The effectiveness of fiscal policy contributions from institutions and external debts

University of Economics Ho Chi Minh City
Journal of Economic Development
**********

The Project:
PROMOTING THE ENHANCE QUALITY OF
THE ENGLISH-LANGUAGE VERSION OF JOURNAL OF ECONOMIC DEVELOPMENT,
INTENDED AS A SCOPUS-LISTED JOURNAL

Topic:

THE EFFECTIVENESS OF FISCAL POLICY: CONTRIBUTIONS FROM INSTITUTIONS
AND EXTERNAL DEBTS

CS- 2017-01TĐ-03


Contents
CHAPTER 1: INTRODUCTION .............................................................................................3
1.1.

Abstract................................................................................................................................................ 3

1.2.

Motivations .......................................................................................................................................... 3

1.3.

Methodology ........................................................................................................................................ 5

CHAPTER 2: LITERATURE REVIEWS ...............................................................................7
2.1.

The effectiveness of fiscal policy .......................................................................................................... 7

2.1.1.

The theoretical framework of fiscal policy’s effectiveness............................................................. 7

2.1.2.

The empirical works ..................................................................................................................... 9

2.2.

The institutions and the effectiveness of fiscal policy........................................................................ 13

2.3.

The external debt burden and the effectiveness of fiscal policy ........................................................ 17

CHAPTER 3: METHODOLOGY AND DATA.....................................................................21
3.1.

Methodology ...................................................................................................................................... 21

3.2.

Data .................................................................................................................................................... 25

CHAPTER 4: RESULTS AND DISCUSSIONS ....................................................................29
4.1. The institutions and the effectiveness of fiscal policy .............................................................................. 29
4.2. The external debt and the effectiveness of fiscal policy ........................................................................... 32
4.3. Robustness check ..................................................................................................................................... 34

CHAPTER 5: CONCOLUSIONS ..........................................................................................39
Acknowledgement ...................................................................................................................41
References................................................................................................................................42


CHAPTER 1: INTRODUCTION

1.1. Abstract
The effectiveness of fiscal policy is an interesting field in literature of macroeconomics. In
this paper, we use panel data from 2002 to 2014 from 20 emerging markets to investigate the
effects of fiscal policy on economic growth under contributions from the differences in
institutions and external debt levels. By using GMM estimators for unbalanced panel data, our
results show positive effects of fiscal policy on economic growth across emerging markets in the
examined periods. Notably, the improvement in institutions promotes higher crowding-in effects
of fiscal policy. In addition, this paper finds interesting evidences that the external debt has nonlinear effects on economic growth, whereas the heterogeneous effects of fiscal policy on
economic growth as positive effects in low indebted level and negative effect in high indebted
level may explain the mechanism of this non-linear relationship. The results have significant
contributions to the literature and useful implications for authorizers in promoting sustainability
of the economy. The authorizers are strongly recommended to focus on improving the
institutional quality that not only boosts the effectiveness of fiscal policy in general, but also
solves the dilemma of high indebted countries when the fiscal policy loses the effectiveness.

1.2. Motivations
Fiscal policy is conducted by government through taxation and public spending with the aims
at sustainable development for the economy. So, fiscal policy and its impacts on the economic

growth tend to be at the center of macroeconomic and political debates. The field of the
effectiveness of fiscal policy has re-highlighted in light of the 2008 global financial crisis with
the new contemporary drivers such as external debt (Ruščáková & Semančíková, 2016). Since
the complexity of the process by which fiscal policy is conducted is not fully captured, that why
different theories provide different answers regarding macroeconomic effects of fiscal policy and
arguments about the suitability and real effects of government expenditures on economic growth
are still interesting field of study (Bouakez, Chihi, & Normandin, 2014). Whereas, the main
question in the literature of the fiscal policy’s effectiveness is that whether fiscal policy presents
crowding-out and/or crowding-in effects in a country and what its drivers. In fact, many


researchers try to find evidences with the parallel existence of both and mixed conclusions (see
Ahmed and Miller (2000), Heutel (2014), Şen and Kaya (2014)).
The studies of the effectiveness of fiscal policy have developed and conducted in long history
through many economic growth models. Many studies use versions of the Solow (1956) model
to study the dynamic effects of taxation on economic growth, while other studies use neoclassical growth model (Easterly & Rebelo, 1993). In this regard, researchers argue that the
effects of government expenditures on economic growth follow two different regimes including
crowding-out effects and crowding-in effects. The neo-classical theory states that government
expenditure crowds out private investment then has negative impacts on economic growth.
While, Keynesian view, in contrast, states that government expenditure stimulates private
investment in the case of un-fully employment, which then has positive impacts on economic
growth, especially in developing countries (Ahmed & Miller, 2000).
Moreover, the effects of fiscal policy on economic growth is driven by many factors such as
the employment in the economy, the transparency of government, the composition of
government expenditures, or even the government size (see Akanbi (2013), Arestis (2011),
Kasselaki and Tagkalakis (2016), Hemming, Kell, and Mahfouz (2002)). In empirical literature
about the determinants of fiscal policy’s effectiveness, there are, in fact, some studies that
consider the role of institutional framework such as corruption situation, economic freedom,
democracy (see Baldacci, Hillman, and Kojo (2004), Martinez-Vazquez, Boex, and Arze del
Granado (2007), Nelson and Singh (1998)).

Meanwhile, the burdens of external debt on the sustainability of fiscal policy are also
concerned. For instance, Amato and Tronzano (2000) document that a restrictive fiscal stance, a
lengthening of average debt maturity and an increase in the share of foreign-denominated debt
are crucial to stabilize the exchange rate in Italia. Bal and Rath (2014) find that central
government debt, total factor productivity (TFP) growth, and debt-services are affecting the
economic growth of India in the short-run and they recommend that Indian government should
follow the objective of inter-generational equity in fiscal management over the long term in order
to stabilize debt-GDP ratio. Which means that the external debt may influence the effectiveness
of fiscal policy. Recent study, Doğan and Bilgili (2014) examine the relationship between
external indebtedness and growth variables in Turkey. They find that external borrowing has
negative impact on growth both in regime at zero and regime at one, but the public debt has


higher negative effects on economic growth and development. Precisely, they conclude that that
the economic development and borrowing variables do not follow a linear path.

1.3. Research questions
In fact, there are very early studies about the effects fiscal policy such as Smith (1937),
Bailey (1971), Buiter (1977), and Arestis (1979), and many recent studies try to investigate the
impacts of both government expenditures on private investment and especially economic growth.
However, the debate with regard to the effectiveness of fiscal policy is still ongoing (Bouakez et
al., 2014; Heutel, 2014; Kameda, 2014a; Şen & Kaya, 2014). Precisely, the literature of fiscal
policy is lacking of the studies about the effectiveness of fiscal policy under the contributions
from the institutions and external debts in a comprehensive work. Therefore, this study is
conducted under the motivations from the study of Doğan and Bilgili (2014) by investigating the
effectiveness of fiscal policy on economic growth under the relationships with the changes in the
institutions and the burdens of external debt in the context of 20 emerging markets including
Argentina, Bangladesh, Brazil, Bulgaria, China, Colombia, Egypt, India, Indonesia, Malaysia,
Mexico, Pakistan, Peru, Philippines, Romania, Russia, South Africa, Thailand, Turkey, and
Vietnam.

In which, this study goes to answer two main questions:
Question 1: do the institutions enhance the effectiveness of fiscal policy in emerging market
economies?
Question 2: does fiscal policy have heterogeneous effects under the difference levels of
external debt?

1.4. Methodology
In this study, we achieve our objectives by implementing following strategy. We firstly
examine the impacts of fiscal policy on economic growth through the modified model of
endogenous growth theory by incorporating government expenditure and controlling other
common drivers of economic growth including capital, labor, financial development, technology,
economic openness (trade and capital flows). Then, the institutional factors including
government effectiveness, regulatory quality, and control of corruption are incorporated,


respectively, to test the impacts of institutions on economic growth. Next, we use the interaction
terms between government expenditure and institutions to examine the effectiveness of fiscal
policy under the associations of institutional framework. We then estimate the growth model
with the explanatory variables including both external debt level to GNI and its square to
examine the non-linear relationship between external debt and economic growth. After that, we
divide our data into two sub-samples (the low indebted countries and high indebted countries) to
investigate the effectiveness of fiscal policy under two regimes. At last, we use GDP per capita
growth rate in replacing GDP growth rate to check robustness of results.
By doing this strategy, we believe that this study has significant contributions to both theory
and practice. Firstly, this study has contribution to the literature of fiscal policy effectiveness and
fiscal indebtedness by adding the effects of government expenditures under the external debt
level and the associations with institutional quality. The results find significant evidences that the
institutions enhance the effectiveness of fiscal policy. Notable, the external debt level presents
the non-linear relationship with economic growth through the mechanism that the fiscal policy
has the heterogeneous effects on economic growth: the crowding-in effect in low indebted level

and crowding-out effects in high indebted one. Secondly, this study has significant implications
for the authorizers in implementing the long-term sustainable fiscal policy in line with borrowing
policy and the solutions for the high indebted countries that face to the dilemma of ineffective
fiscal policy.
This study is structured as following. Chapter 1 focuses on our motivations of this study.
Chapter 2 presents literature reviews and then our arguments on the effectiveness of fiscal policy
under the contributions from institutions and external debt. Methodology and data are provided
in Chapter 3. Chapter 4 presents the results and our discussions. The concluding remarks are
discussed in Chapter 5.


CHAPTER 2: LITERATURE REVIEWS
2.1. The effectiveness of fiscal policy
2.1.1. The theoretical framework of fiscal policy’s effectiveness
In the literature of fiscal policy effectiveness, it is natural place to start with the Keynesian
theory. In Keynesian model, the sticky price and excess capacity are assumed that contraries to
the classical economics, so that aggregate demand determines output and government
expenditures have a multiplier effect on aggregate demand and output (Coddington, 1976).
Therefore, Keynesian economics call for the government intervention and incorporate
government expenditure into the aggregate demand function. The Keynesian views argue that
there is very rare case for an fully employed economy, thus the sensitivity of investment to
interest rates would be low and then an increase in interest rates due to expansionary fiscal policy
would be minimal, the government expenditure, in turn, has positive impacts on economic
growth (O’Hara, 2011; Şen & Kaya, 2014). This view is also called as the crowding-in effects of
fiscal policy, where the government should undertake the expenditure in the recession time to
cover the lack of private consumption and investment (Jahan, Mahmud, & Papageorgiou, 2014).
However, some of extensions in the line of Keynesian model allow for crowding-out effects
of fiscal policy, which means the expansion of government expenditure crowds out the private
demand and then influences negatively on output, through the changes in interest rates and
exchange rate in the case of open economy. With the assumption that the private investment is

negative impacted by the increase in interest rate, the expansionary fiscal policy that backed by
borrowing leads to the lower private investment due to higher interest rates. Moreover, the higher
interest rates due to the expansionary fiscal policy attract capital flows in the case of open
economy that appreciate exchange rate and then results the deterioration in current account (see
Mundell (1963), Fleming (1962)).
The neo-classical economics address the shortcomings of Keynesian economics on its lack of
microeconomic foundations. The neo-classical views focus on the determination of goods,
outputs, and income distributions in markets through both supply and demand sides by adding
the assumption of utility maximization of income-constrained individuals and firms under the
boundary of factors in production and available information (see Gaffney (1994), Goodland and
Ledec (1987), Davis (2006)). In which, the neo-classical economics raise the rational


expectations in comparing to the adaptive expectations in Keynesian economics. This brings
forward adjustments in economic factors that occur more progressively so that fiscal policy
matters in not only long-term but also short-term period. And the permanent fiscal changes can
lead to the crowding-out effects since private sectors expect the persistent changes in interest
rates and exchange rates in this case (see Buiter (1977), Arestis (1979), Mundell (1963), Fleming
(1962)).
In addition to neo-classical economics, the Ricardian view that is based on Ricardian
equivalence theorem assumes that the individuals are forward-looking in the current activities,
which is also in contrasting with the Keynesian economics view as individuals rely on current
income (see Barro (1988), McCallum (1984)). In Ricardian view, individuals anticipate a present
tax cut as higher government borrowing that turns into the higher taxes in the future so that there
is no change in permanent income. This condition in along with the assumptions of no liquidity
constraints and perfect financial markets lead to no change in private consumption in general
(Barro, 1974). Thus, Ricardian view suggests neither crowding-in nor crowding-out effects of
fiscal policy (Arestis, 2011; Şen & Kaya, 2014). However, if governments change lump-sum
taxes for the fiscal policy, the features of progressive taxes will have impacts on permanent
income and then the aggregate demand and output. As a result, the effectiveness of fiscal policy

most likely depends on how it is paid in the future and the productivity of government
expenditures (Hemming et al., 2002).
As a brief summary, the government expenditure, as according to the Keynesian views, is
needed to cover the lack of consumption in private sectors, which means the fiscal policy
presents a positive effect on economic growth. However, the Keynesian view is lacked of
considering other factors such as institutional environment or debt burden on the effectiveness of
fiscal policy. The neo-classical economics views further explains the effectiveness of fiscal
policy in the some manner relationship with the public debt. In neo-classical views, today’s
individuals think that the existing budget deficits due to the expansionary fiscal policy to
increase the consumption level have to pay back through taxes for future generations. While,
government expenditure is less productive than private investment thus the increased output as a
result of the debt financed government expenditure does not fully offset the negative effect due
to the crowding-out to private investment on output. Therefore, the fiscal policy presents
crowding-out effects at the end. Meanwhile, the Ricardian view suggests that fiscal policy


presents neither crowding-in nor crowding-out effects since private investment and government
spending are considered to behave independently from each other. Where, the increase in
government spending is anticipated to be accompanied by a rise in taxes in the future, thus
government expenditure financed by debts is expected to be repaid by revenue generated through
taxes levied in the future. As the result, interest rates and private investment remain unchanged.
All above economic views require assumptions to be presence such as no liquidity
constraints, perfect financial markets in Ricardian equivalence. However, these assumption are
usually un-existed thus the significance of theories is questioned in both theory and practice
(Haque & Montiel, 1989). Furthermore, there are some cases that the effectiveness of fiscal
policy is explained by all of these views. For instance, if government is restricted by the fiscal
rules to balance the fiscal budget in the long run, thus individuals may partial adjust their
behaviors if they have short-term horizon which presents the presence of both Ricardian and neoclassical views. In the same idea, if the current path of government debt is not sustainable and
future tax increases will be required to lower the debt, the Ricardian view may be presence in
expansionary fiscal policy seemingly with the Keynesian view which depends on the level of

public debt (Sutherland, 1997). Or, if the government expenditure is in line of an upwardtrending stochastic process that individuals believe a sharply fall when it approaches a specific
“target point”, there will be a non-linear relationship between private consumption and
government expenditure (Bertola & Drazen, 1991). Therefore, the argument of a non-linear
relationship between fiscal policy and economic growth makes sense in literature. However, the
literature needs the explanations for the mechanism and empirical evidences.
2.1.2. The empirical works
In fact, many previous studies have investigated the effects of fiscal policy in many
countries, especially in advanced countries such as US, Japan, European area 1. In which,
empirical works usually focus on the relationships between fiscal policy, interest rates, private
investment, exchange rates, and the existence of Ricardian equivalence with three main streams
including the estimation of fiscal multiplier from macroeconomic model simulations, the lesson
studies of fiscal policy, and the determinants of fiscal multipliers (Hemming et al., 2002).
Hemming et al. (2002) summary that the fiscal policy presents mostly with positive multipliers,
it means that government expenditure has positive impacts on economic growth in the short run.
1

See Hemming et al. (2002) for the more detail summary


In addition, they find few evidences of negative short-term multipliers. They also document that
the spending changes have higher fiscal multipliers than the tax changes. However, the long-term
fiscal multipliers, in contrasting to the short-term, are generally smaller and reflect the crowdingout effects of government expenditures.
Recently, Afonso and Strauch (2007) find that the European fiscal policy in 2002 makes
market swap spreads response in mostly around five basis points or less. Similarly, the study of
Kameda (2014a) finds that a percentage point increase in both the projected/current deficit-toGDP ratio and projected/current primary-deficit-to-GDP ratios raises real 10-year interest rates
by 26–34 basis points that implies a crowding out effect of Japanese fiscal policy. Kameda
(2014b) also studies the determinants of the effectiveness of fiscal policy and documents that the
diffusion index of the attitudes of financial institutions have a definite impact on fiscal expansion
effects. In particular, the demand-enhancing effects of government expenditure should be nonKeynesian effects if the existence of liquidity-constrained households when banks’ attitude
toward lending is tight and the financial condition of the government is bad. Bhattarai and

Trzeciakiewicz (2017) use a DSGE analysis to examine the fiscal policy in UK. They note the
highest GDP multipliers for government consumption and investment in the short-run, whereas
capital income tax and public investment have long-run crowding-out effect on GDP. Moreover,
they emphasize that the effectiveness of fiscal policy decreases in a small open-economy
scenario.

Table 1. Some additional evidences of fiscal policy’s effectiveness
Study
Nguyen and Turnovsky (1983)

Sample
Australia,

Empirical evidences
Crowding-out effects

UK, US
D. Cohen and Clark (1985)

US

Crowding-out and crowding-in effects

Kormendi and Meguire (1985)

47

No relationship between government spending

countries

and economic growth

Bradley (1986)

US

Crowding-out effects

Engen and Skinner (1992)

107

Crowding-out effects in the balanced –budget

countries

increase the government expenditure


Easterly and Rebelo (1993)

100

Public

investment

in

transport

and

countries

communication is consistently correlated with
growth

van de Klundert (1993)

Europe,

Crowding-out effects on exports

US
Devarajan, Swaroop, and Zou 43

Crowding-in

(1996)

expenditures

developing

effects

of

public

current

countries
Kneller, Bleaney, and Gemmell 22 OECD Crowding-in

effects

(1999)

crowding-out

countries

expenditures,

for

productive
effects

for

distortionary taxations
Ahmed and Miller (2000)

39

Public investment in transport has crowding-in

countries

effects, tax-financed expenditure has crowdingout effects

Leleux and Surlemont (2003)

15

Public investment crowding-ins the private

European

venture-capital

countries
Heutel (2014)

Charity

government grants crowd in private donations

activities
Şen and Kaya (2014)

Turkey

Government

current

transfer

spending,

government current spending, and government
interest spending crowdout private investment,
whereas government capital spending crowds-in
private investment
Kasselaki

and

Tagkalakis Greece

(2016)

A

government

spending-based

fiscal

consolidation improves financial markets and
boosts economic sentiment

da Silva and Vieira (2017)

113

Fiscal policy behaves in a procyclical way only

countries

in the pre-crisis period and there is no
significant relationship between output gap and
government spending for the post-crisis period


Besides the presence of plentiful empirical literature in the effectiveness of fiscal policy, this
field of study is got much less evidence on the short-term effects in developing countries due to
data deficiencies, the structural/institutional factors in the last century (see Hemming et al.
(2002)). For instance, Haque and Montiel (1989) find that the Ricardian equivalence is not
supported in the developing countries due to liquidity constraints. Montiel and Haque (1991) go
further by using the Mundell-Fleming model with rational expectations and full employment for
31 developing countries and conclude that the increasing of government expenditures have
contractionary short-term and medium-term effects. Previous, M. S. Khan and Knight (1981)
find positive nominal income elasticities of government expenditures and taxes and they are
close to unity in 29 developing countries. Then, other empirical studies such as Agenor and
Montiel (1996), Easterly, Rodriguez, and Schmidt-Hebbel (1994), Rama (1993) document
evidences that fiscal policy has crowding-out effects on private investment through the impacts
on interest rates in developing countries. Meanwhile, empirical studies also provide evidences
supporting for partial or/and fully existences of the Ricardian equivalence in developing
countries such as Agenor and Montiel (1996), Corbo and Schmidt-Hebbel (1991), Masson,
Bayoumi, and Samiei (1995), Giavazzi, Jappelli, and Pagano (2000).
However, the economic development in emerging market economies, which is a new
definition of the development level of economies and nearly relating to the developing countries
definition, boosts their roles in the world economy. In addition, the better fulfill of data have rehighlighted the interesting in investigating the effectiveness of fiscal policy by adding more

methods and conditions into model for this group. For example, Cuadra, Sanchez, and Sapriza
(2010) note that emerging market economies typically exhibit a pro-cyclical fiscal policy, where
governments increase (decrease) expenditures in economic expansions (recessions) and rise
(reduce) tax rates in bad (good) times. This situation is in line with the characteristic of countercyclical default risk in their business cycle. They also document the importance of incomplete
markets and sovereign default risk premium in explaining the pro-cyclicality of public
expenditures and tax rates in these economies. Therefore, the assumptions of Ricardian view are
not existed that propose for the Keynesian or neo-classical views of fiscal policy.
For instance, Papageorgiou (2012) emphasizes that government should decrease the tax rate
on labour income and increase the consumption tax rate to stimulate the economy and increase


welfare, while higher public investment spending is good for the economy. In the same direction
of study in Greece, Kasselaki and Tagkalakis (2016) find that the tax based fiscal consolidation
has more pronounced and more protracted negative effects on output, while the government
spending-based fiscal consolidation improves financial markets and boosts economic sentiment.
While, Akanbi (2013) tests the effectiveness of fiscal policy with existing structural supply
constraints versus demand-side constraints in South Africa for the period 1970 – 2011. The
results suggest that fiscal policy is more effective in conditions of limited or no supply
constraints. In addition, expansionary or consolidating fiscal policies through government
spending changes will be more effective in condition of no structural supply constraints, while
tax changes will be more effective in contrasting cases. Jha, Mallick, Park, and Quising (2014)
go further to examine fiscal policies in 10 emerging Asian countries and find that tax cuts have a
greater countercyclical impact on output than government spending.
No surprising that the debate on the role and the effectiveness of fiscal policy are continuous
argued broadly in both literature and practice. Recently, Arestis (2011) notices that the “New
Consensus in Macroeconomics”, recent developments in macroeconomics and macroeconomic
policy, downgrades the role of fiscal policy in contrasting with monetary policy due to its
ineffective. Through a careful literature review and discussion at recent developments on the
fiscal policy front, he then concludes that fiscal policy does still have a significant role to play as
an instrument of economic policy through its impact on allocation, distribution and stabilization.

However, researchers and authorizers have to careful consider the assumption in economic
theories of fiscal policy’s effectiveness as the existence of Ricardian and non-Ricardian
economic agents, liquidity-constrained households, and the endogenization of labour supply and
capital accumulation. Whereas, other features of the economy should be considered in study the
effectiveness of fiscal policy such as the institutional framework and the debt burden.

2.2. The institutions and the effectiveness of fiscal policy
In fact, the dependence of fiscal policy’s effectiveness on institutional aspects is discussed
under the literature with two main strands including the inside and outside lags of effects and the
political economy considerations (Hemming et al., 2002). First, the fiscal policy has inside and
outside lags, where the inside lags present the needed time to see that fiscal policy should
changes, the outside lags are the function of the political process and the effectiveness of fiscal


management that is the time for fiscal measures take effects on aggregate demand (Blinder &
Solow, 1974). Due to the long time to design, approval, and implementation, the inside lag may
be longer, while the outside lag is more variable depending on the institutional environment.
Second, the fiscal policy is impacted by the political considerations such as the fiscal illusion of
public and policy-makers, the favor of transferring current fiscal burden to future generations,
the limitation of government due to the debt accumulation, the delay of fiscal consolidations due
to the political conflicts, and the function of current budget institutions that leads to high
spending.
The institution is defined as the social rules of the game (Douglass C North, 1990), which
includes “humanly devised”, “the rules of the game” to set “constraints” on human behavior, and
the economic incentives (see Douglass Cecil North (1981), Acemoglu and Robinson (2008)).
The better institutions reduce asymmetric information problem, transaction cost, and risk, while
they improve the market efficiency, especially efficiency of asset allocation (K. J. Cohen,
Hawawini, Maier, Schwartz, & Whitcomb, 1983; T. S. Ho & Michaely, 1988; M. Khan, 2007;
Williamson, 1981). Therefore, the better institutions should have positive associations with the
effectiveness of fiscal policy since the lower asymmetric information problem, transaction cost,

and higher market efficiency reduce both the inside and outside lags that then increase the
efficiency of fiscal policy, especially the short-term effects. Moreover, the problems of inside
and outside lags are more important in emerging market economies, thus the improvement in
institutional framework is expected with higher enhancing impacts on the effectiveness of fiscal
policy. In addition, the better institutions also reduce the fiscal illusions, the political conflicts,
while it pushes more responsibility of governments in building and implementing fiscal policy,
the fiscal policy, in turn, should be more effective.

Table 2.Studies relating to the institutions and its impacts to fiscal activities
Study
Kaufmann,

Sample
Kraay, 150

and Zoido-Lobatón countries

Results
a strong causal relationship from better governance to
better development outcomes

(1999)
Kaufmann

and 175

The improvement in public governance shapes public


Kraay (2002)

countries

policy.

Neumayer (2002b)

136

Governments are accountable, respect democratic rights

developing

as well as refrain from imposing burdens on business

countries

have a statistically significant influence on country’s debt
forgiveness

Baldacci

et

(2004)

al. 39

low- Factor productivity is some four times more effective

income

than investment as a channel for increasing growth

countries

through fiscal policy in low-income countries. Highdeficit low-income countries can nonetheless benefit by
reducing

unsustainable

fiscal

deficits

because

of

governance-related factor productivity responses that
increase growth.
Cubbin and Stern 28

A regulatory law and higher quality governance is

(2004)

developing

positively and significantly associated with higher per

countries

capita generation capacity levels and higher generation
capacity utilisation rates, and this positive regulatory
impact appears to increase with experience at least for
three years or more.

Azmat and Coghill Bangladesh

the need for integrated governance linking government,

(2005)

business and civil society as paramount for promoting
good governance for the success and sustainability of the
reforms

Capuno (2005)

Philippines

The uneven quality of local governance thus may have
contributed to imbalanced regional growth in the fiscal
decentralization process

Quibria (2006)

Gani (2007)

29

Asia no strong positive link between governance and growth:

countries

paradoxically

countries

The rule of law; control of corruption; regulatory quality;

from Asia government effectiveness and political stability are
and

Latin positively correlated with FDI

America


Rajkumar

and 91

Public spending on primary education becomes more

Swaroop (2008)

countries

effective in countries with good governance

Cooray (2009)

71

Both the size and quality of the government are important

countries

for economic growth

Dzhumashev (2014)

Simulations Corruption improves economic efficiency only when the
for

UK, actual government size is above the optimal level

Turkey,
Kenya

In fact, the empirical literature in the field of fiscal policy had considered the role of
institutional framework in some manners such as politics, democracy, economic freedom, and
corruption in recent decades. Nelson and Singh (1998), for instance, argue that a democratic
political system permits active in a voluntary way, at the same time it creates competitive market
forces conditions for economic growth. They also emphasize that the ineffective democracy
regimes in developing countries detriments the growth. Lockwood, Philippopoulos, and Tzavalis
(2001) add that the political equilibrium determines the path of government expenditures, taxes
and debt in Greece in the period 1960-1972, which means the fiscal policy may not follow a

long-term efficiency for the country. Martinez-Vazquez et al. (2007) notice that the elimination
of corruption is not usually an economic objective for the development, but the frustration with
the lack of effectiveness of traditional economic theories and the recognition of the important
roles of institutions and good governance practices have led the more attention to the corruption.
Precisely, Dimakou (2015) finds that corruption constrains the fiscal capacity to tax and
increases the reliance on inflation.
However, no comprehensive study has considered the fiscal policy’s effectiveness under the
institutional framework. More interesting, it lacks of empirical study in emerging market
economies, which have more space in improving institutional quality and the economic growth.
For example, the study of Aidt, Dutta, and Sena (2008) document that corruption has a
substantial negative impact on economic growth in high institutional quality economies,
otherwise it has no impact on economic growth in low quality one. P.-H. Ho, Lin, and Tsai
(2016) find that the improvement in country governance just enhances the effectiveness of banks
and then promote the economic growth in developing countries, while it reduces these effects in


developed countries due to smaller spaces for improvement. In addition, Wang, Cheng, Wang,
and Li (2014) argue that the improvements in institutional quality just have strong effects on
promoting economic development only when institutional quality is within a certain range.
Therefore, we can argued that the improvement in institutions has strong impacts on the
effectiveness of fiscal policy in emerging market economies.
2.3.The external debt and the effectiveness of fiscal policy
The debt burdens, on the other hand, are also concerned in the literature of fiscal policy
effectiveness.

Table 3. Studies relating to the external debts and fiscal policy
Study

Sample

Results

Sutherland

Theoretical

At moderate levels of debt fiscal policy has the traditional

(1997)

arguments

Keynesian effects.

Amato

and Italia

All fiscal and debt management indicators display

Tronzano (2000)

significant

effects

on

exchange

rate’s

devaluation

expectations
Alt and Lassen OECD

A higher degree of fiscal transparency is associated with

(2006)

lower public debt and deficits

countries

Ardagna, Caselli, 16
and Lane (2007)

OECD The effect of debt on interest rates is non-linear: only for

countries

countries with above-average levels of debt does an
increase in debt affect the interest rate

Lima,

Brazil

The fiscal austerity led to very low rates of growth for the

Gaglianone, and

Brazilian economy, with negative impact on employment

Sampaio (2008)
Jayaraman
Lau (2009)

and 6

major External borrowing contributes to growth in PICs in the

Pacific island short run, growth enhances the image of a PIC as an
countries

efficient user of borrowed funds, enabling it to borrow from
abroad on better terms; consequently, higher growth
results in further rise in external debt level


Drine and Nabi 27 developing An increase of the public external debt share increases the
(2010)

countries

production efficiency through a positive externality effect.
However, it generates an opposite effect via the reduction
of the formal sector’s size in favour of a less efficient
informal sector. The resultant effect becomes negative

beyond an optimal level

Claeys, Moreno, OECD

The crowding out effect of public debt on domestic long

and

term interest rates is small. Emerging markets are not as

Suriñach countries

(2012)

well integrated into international capital markets, causing a
stronger crowding out effect

Bal

and

Rath India

Central government debt, total factor productivity (TFP)

(2014)

growth, and debt-services are affecting the economic
growth in the short-run

Ramzan

and Pakistan

Ahmad (2014)

External debt has a negative impact on growth, but this
adverse effect can be reduced or even reversed in the
presence of sound macroeconomic policy

Siddique,

40

heavily debt as a share of GDP has a negative influence in the

Selvanathan, and indebted poor short-run as well as in the long-run growth
Selvanathan

countries

(2016)
Galstyan
Velic (2017)

and 10

emerging The nominal exchange rate and inflation differentials are

market

more important determinants in states of high debt than in

economies

states of low debt

According to the review of Hemming et al. (2002), the debt accumulation may be used as a
strategic instrument to limit the fiscal room for the maneuver of future government, while the
availability and cost of domestic and external financing is often a major constraint on fiscal
policy in developing countries. Thus, an emerging market economy with highly level of debts
will determine the size of fiscal deficit in facing with more difficulties in assessing to
international capital market (inaccessible or accessible with unfavorable terms), which then leads


to the stronger crowding-out effects. Meanwhile, the low indebted countries have higher fiscal
room for future government in implementing fiscal policy, which may undertake with the
favorable terms of debt-financing, and that in turn promotes the crowding-in effects.
Moreover, the individuals in high indebted countries are more sensitive to the government
expenditures in following the framework of neo-classical views. The public may expect that the
increasing of government expenditures in this case be in along with the less favorable terms of
government’s borrowings and less efficiency of spending, which then stimulate individuals to cut
back their current consumption more and more. As a result, this proposes higher crowding-out
effects of fiscal policy. In contrast, the individuals in low indebted countries may less sensitive to
the government expenditures, especially through the debt-financing spending, since the interest
rates are less responsive and they are easier to access the financial markets, thus the fiscal policy
is argued with the existence of crowding-in effects.
According to Kirchner and Wijnbergen (2016), the effectiveness of fiscal stimuli is impaired
when banks are substantially invested in sovereign debt since deficit-financed fiscal expansions
reduce private access to credit in this case. Therefore, we use the total external debt, which

includes public debt and private debt in this study to examine the impacts of debt on
effectiveness of fiscal policy. This helps us consider the constraints of external debt of ability of
private sector in accessing international financial markets. We argue that the expansionary fiscal
policy in the highly indebted countries not only creates the crowding-out effects for the private
sectors through the impacts on interest rates and exchange rates, but also crowds out the
availability of private sectors in accessing into the international financial markets that creates
more constraints for private sectors to implement economic activities. In contrast, these effects
may not exist or less significance in the case of low indebted countries. As a summary, our
hypothesis is argued that the relationship between fiscal policy with the economic growth is nonlinear one as the positive effect in the low indebted level and the negative effect in the high
indebted level.
In fact, the non-linear relationships between fiscal policy and economic factors are examined
under some manners. Adam and Bevan (2005) investigate the relationship between fiscal deficits
and growth for a panel of 45 developing countries and find evidence of a threshold effect at a
level of the deficit around 1.5% of GDP. They also find evidence of interaction effects between
deficits and debt stocks with high debt stocks exacerbating the adverse consequences of high


deficits. While, Catão and Terrones (2005) examine inflation as non-linearly related to fiscal
deficits through the sample of 107 countries over 1960–2001 period. They find a strong positive
association between deficits and inflation among high-inflation and developing country groups,
but not among low-inflation advanced economies.
This fact suggests that we should consider the non-linear relationship between fiscal policy
and eonomic growth in the emerging market economies. Emerging market economies are an
emerging group of countries with interesting economic features in developing countries. While,
the expected future revenue plays an important role in explaining the low fiscal limits of
developing countries relating to developed countries (Bi, Shen, & Yang, 2016). Therefore, the
study of the relationships between institutions, external debts and the effectiveness of fiscal
policy is more significant for both literature and practice. Next section presents the methodology
and data.

CHAPTER 3: METHODOLOGY AND DATA
3.1. Methodology
It is easy to begin with the production function of an economy in the methodology to
examine the determinants of economic growth following the founding study of Cobb and
Douglas (1928) as:
𝑌 = 𝐴𝐾 𝛼 𝐿𝛽

(1)

in which: Y is the total output; K is the capital input; L is the labor input; A is the total factor
productivity; α and β are the output elasticities of capital and labor, respectively. Then, taking
logarithm both sides of eq.1, we have:
𝐿𝑜𝑔(𝑌) = 𝐿𝑜𝑔(𝐴) + 𝛼𝐿𝑜𝑔(𝐾 ) + 𝛽𝐿𝑜𝑔(𝐿)

(2)

According to eq.2, the changes in capital input, labor input and total factor productivity
determine the output growth as the simplest function of economic growth. Thus, economic
growth can be presented as the function:
f(g) = f(T, K, L, e)

(3)

in which: g is the proxy of economic growth, T is the technology, K and L are same as previous,
and e the random shocks in output.
In fact, technological factor in economic growth is anything that is new in product or
production method, or raw material, or business area, or financial method, or organization
scheme (Schumpeter, 1934). This then is deeply studied in many previous works including the
pioneer works as Solow (1956), Arrow (1962), Lucas (1988), Paul M Romer (1986), Paul

Michael Romer (1991). And, recent studies such as Shahiduzzaman and Alam (2014), Luo,
Olechowski, and Magee (2014), Pradhan, Arvin, and Norman (2015), Maswana (2015),
Salahuddin and Alam (2016), Bhattacharya, Rafiq, and Bhattacharya (2015), and Kurt and Kurt
(2015). In addition to technology, the capital formation is other important determinant of
economic growth especially in low developed economies (Ghosal & Nair-Reichert, 2009;
Kapelko, Oude Lansink, & Stefanou, 2015; Malaga-Toboła, Tabor, & Kocira, 2015). From the
Harrod–Domar growth model in 1940s through Harrod (1939) and Domar (1946), economic
growth can explained in terms of the saving level and capital productivity, thus the increasing of
capital investment is a stimulating factor for economic growth (Sato, 1964; Solow, 1988).


Due to the lack of concerns about the productivity in Harrod-Domar model, the exogenous
growth model (or Solow model) was developed by Solow (1956) and Swan (1956) that had
included the term of productivity growth to explain long-run economic growth through capital
accumulation, labor or population growth, and increases in productivity or technological
progress. Then, Paul M Romer (1986), Lucas (1988), and Rebelo (1990) propose the endogenous
growth theory in the mid-1980s from the basic studies of Arrow (1962), Uzawa (1965), and
Sidrauski (1967). Endogenous growth theory suggests that besides other main drivers of
economic growth the endogenous factors of economic growth are investment in human capital,
innovation, and knowledge. This theory also focuses on positive externalities and spillover
effects of a knowledge-based economy such as the contributions from the capital flows and trade
openness. In fact, many empirical studies recently have applied the endogenous growth theory in
examining the determinants of economic growth (see Lee (2005), Bronzini and Piselli (2009),
Arvanitis and Loukis (2009), Teixeira and Fortuna (2010), Banerjee and Roy (2014), van Lottum
and van Zanden (2014), Chowdhury, Schulz, Milner, and Van De Voort (2014), Scherngell,
Borowiecki, and Hu (2014), Breton (2015), Zlate and Enache (2015)).
Beside the common elements of economic growth as stated, there is an extensive literature
has focused on the impacts of financial development on economic growth (Levine, 2005).
Schumpeter (1912) is seen as the first economist who addressed the relation between financial
development and economic growth. He argued that a well-functioning financial system should

promote economic growth through the selection of the productive investments which are the
most likely to be successful and the efficient allocation of resources (via bank credits) to these
innovative technologies. Financial markets overcome transaction costs and informational
asymmetries to reduce liquidity constraints and improve the allocation of capital, it increase the
physical capital accumulation and productivity growth to affect to economic growth (Rioja &
Valev, 2009). In which, stock markets and banking sector both enhance economic growth (Beck
& Levine, 2001, 2004; Rousseau & Wachtel, 2002), but banking sector is more important in
developing countries.
Therefore, we recruit the common determinants of economic growth including capital,
technology, labor, technology, capital flows, trade openness, and add the credit element for the
basic model of economic growth in this study. With this beginning of basic model, we
incorporate government expenditure to examine the impacts of fiscal policy on economic growth


for 20 emerging market economies in the period 2002-2014, and follows the empirical model in
Miller and Russek (1997):
𝑔𝑖,𝑡 = 𝜕1 𝑔𝑖,𝑡−1 + 𝜕2 𝑔𝑑𝑝𝑝𝑐𝑖,𝑡−1 + 𝛼𝑋𝑡 + 𝛽1 𝐺𝑜𝑣𝑒𝑥𝑔𝑖,𝑡 + 𝜀𝑡,𝑠 with ε s  i.i.d .N (0, δ s2,t ) (4)

in which: i and t is country i at time t. g is GDP growth rate (gdpg) that proxies for the economic
growth. The lag of g is put into the model to control for the dynamic of economic growth model,
while the gdppc is logarithm of GDP per capita that presents for the starting economic
development level. X is vector of control variables including: the capital investment factor that
presented by the gross capital formation growth rate (capg); the labor factor that presented by the
population growth rate (popg); the credit factor that presented by the logarithm of domestic
credit to private sector by banks (credit); the technology factor that presented by the logarithm of
total patent applications by both residents and non-residents (patent); the trade openness that
presented by the logarithm of total trade to GDP (trade); and the capital flow that presented by
the net inflows of foreign direct investment to GDP (fdi). govexg is the proxy for fiscal policy
that presented by the general government final consumption expenditure growth rate. All the
definitions and sources of variables are presented detail in Table 1.

Table 4. Variables, definitions and sources
Variables
Dependent
variables

Gdpg
Gdppcg

Gdppc
Capg
Popg
Control
variables

Credit
Patent
Trade
Fdi
Debt

Explanatory Govexg
variables Goveff
Regu

Definitions
GDP growth rate (% annual)
GDP per capita growth rate (% annual)
Independent variables
Logarithm of GDP per capita
Gross capital formation growth rate (% annual)

Population growth rate (% annual)
Logarithm of domestic credit to private sector by
banks
Logarithm of total patent applications by both
residents and non-residents
Logarithm of trade ratio to GDP
Net inflows of foreign direct investment to GDP
(%)
Ratio of External debt stock to GNI (%)
General government final consumption
expenditure growth rate (% annual)
Government effectiveness indicator
Regulatory quality indicator

Sources
WDI
WDI
Calculation from
WDI
WDI
WDI
Calculation from
WDI
Calculation from
WDI
Calculation from
WDI
WDI
WDI
WDI

WGI
WGI


Concor

Control of corruption indicator

WGI

In next step, we also incorporate institutional factors into the model to investigate the effects
of institutional quality on economic growth following the empirical model suggesting in Lee and
Hong (2012). In this step, we collect three dimensions of institutions from World Governance
Indicators (Worldbank) including the government effectiveness (Goveff), regulatory quality
(Regu), and control of corruption (Concor) to proxy for the institutional framework, respectively.
Despite of critics about bias or lack of comparability and the utility of institutional quality in
World Governance Indicators (Thomas, 2010), there are many previous studies that use these
indicators as the best proxies for institutional quality (see Kaufmann and Kraay (2002),
Neumayer (2002a), Neumayer (2003), Dollar and Kraay (2003), Naudé (2004), Lash (2004),
Llamazares (2005), Neumayer (2005), Andrés (2006), SCHUDEL and Schudel (2008), Clist
(2011), Park (2012), In’airat (2014), Herrera-Echeverri, Haar, and Estévez-Bretón (2014),
Nordveit (2014), Barry and Tacneng (2014), Zhang (2016)).
As explained in World governance indicators, government effectiveness “reflects perceptions
of the quality of public services, the quality of the civil service and the degree of its
independence from political pressures, the quality of policy formulation and implementation, and
the credibility of the government's commitment to such policies”. Regulatory quality “reflects
perceptions of the ability of the government to formulate and implement sound policies and
regulations that permit and promote private sector development”. And control of corruption
“reflects perceptions of the extent to which public power is exercised for private gain, including
both petty and grand forms of corruption, as well as "capture" of the state by elites and private

interests”. We believe that these aspects of institution have the highest significance for the
effectiveness of fiscal policy as discussed in literature review. As one of the most important aim
of this study, we use the interaction terms between fiscal policy with each indicator of
institutions to examine the effectiveness of fiscal policy under the associations with institutional
framework.
Next, we estimate the growth model with the explanatory variables including both external
debt to GNI and its square to examine the non-linear relationship between external debt and
economic growth. Basing on the results of these estimations, we then divide sample into two
sub-samples. First sub-sample includes 8 low indebted economies that have the average external


debt to GNI lower than 40% in the period of 2002-2014, which are Bangladesh, Brazil, China,
Colombia, Egypt, India, Mexico, and South Africa. Second sub-sample includes 12 high
indebted economies that have the average external debt to GNI higher than 40% in the testing
period, which are Argentina, Bulgaria, Indonesia, Malaysia, Pakistan, Peru, Philippines,
Romania, Russia, Thailand, Turkey, and Vietnam. Our method to divide sample following the
instruction of IMF in debt management for governments that 40% is the threshold for the
sustainability of a country with external debt. Then, we apply the previous procedures to two
sub-samples separately to investigate the effectiveness of fiscal policy under two debt regimes.
At the last step, in order to ensuring the robustness of our results, we use GDP per capita
growth rate to replace for GDP growth rate and replicate these strategy. By doing this strategy,
we ensure the significant and robustness of our results, which then have significant contributions
to both theory and practice.
There are some serious biases with fixed effects model when we estimate the dynamic panel
data in the eq. 4. It is due to likely of endogenity for the most variables on the right side equation
such as government expenditure, the capital investment and the correlation between lag
dependent variable with error term in the context of a dynamic panel data model (Nickell, 1981).
Therefore, other methods are proposed in the econometric literature (e.g. Anderson and Hsiao
(1982), Manuel Arellano and Stephen Bond (1991), Blundell and Bond (1998)). M Arellano and
S. Bond (1991)) to use difference GMM estimator that is better dealing with endogeneity,

heteroscedasticity, and serial correction. However, this method technique may sometimes
generate many instruments and the variance of the estimates may increase asymptotically and
create considerable bias (Blundell & Bond, 1998; Soto, 2009). Then, the system GMM estimator
is also developed in this case and it displays better estimators (Soto, 2009). So, we recruit both
the first difference and system-GMM estimator as our econometric methods in this study and
only present better unbiased estimators.

3.2. Data
Our data is collected yearly from the period of 2002 - 2014 for 20 emerging countries 2 due to
the time limitation in World governance indicators that have continuous data from 2002 to 2014.
2

20 emerging markets are defined in introduction section and the number of emerging market economies is due
to the availability of data.