter 3: RESEARCH METHODOLOGY

1.Overview of external debt and economic growth in developing countrieS

2.Analytical Framework

3.The Econometric Model

4.Data

4.1 Data Description

4.2 Summary table of variables and data source

5.Estimation Approach

Chapter 4: RESULTS

1.Descriptive Statistic Data

2.Econometric Results

3.Results Expression

3.1 Linear equation: Yi,t = αXi,t + γDi,t + ui,t

3.1.1 All coefficients are constant across time and individual

3.1.2 Using Fixed-effects Technique

1

3.1.3 Using Random-effects Technique

3.1.4 Choosing between Fixed –effects (FEM) and Random-effects Model (REM)

3.2 Quadratic equation: Yi,t = αXi,t + γDi,t + δD2i,t + ui,t

3.3 Tests for correcting the chosen model – FEM

3.4 Analyzing the estimation results and economic meanings of chosen model - FEM (eq3)

3.5 Discussion

Chapter 5: CONCLUSIONS

1.Conclusions

2.Policy Implications

3.Limitation of thesis

REFERNCES

APPENDIX

TABLES & FIGURES

Figure 1.1 Solow Model Production Function ……….8

Figure 1.2 Laffer Debt Curve ………13

Figure 1.3 Debt Threshold Curve 18

Table 3.1 External debt and GDP in main areas of developing countries …… 20

Figure 3.1 Extenal Debt and Economic Growth Framework ……… 24

Table 3.2 Summary of Descriptive Variables ………28

Table 4.1 Summary Statistics of Variables ……… 30

Table 4.2 Summary of Regresssion Result 32

APPENDIX EXTERNAL DEBT OVERVIEW 55-59 DETAILED POLICIES ON EXTERNAL DEBT ISSUE 60-64 TABLE 1: OLS RESULT ………65

TABLE 2: OLS WITH DUMMY VARIABLE ……… 66

TABLE 3: FIXED-EFFECTS MODEL RESULT ……… 67

TABLE 4: RANDOM-EFFECTS MODEL RESULT ……… 68

TABLE 5: HAUSMAN TEST ……….69

TABLE 6: FIXED-EFFECTS MODEL RESULT FOR EQUATION 2 ………70

TABLE 7: TESTING FOR MULTICOLLINEARITY ……… 71

TABLE 8: TESTING FOR HETEROSKEDASTICITY ……….72

3

Policy makers and economists have paid a great attention to investigate therelationship between external debt and economic growth to evaluate the impact ofexternal debt on economic growth and find out the reasons Under the presenteconomic circumstances, the more developing countries approach theirglobalization and interrogation, the higher risk of debt crisis they have to bear due

to the easy accession to the external sources of foreign loans Khan and Ul Haque(1985) considered this risk as an explosive one to emphasize its potential threat tothe whole economy Therefore the issue of external debt has always kept asignificant concern related to development economics and become the never oldtopic of various researches from the economists to policy makers

Any economy which experienced a fiscal deficit can finance the publicdeficit by borrowing domestically from a private sector through financialinstitutions or from other international sources For developing countries, due to

the lack of the strong private sectors and well-established banking system, theavailable source of domestic capital input are almost insufficient all the time Facingthe nonstop demands for investment and development, many developing countrieshave no way but borrowing extensively from international lenders and otherexternal sources This is also the reason why the relationship between external debtand economic growth has an outstanding correlation in developing countries andmany studies also choose developing countries as main object of research If thegovernment or policy makers do not have a correct and comprehensive evaluation

on this relationship as well as its impact, the issue of economic growth can be a bigpuzzle

Some countries believe in the positive impact of external debt on economicgrowth and easily get trap in the debt crisis As Amoateng and Amoako-Adu (1996)found that GDP growth shared a positive relationship with debt service Otherwise,the others are scare of debt burden and minimize the external loans but they cannotgenerate the economic growth due to the shortage of capital According to Maureen(2001), external debt has a negative impact on private investment and hinders theeconomic growth as a result Practically, the external debt issue which has beenwidely known for the debt overhang theory and crowding out effect can definitelyaffect the economic growth in some ways by imposing the threats and vulnerabilityfor the developing economies Specifically, external debt which is larger than theeconomy scale of borrowing countries can certainly lead to a risky capital deficit tohinder private investment channels, which can result in a retrograde economicgrowth rate Besides, using export to service the debt of developing countries maymake a negative impact on economic growth through drawing the income from theservice activities But it cannot be denied that external debt can provide the physicalcapital input that helps to boost the economic growth rate Therefore, finding thelinkage between external debt and economic growth can give best tool fordeveloping countries to adjust and boost their economic growth with the reasonablepolicies

2

2 Research objectives

The study aims to explore whether there is a relationship between theexternal debt and economic growth or not by using the panel data of selecteddeveloping countries in the fixed period Focusing on the tests to answer thequestion if there is a linkage, how the relationship can turn out Under the specificcollected data, this study aims to answer the questions if the relationship is positive

or negative This relationship is expected to be linear but with reference to previousempirical researches, this study also attempts to find out whether a nonlinearrelationship exists or not

Since developing countries are striving for sustainable economic growth,they likely need to deal with the problem of debt most of the time, especially thelevel of external debt There are many researches on the impact of the external debt

on economic growth of a specific country with time series or a group of countries inthe same area but there are few ones covering the range of many areas Moreover,many existing researches have focused on evaluating the impact of external debt oninvestment and saving levels rather than the analysis of the relationship between theexternal debt and economic growth itself

Although there are many studies on this topic with sample of differentdeveloping countries in previous period far back then, there are still few researchesfocusing on this issue at the recent time with detail This is the reason why thisstudy chooses to explore the nature of relationship and cover the sample of 2representative developing countries which are equally distributed in Africa; LatinAmerica and Asia as: South Africa, Nigeria; Mexico, Brazil; Vietnam and India forthe period of 20 years from 1990 to 2009

4 Structure of thesis

This thesis will consist of 5 chapters and each chapter will cover the

following content Chapter 1 gives the general introduction about the research topic

including the problem statement to explain the importance of external debt inrelationship with economic growth; the main research objectives to find out whatshould be studied in this relationship and the scope of study to limit the researchsample.

Then, chapter 2 which is named literature review presents two main parts Asthe first part, theoretical review covers the theories and hypothesis about relationshipbetween external debt and economic growth Some models or citation can be used toillustrate for the theories and the determinants The second part mainly focuses onempirical reviews on this relationship as well as the related determinants with thecitation and brief findings drawn from the previous researches

Next, chapter 3 aims at explaining the methodology of research A briefoverview of practical problem can be reminded, then the main part focuses on thebuilding of econometric model The data is then described in detail and the variables

in function can be clarified

Chapter 4 presents the results from regression with a descriptive statisticdataset Some discussions can be generated in the process of looking back andcomparing with the literature review

And finally, chapter 5 gives a conclusion for what has been found and raisessome policy implications for what to do when determining the relationship as well

as the impact of external debt on economic growth

4

Chapter 2 LITERATURE REVIEWS

Before entering the literature review focusing on relationship between theeconomic growth and external debt, a brief definition of external debt should beincluded to present a clear overview External debt which is also called as foreigndebt is known as the part of the total debt in a country that is owed to creditorsoutside the country The governments, corporations or private households are able tobecome the debtors under the circumstance In the other ways, “total external debt is

a debt owed to non-residents repayable in foreign currency, goods or services”(World Bank World Indicators Definition, 2000)

To sum up, total external debt is the sum of public, publicly guaranteed, andprivate nonguaranteed long-term debt, use of IMF credit, and short-term debt.Short-term debt includes all debt having an original maturity of one year or less andinterest in arrears on long-term debt Total external debt is the debt at any giventime, as total loans of liabilities at that time, not including debt service reserve,requires the borrower to pay the original debt with or without interest, in the future,and this debt is owed by residents with no residence in the country, according to thedefinition of eight international statistical analytical organizations of external debt,including Bank for International Settlements, the Commonwealth Secretariat, theEuropean Statistical Organization, International Monetary Fund, the Organizationfor Cooperation and Development Economics, Secretariat of the Paris Club,Conference on Trade and Development United Nations The external debt of thecountry's balance comprise of all current liabilities (excluding debt service reserve)

to pay the principals and interest in respect of external debt in the country.Besides,eExternal debt can also be defined by national loans for creditors whoreside outside the country (including the national debt by a non-resident in thatcountry holds) According to the Glossary of banking and Finance Peter Collin

Publish (1997), external debt is the debt of a country from another country includingthe debt on the domestic debt market, but creditors are non-residents in the domesticmarket.

Another issue which needs to be considered carefully is the debtsustainability This concept covers represents the level of debt which permits theborrowing country to finance its current and up-coming compulsory debt servicefully without restructuring or avoiding the accumulation of arrears but allowing areasonable level of economic growth According to the World Bank and IMF report

"a country can be said to achieve external debt sustainability if it can meet itscurrent and future external debt service obligations in full, without recourse to debtrescheduling or the accumulation of arrears and without compromising growth”.Then, analyzing the external debt sustainability under the circumstances of viewingthe behavior of economics variables and other determinants can shed a light on thecondition in which the level of debt and other indicators can be kept at a stable stateand determine the reasonable adjustments as well as the scale for making thepolicies

There are various indicators for the sustainability of external debt but

to current fiscal revenue ratio…and their respective service ratio are the mostgenerally applied ones In this study, the ratio of external debt on export and totaldebt service as the main indicator for debt variable

Finally, the more detailed issues which are related to the structure, theclassification and the general indicators of external debt can be discussed and found

in the Appendix Having a glance through these parts is also important because it

can help to approach the external debt more comprehensively and understand theproblem more easily as well as deeply

6

1.2 Economic Growth Theories and Models

First of all, the theory of growth should be taken into account and sometypical models should be employed to get a better view of the factors affecting theecomomic growth to find the answer for the question whether a relationshipbetween economic growth and external debt exists or not One of the best-knownmodels of growth which can give a general look on determinants is NeoclassicalGrowth model developed by Solow-Swan (1956) This model explained the longrun economic growth clearly by considering the basic inputs including the labor (L),the capital (K) and other factors Specifically, a linkage between economic growth

in productivity, commonly referred to as technological progress can be found

A quick glance through the Solow diagram will present a clear look into theserelationship:

Figure 1.1 Solow Model Production Function

The Solow diagram can be drawn using the two key equations of the Solow model in terms of output per worker and capital per worker These equations are:

The respective steady state quantity of capital per worker and steady state

quantity of output per worker can be demonstrated by:

k* = (s/n+d) 1/(1-a)

y* = (s/n+d)a/(1-a)

Where: k* = steady state quantity of capital per worker

y* = steady state quantity of output per worker

Looking the diagram of production function, the relationship between theeconomic growth and its determinants including the investment rate, populationgrowth rate and technology can be demonstrated fully Specifically, if an increase ininvestment rate in an economy with its steady value happens, the sy curve will shiftupward to s´y, which also makes the production function shift upward and result in

an economic growth as well as a higher steady value capital per worker Otherwise,

if population growth rate increases, the (n+d)k curve shifts upward to (n’+d)k Andthis will make a loss in steady state capital stock per worker value

8

Finally, this model also figures out a significant points that a sustainableeconomic growth will be only ensured when being accompanied by a progress intechnology which occurs when there is an increase in “A” which is a newcoefficient called “labor augmenting” or “Harrod –neutral” labeled by Solow in the

be only achieved if there are the sufficient sources of capital input and theproductivity of a labor unit All will lead to the undeniable correlation betweeneconomic growth and the accumulation of capital as an important input factor of

Therefore, in specific view, a definite relationship between economic growthand external debt can be expected due to the fact that not all countries, especiallythe developing ones can have enough the sound saving and investment rate, so thealternative capital input comes from the external sources will be inevitable Or inthe other ways, considering all the factors affecting economic growth, a linkbetween it and external debt can be found definitely but the fact that how thisrelationship would like to turn out has a different approach drawn from the variousexisting researches

1.3 Theories and hypothesis on the relationship between external debt and economic growth

There are various theories and hypothesis relating to the relationship betweenexternal debt and economic performance, which can be approached under thedifferent point of view It is hard to identify clearly the impact of external debt oneconomic growth since it will depend on the level of debt Positive relationshipbetween the external debt and economic growth will be determined if an acceptablelevel of debt can compensate the capital deficit and improve the social welfare.Whereas a negative relationship will be likely to happen if an excessive level ofexternal debt can affect economic growth through debt overhang and crowding outeffect

Firstly, some hypothesis and theoretical models have shown the support forthe positive relationship between external debt and economic growth at a reasonablelevel of debt to improve the social welfare and encourage the investment Thegeneral core point in those views can be drawn to the fact that economic growthwould be explained and booted by the accumulation of capital Typically, theHarrod-Domar model (1946) which was used widely in development economics put

an effort to show the direct linkage between economic growth and capitalaccumulation and indicated that if the debt can supply the capital accumulation, theeconomic growth will be targeted This model was developed initially by Sir RoyHarrod in 1939 and then, Evsey Domar continued developing it in 1946 to find theexplanation for economic growth rate in terms of the level of saving and

of generating the economic growth which can be explained through the quantity oflabor and capital input The more capital stock is accumulated, the more easily theeconomic growth target can be achieved The main function Y=f(K) implies thecapital is necessary for output This is the key issue to the developing countrieswhere the labor is abundant but the physical capital stock is always in condition ofshortage, so the economic growth has been slowed down

Dealing with this question, many developing countries have approached theexternal debt with the motivation of making the positive impact to finance thenational economic growth for a long time These external debts can definitely help

to generate the economic growth of that country in turn if they do not exceed theacceptable limit to become a debt burden According to the model of Growth, it cannot be denied that the external debt can definitely help the countries, especially thedeveloping ones, to fill the gap of capital shortage for growth as long as the externaldebt is controlled at a reasonable point

Consequently, a positive relationship between external debt and economicgrowth can be achieved at some level since the critical function of external debt is

to supply the adequate source of capital for the domestic demand in wide range of

10

investment and transformation to boost the economy as mentioned by Chenery andStrout (1966) Then a stimulated economic growth can bear a higher output because

it can finance the productivity of labor And this good performance can help acountry to service its own external debt as long as the level of debt is correspondent

to its pay-back capacity

On the other hand, some theories has supported the idea that a certain amount

of external debt can affect the economic growth negatively on through the debtoverhang effect on investment channel It can be said that “debt obverhang” theory

is one of the most well-known ones to demonstrate the adverse impact of externaldebt on growth According to (Myers, 1977), the term “debt overhang” wasoriginally stemed from the corporate finance literature which portrays a situation inwhich a firm’s debt is larger than any earnings generated by new investmentprojects, even projects with a positive net present value Therefore, both the currentstock of debt and the present value of a firm cannot be improved by any ways underthat circumstance as the beginning of “debt crisis” in microeconomic view Then theconcept of “debt overhang” was developed in the international finance literature bymany economists

According to Krugman’s (1988), debt overhang condition happens in onecountry when the face value of its debt is larger than the expected present value ofits future transfers The “debt overhang” theory is also characterized by Sachs(1989) as the excess of the expected repayment on external debt over the contractualvalue of debt This theory emphasizes the negative impact of external debt oneconomic growth through the channel of investment and the physical capitalaccumulation

Specifically, when a country, especially the developing one, bears a highlevel of external debt, it can easily discourage the domestic and foreign investmentfor physical capital accumulation, then slow down the economic growth as a whole

It can be easy to recognize the risk with harmful effect when a large accumulation

of physical capital coming from the external sources exceeds the ability of one

county to repay both its past and current debt In turn, the investment activities have

to bear many losses when potential lenders and investors will be steered away due

to the uncertainties and instabilities of repayment Even if not in the present timebut in the probability of future, can the estimated level of debt unserviceable alsoimpose the similar negative impact on the economic performance through the level

of output Then, the returns from the domestic investments are "taxed away" byexisting foreign creditors to service the debt and this will discourage the incentives

to invest in both domestic and foreign sectors and affect negatively on economicgrowth, according to Claessens et al(1996)

The impact of debt overhang to illustrate for the relationship betweenexternal debt and economic performance can be tracked through the “crowding out”effects It means that the increase in external debt service will strike down the level

of investment in turn The main measures of external indebtedness are the indicatorsover GDP and exports, so “crowding out effect” emphasizes the negative impact ofthe fact that the income from export has to be taken away to repay the external debt

As a result of this, the lack of capital for stimulating the investment activities toescort the economic growth is inevitable According to Hansen (2004), the debtpayment service can become the severe restriction to economic growth by such thatcrowding out effect

Then Cohen (1993) found the empirical evidence to oppose the “debtoverhang” hypothesis and support the “crowding out” effect Besides, when acounty is in the debt overhang condition, the government can take some actionswhich may depress the investors like the distortionary measures including the fiscaland restructural reforms to finance the debt –service obligations These changes inthe policy of government can set a barrier or build the hesitation to hinder theinvesting activities and result in a loss of output as well as a negative economicgrowth rate at last This effect can be demonstrated by the the Debt Laffer curvewhich was introduced by Sachs (1998) through the theory of “debt overhang” asbelow:

12

do not have any incentives to work more when they have to bear a larger tax for that

at all Similarly, the Debt Laffer Curve demonstrates clearly the situation when ahigh level of debt is accompanied with a loss in the capability of expectedrepayment because of the discouragement effect and sabotage impact

Specifically, it means that when the external debt stock rises, the indebtedcountry is discouraged by the lack of investment from foreign investors and has noincentives to produce more due to the fact that the output will be used to service theexisting external debt Thus, a loss in economic growth is inevitable and clear.Looking at the Debt Laffer Curve, if a country is on along its left side, an increase indebt stock can be accompanied with an increase in repayment capability but afterreaching a critical point called peak, any increases in debt stock means a decrease inexpected repayment However, it is not easy to figure out that critical point and itcan be different depending on each countries Any countries on the right side of thatpoint have to bear the condition “debt overhang” with negative impacts on theeconomic growth rate

2 Empirical Literature Reviews

Besides a number of theoretical literature on this issue, many empiricalstudies has been done to find out the relationship between external debt andeconomic growth There are various existing results but they can be divided intothree groups The first group belongs to the empirical evidence that has supportedthe positive effect of external debt on economic growth at the certain level Thesecond group relates to those which have indicated the negative relationship of highlevel of external debt with economic growth; then the third one combines two aboveeffects to argue that there is a nonlinear relationship between external debt andeconomic growth by nature Besides, there are some empirical studies have found

no relationship between external debt and growth

Firstly, a few studies have found the positive relationship between theexternal and economic growth at the certain level Chowdhury (1994) attempted toshed a light in the debate of determining the causes and relationship betweenexternal debt and economic growth, by conducting granger causality tests for Asianand Pacific Countries over a period of 1970-88 With the dependent variable isGDP, a set of independent variables including debt payment (negative), inflation(negative), interest rate (positive), agricultural labor (negative), it has been foundthat any increase in GDP is followed by a higher level of external debt and theoverall external debt does not have any negative impact on economic growth

A simple neo-classical model was employed to find out the relationshipbetween external debt and economic growth and the capability of external debtservicing with the cross sectional data of 31 Sub-Saharan African countries byGerald Scott(1994) This study used the dependent variables of GNP per capita inlog and a set of independent variables including exports (positive), domestic Capital(positive), technology (negative), imports (positive), exchange rate (positive) It hasbeen concluded that there is positive evidence of the impact of external debt ongrowth at the certain level Employing data from 59 developing and 24 industrial

14

countries over a period of 1970-2002, Schclarek (2004) found no proof of the factthat external debt can affect total factor productivity (TFP) but got the evidence that

a low level of external can accompany with a favorable rate of economic growth.This also hinted a positive relationship between low debt and economic growth

Secondly, the empirical studies which favor the negative relationshipbetween external debt and economic growth account for substantial number andbecome the trendy and attentive hardcore on this topic This study will also focus onthis direction in the process of find out how this relationship can be For instance,Iyoha (1999) used a small macro-econometric model with the simulation approachwith both Growth equation and Investment Equation to look into the impact ofexternal debt on economic growth in sub-Saharan African countries in the period1970-1994 A negative relationship came out the light with the consistent hypothesiswith the debt overhang theory and crowding out effect

Besides, Sachs (1984, 1988), Cohen and Sachs (1986), and Krugman (1988),Adepoju et al (2007) studied that relationship in the context of debt crisis in 1980sand employed the debt overhang theory models to find out the negative impact ofexternal debt on economic growth rate For the case of Nigeria over a period of

1962 to 2006, by using the simultaneous model with time series data, Adepoju(2007) jumped to the conclusion that the accumulation of external debt really heldthe economic growth back for times in Nigeria Another case in Pakistan, Hameed

et al.(2008) investigated the dynamic effect of external debt servicing, capital stockand labor force on the economic growth for a period of 1970-2003 and found anadverse relationship of external debt and economic growth through the channel of

variable of GDP growth rate and a set of independent variables including inflation(negative), lagged inflation (positive), real public investment as a ratio of GDP(positive), private investment (positive), fiscal deficit/GDP (negative), HumanCapital Development (positive), Debt/GDP(negative), Debt Service to export

(positive) with time series data of Keyna A negative relationship between externaldebt and economic growth was found as final result.

Then, the panel data of 14 Pacific Island Countries from 1988–2004 wasanalyzed by T.K Jayaraman, Evan Lau (2008) to do tests for finding a linkagebetween external debt and economic growth rate Real GDP was set as dependentvariable with a set of independent variables as follows: External Debt (negative),Exports (positive), Budget Deficit (negative) And one of the most popularempirical study, "What Are the Channels Through Which External Debt AffectsGrowth?", Pattillo(2004) used the data of a group of 61 developing countries in theperiod of 1969–98 to build a growth-accounting framework and found that theaverage level of external debt slow down the economic growth by holding back thephysical capital accumulation and the growth of total factor productivity (TFP)

Thirdly, many existing empirical researches which support the nonlinearrelationship between external and economic growth have caught the attention moreand more Typically, by using fixed and random effects panel data to estimate theregression model, Elbadawi et al (1997) found out the nonlinear impact of externaldebt on economic growth The relationship was characterized and consistent withthe debt Laffer curve with a critical threshold This research showed both in linearand quadratic form of debt-to-GDP ratio in the regressions and found the growthmaximizing debt to GDP ratio of 97 percent

Then, Patillo et al (2002, 2011) used many methods of estimation includingOLS, Fixed effects, and GMM system to come to conclude a nonlinear relationshipbetween external debt and economic growth with the panel data of 93 developingcountries Patillo et al (2004) continue to research in detail about the channelsthrough which external debt can make a non-linear impact on growth in developingcountries, and found that the physical capital accumulation accounted for one-third

of the negative effect of debt on growth and two-thirds were stemmed from totalfactor productivity growth (TFP) This study of 93 developing countries in theperiod of 1969-1998 found that the negative impact of external debt on growth

16

happened when its net present value (NPV) exceeded 160–170 percent of exports

Reinhart and Rogoff’s (2010) set the first step in orientating the researchtendency in estimating the debt thresholds and threshold effects of debt on growth.Both linear and non-linear effects of external debt on growth were investigatedthrough the model specifications by employing different econometric techniques.The linear estimations directly indicated a negative relationship between debt andinitial growth, while the nonlinear estimations found a threshold of 90% debt-to-GDP ratio at that level external debt to make a significant negative impact ongrowth Therefore, unlike a low and reasonable limit, a high external debt is likely

to have the definite negative impact on the economic growth Or in the other ways,there is a threshold of level debt at which the external debt can impose the twoopposite effects on the economic performance of one country This implies that arelationship between external debt and economic growth also has an inverted Ushape with a critical point called the threshold beyond which the initial positiveimpact of external debt on economic growth will turn out the negative onecompletely That is beginning from zero up to the thresh hold, any increases inexternal debt or its net present value can make the expected repayment reliable, thenincrease its contribution to economic growth through accumulating the physicalcapital However, after reaching the threshold, any further increases in external debtonly results in the risk of debt unserviceableness In other words, the negativeeffects of debt overhang theory occur only after a certain threshold level has beenreached The debt threshold curve can be inspired by the shape of Debt Laffer Curvelike this:

Figure 1.3 Debt Threshold Curve

Source: Patillo et al (2004)

Looking at the debt threshold curve, A is determined as debt threshold andkeeps role to divide the impact of external debt on economic growth into twodistinguished ones On the left side of A, external debt bears a positive relationshipwith economic growth or in other words the marginal productivity of each availableexternal debt is greater than or equal with the principal and the interest payment,according to Cline (1985) However, on the right side of A, one country has to sufferthe negative influence of external debt on economic growth due to that fact that thepayment of interest and repayment of principal can discourage the privateinvestment or make the changes in public expenditure Higher external interestpayments can increase a country's budget deficit, thereby reducing public savings.And when reaching the point B of net present value (NPV) of external debt, itscontribution to economic growth not only has downward trend but also keeps anegative value

It can be seen that there are three different flows of main ideas onrelationship between external debt and economic growth which has been based onvarious empirical studies However, there are still some researches on external debt

18

and growth which cannot find any relationship Focusing on one of the HeavilyIndebted Poor Countries (HIPC), Kenya, Were (2001) focused on analyzing the debtoverhang and tried to find out any relationship between external debt and thepresent low economic growth rate Using time series data from 1970-1995, thisstudy could not find any evidence to favor for the negative or positive impact ofdebt servicing on economic growth; instead of it, some crowding-out effects onprivate investment were confirmed While employing 13 developing countries datafor a period of 1960-1981 and 1982-1989, Warner (1992) could also not find anyempirical relationship between external and debt servicing on economic growth aswell.

Chapter 3 RESEARCH METHODOLOGY

Table 3.1: External debt and GDP in main areas of developing countries

Unit: Billion USD

Area

Pacific and East Asian countries

European and Middle Asian countries

Latin American and Caribean countries Middle, East and North African countries South Asian countries South Sahara African countries

Source: World Bank and IMF Indicators and Data (2010)

Recently, the issue of external debt has risen globally at high speed

Developing countries in three areas: Asia, Latin –America and Africa are facingwith this problem and struggling for the target of achieving the sustainableeconomic growth at the same time This has pushed these countries into a dilemma

20

which they cannot find the sufficient balance in the relationship of debt and growth.Specifically, in Asian countries, external debt had a historic sharp rise back far in the1990s due to the motivation of boosting the economic growth from foreign loans In

1991, the ratio of debt to GDP remained approximately 19% and continued to rise to

a peak of 35% only after 7 years (World Bank Statistic Yearbook, 2010) When USFederal Reserve tightened their interest rates in 1994 and drawn the money awayfrom Asian developing countries, the governments had to prohibit their currencybegs to dollar due to the lack of foreign reserves which leaded to in capability ofmaintaining the fixed exchange rate This thing brings up the problem of currencydevaluation and debt service obligations

Under the circumstance of exposing to the external debt shock, developingcountries in Asia tried to limit the ratio of external debt to GDP by 15% in 2008.However, the global financial crisis in 2008 broke out and caused a big hole ofcapital mislocation The general economic growth rate slows down and manyeconomies fell into the stagflation This forced these countries found the externalcapital aids to compensate the demand and target of recovering the nationaleconomic performance The excessive loans from external sources which most ofdeveloping countries in Asia considered as the main momentum of generatingeconomic growth now totally turn out a debt burden when the existing external debt

is larger than the expected repayment capability Serving the external debt can be adifficult problem and become a crisis if the relationship as well as the impact ofexternal debt on economic growth still remains vague In that stage, almost all of thedeveloping countries have to bear the risk of exposing to the short term externaldebt, including Vietnam

In 2011, Vietnam has a high level of short-term external debt which implied arisky exposure to the external shocks It is the same as in India or Indonesia.However the debt crisis did not happen because the external debt is not larger thanthe acceptable level respectively to the foreign reserves in one country but the riskhas already existed Besides, the economic growth in those typical countries

21

experienced a downward trend as well as some difficulties The Asian developingcountries still have the troubles in dealing with the external debt and evaluating itsimpact on economic performance Therefore, to any policymakers or economists,the relationship between external debt and economic growth needs considering andput into the right place.

Next, the external debt issue can also be found in developing countries ofLatin-America Debt crisis has not been a strange theme in the history of these

countries by far The debt crisis in 1980s really broke out in Latin America with the

Mexico’s announcement of incapability of servicing its debt It marked one of thedarkest time to the economic growth rate in specifically and the stability of thewhole country in generally Most of developing countries in this time faced a largeexternal debt because of not being aware of the relationship between the debt and

economic growth to have an accurate evaluation of its impact After that, a lot of

countries underwent the restructuring and defaulting the debt such as Ecuador in

1999, Arghentia in 2002…to adjust the economic performance in close relationshipwith debt Governments of key economies like Brazil, Mexico, etc have paid a greatattention to external debt issue and tied the fiscal balance to this problem tightly Arelationship between external debt and economic growth has been considered andgained so much attention from policy makers

Similarly, in Africa, developing countries have also struggled for the issue ofexternal debt for a long time Due to the frequent lack of physical capital input,those countries have to borrow from external sources but they have a limitedcapability of repayment Inevitably, a debt overhang situation occurs and economic

growth is tied at low level Specifically, in 1983, Africa’s total external debt over

Africa. And at the peak of debt crisis in 1990s, African countries had to use more

than 40% of its export earnings to service its external debt (IDA and IMF, 2009).

Especially, the developing countries in sub-Saharan Africa rely heavily on foreignborrowings as an official part of development strategy Therefore, the external debt

has been accumulated year by year without being aware of reaching theunsustainable level For example, from 18 billion dollar in 1975, external debt rose

to over 200 billion dollar in 1995, which showed an extensive buildup of debt InSouth Africa, the ratio of external debt to export was above 300 percent in terms ofthe average ratio of 200 percent for the rest of countries in Africa (World Bankreport, 1975-1995) Thus, the economic growth in African developing countriesalmost fast nonstop at same time and still keeps at low rate due to the condition ofdebt overhang and crowding out effect

Generally speaking, external debt and economic growth has been proved to

be linked together definitely The negative relationship with the retrograde impactcan be expected when developing countries are in condition of debt overhang withhigh level The only incentive for borrowing external loans is motivated by theinitial compensation for the shortage of physical capital input

The output which represents for the economic performance of one countrydepends on various factors but can be written under the function like this: Y =Af(K,L), according to Cobb-Douglas Then, each factor K and L also has its owndeterminant and can be written under a similar function form like that: K= H(Debt,

that there is a linkage between the external debt and economic growth:

23

framework to understand the reason why the certain variables are selected to run theregression All the determinants have shared the link with economic growth andexternal debt can be found in the whole picture This simple framework can help tobuild a flowchart of ideas before constructing the econometric model :

Investment

Figure 3.2 Extenal Debt and Economic Growth Framework

(Ex: Export; Im: Import)

the whole

external debt can be demonstrated simply From the function Cobb –Douglas and

which can be measured by the schooling years of a certain object and labor is theinputs of human capital Besides these two main factors, economic growth is alsoinfluenced by the other determinants Savings, Import and Export are definitelypositive inputs of economic growth Then external shock which can be clssified asinflation is also another factor to be paid attention because economic growth wil

important role onorientating economy and correcting the market to achieve the proposed target Then

24

external debt appears in the whole cycle of economic growth and its determinantbecause it has direct link with capital issue and external debt service can make animpact on economic growth due to its burden of repaymnet To finance the externaldebt, one country usually its returns on export to pay for the interest of external debt

or service it Therfore, it can also be concluded that external debt issue of onecountry, especially the developing ones has a close linkage with export The impact

of debt service is expected to be negative but the relationship between external debtand economic growth is still a question An econometric will be built based on theideas from this framework to serve the object of finding the nature of external debtissue

The model adopted in this study will be based on Pattilo et al (2002) modelspecification of study on the relationship between external debt and economicgrowth and Barro Growth Model (2003) on study of economic growth’sdeterminants The basic regression equation which is used to uncover therelationship between debt and economic growth has the form as following:

Y i,t = αX i,t + γD i,t + u i,t (eq1)

Where :Y i,t is the dependent variable which is GDP per capita or LogGDP (inlogged term) to represent for economic growth rate

Then, D i,t is the debt variable includes

capabiltity of servicing debt

Finally, X i,t represents the set of explanatory variables, including tthe set ofdeterminants of growth (capital, human capital, macroeconomic environment andfiscal gap)

input;

 Initial GDP: Log of GDP per capita in year t-1 which represents for thebenchmark value of economic growth is based on Growth Model of Barro(2003)

for human capital;

and export on GDP

 Expenditure (government’s expenditure): covers the effects of the fiscalpolicy due the government’s expenditure can affect the government’s fiscalbudget

+I + G + NX (net export)

Equation (1) is linear in nature and is used to test the linear relationship.However, we are also interested in finding out whether any nonlinear relationshipbetween debt and economic growth exists or not Then, a quadratic specification isalso considered to determine the marginal point at which the impact of external debt

on economic growth can be considered as:

Y i,t = αX i,t + γD i,t + δD 2 i,t + u i,t (eq2)

Specifically, we focus on investigating if there is an inverted-U shaperelationship between debt and growth or not Thus, in order to determine the non-

linear impact of external debt on economic growth, the coefficient δ is expected to

have a negative sign and the point at which external debt starts to give a negative

impact will be calculate mathematically as: -γ/2δ The negative sign of δ coefficient

is the prerequisite condition to have the graph of quadratic function like the Laffer

Debt Curve discussed in the literature review If the sign of δ coefficient is positive,

then it means that the nonlinear relationship with the shape like Debt Curve does notexist in the range of data and study for research At that time, the oppositerelationship with U shape between external debt and economic growth appears but

26

this result is contrary to what literature reviews and theoretical reviews have found.Then, the linear relationship will be the main one to focus and explain further more

Data for each country on the previous mentioned variables in the period fromthe year 1990 to 2009 Therefore, there are 6 cross-sectional units and 20 timeperiods The dependent variable used in this study is GDP which represents theeconomic growth Then a set of independent variables are established base on therelevance of affecting factors including the determinants of growth (investment,human capital, policy, macroeconomic factor) and external debt indicators.Specifically, investment to GDP is chosen to present for investment factor; thenpopulation and education which is estimated by the schooling year are chosen asvariables for human capital GDP per capita in year t-1 is chosen to represent forExpenditure which is calculated as the total can be the proxy variable for policystability Openness index and inflation are variables covering for trade andmacroeconomic fluctuation which can impact on the sustainable growth rate

Finally, external debt variables included the ratio of external debt to export and total

external debt service The specific description and calculation as well as the data

sources can be summarized as the following table below:

Table 3.2: Summary of Descriptive Variables

Dependent variable

LogGDP

Independent variable with expected sign

28

It can be seen that apart the debt indicators with the unknown sign, the aboveset of control variables can be grouped as: investment (+), income (+), populationgrowth rate (-), schooling (+) with the expected sign in terms of reflecting theimpact of physical and human capital accumulation on economic growth as TotalFactor Productivity factors They are expected to be consistent with the contentrepresented in the literature review Then, the terms of trade, openness and inflationcovers the external shocks on international market Finally, expenditure(government’s expenditure) is representative for the impact of fiscal policy on thestability of the economy.

There are some popular estimation techniques for panel data as: OLS, Fixed

and Random effect, and GMM system Function of (eq1) depends on the

assumptions about the intercepts, the slope coefficients, and the error term.There are several cases:

OLS regression to estimate the function of GDP)

known as one –way fixed effects model)

from others over countries and time (is known as two –way fixed effects model)

known as random fixed effects model)

(is known as one –way random effects model)

Therefore, in the range of this study, OLS method and FIX/RANDOMmodel will be tried in turn to find out the most suitable one Some tests can also

be applied to fix the model and get the best result at last as well Firstly, thelinear equation is used to test for the most appropriate model, then the quadraticwill be applied

29

Chapter 4 RESULTS

The complete dataset comprises of 120 observations over the period of

1990-2009 from 6 countries including: Brazil, Mexico, South Africa, Nigeria, India andVietnam The ten variables can be summarized statistically in a table like below:

Table 4.1 Summary Statistics of Variables

Variable LogCGDP LogExternal Debt Initial GDP Investment Inflation Schooling Population Openess Expenditure Savings Total External Debt

Generally speaking, panel data can allow the control for variables which we

30

heterogeneity Besides, by combining time series of cross sectional observations, thepanel data like this can give “more informative data, more variability, lesscollinearity among variables, more degrees of freedom and more efficiency”( Gujarati, 2003).

The results from all regression including OLS, fixed effects and randomeffects method for both linear and quadratic equation can be summarized in onetable for convenient monitor The linear equation is tried with OLS (with andwithout dummy variable); Fixed effect and Random effects method, then a huasmantest is employed to decide whether Fixed Effects Model (FEM) or Random EffectsModel (REM) is more suitable After testing, FEM is chosen as the most suitableone Therefore, fixed effects is applied to the quadratic equation Then, the test formulticiollinearity and heteroskedascity is also carried out to test the chosen model

After regressing many function with different methods, a table of result issummarized to make the monitor and comparison more easily The coefficient andstandard error which is demonstrated in the bracket is recorded in the below table.Further details of regression results can be found clearly in the tables in theAppendix After the viewing the table of summary for regression results, thespecific analysis and presentation will be disussed in details