The nexus among corruption, public expenditure and economic growth evidence from provinces of vietnam

The findings are: 1 The lower corruption is associated with the lower in public expenditure in development investment; 2 The lower corruption has a positive effect on economic growth; 3

VIET NAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE NEXUS AMONG CORRUPTION, PUBLIC

EXPENDITURE AND ECONOMIC GROWTH: EVIDENCE

FROM PROVINCES OF VIETNAM

By

NGUYEN THI MY PHUC

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH, November 2015

VIET NAM THE NETHERLANDS

VIET NAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE NEXUS AMONG CORRUPTION, PUBLIC

EXPENDITURE AND ECONOMIC GROWTH: EVIDENCE

FROM PROVINCES OF VIETNAM

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

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

NGUYEN THI MY PHUC

Academic Supervisor:

DR NGUYEN HOANG BAO

HO CHI MINH, November 2015

i

economic growth under the presence of corruption Three stage least squares method is employed in order to indicate this effect in the panel data of 63 provinces/cities in Vietnam from the period 2008-2013 The findings are: (1) The lower corruption is associated with the lower in public expenditure in development investment; (2) The lower corruption has a positive effect on economic growth; (3) Public spending in development investment has a positive effect on economic growth and the presence of corruption do not reduce this effect; (4) The lower corruption has a positive effect on private investment; (5) In term of the openness, foreign investment can reduce corruption Thereby, some implications have been proposed to foster the growth of economy

Keywords: corruption, public investment, public expenditure, Three stage least squares method

iii

ACKNOWLEDGEMENT II TABLE OF CONTENTS III ABBREVIATIONS V LIST OF FIGURES VI LIST OF TABLES VII

CHAPTER 1 INTRODUCTION 1

1.1 PROBLEM STATEMENT 1

1.2 RESEARCH OBJECTIVE 2

1.3 RESEARCH HYPOTHESES 2

1.4 THESIS STRUCTURE 3

CHAPTER 2 LITERATURE REVIEW 4

2.1 T HE K EYNESIAN THEORY ABOUT GOVERNMENT EXPENDITURE 4

2.2 G OVERNMENT EXPENDITURE IN N EO - CLASSICAL THEORY 5

2.3 A BOUT CORRUPTION IN PUBLIC EXPENDITURE IN INVESTMENT 6

2.4 E MPIRICAL REVIEW 7

2.4.1 Public spending in investment and economic growth 7

2.4.2 Corruption and public spending in investment 8

2.4.3 Corruption and growth 9

2.4.4 Private investment in the relationship with growth, corruption and public investment 9

2.4.5 Foreign investment in the relationship with growth and corruption 11

2.5 C ONCEPTUAL F RAMEWORK 12

CHAPTER 3 METHODOLOGY AND DATA 13

3.1 METHODOLOGY 13

3.2 DATA SOURCES AND CHARACTERISTICS 14

3.2.1 DATA SOURCES 14

3.2.2 MEASURE GDP FOR EACH PROVINCE 15

3.2.3 CHOICE OF CORRUPTION VARIABLE 16

3.2.4 ENDOGENOUS AND EXOGENOUS VARIABLES 17

3.3 MODEL SPECIFICATION 18

3.4 ESTIMATION METHOD 20

iv

4.1.2 CORRELATION OF COEFFICIENTS 25

4.2 THE IDENTIFICATION PROBLEM OF THE SIMULTANEOUS EQUATION 28

4.3 RESULTS 29

CHAPTER 5 CONCLUSION AND SUGGESTION 38

5.1 CONCLUSION 38

5.2 POLICY IMPLICATIONS 38

5.3 SUGGESTIONS FOR FURTHER RESEARCHS 39

5.4 THE LIMITTED OF THESIS 39

REFERENCE 41

APPENDICES 46

v

FDI Foreign Direct Investment

FGLS Feasible generalized least squares

GDP Gross Domestic Product

GMM Generalized Method of Moments

GSO General Statistics Office

IMF International Monetary Fund

OLS Ordinary least squares

PCI Provincial Competitiveness Index

TSLS Three Stage Least Squares

USAID United States Agency International Development UTC The Unmatched Count Technique

VCCI Vietnam Chamber of Commerce and Industry VND Vietnam Dong

vii

Table 4.2: Correlation of coefficients 25

Table 4.3: Three stage least squares regression with impact of corruption on government expenditure in investment and economic growth (when i.id variable is not considered) 30

Table 4.4: Three stage least squares regression with impact of corruption on government expenditure in investment and economic growth (when i.id variable has been used) 31

Table 4.5: Three stage least squares regression with impact of corruption on government expenditure in investment and economic growth 33

Table 4.6: Three stage least squares regression with corruption affecting government expenditure in investment, taking an interaction variable into consideration 34

Table A.1: The list of provinces 46

Table A.2: Data sources 47

Table A.3: The development investment expenditure of Vietnam (2007-2013) 47

Table A.4: The data on the development investment capital of Vietnam (2007-2013) 48

Table A.5: The ranking of 63 provinces/cities in the PCI index in the period from 2008 to 2014 50

of public investment depended on institution It was found that a better institutional setting may foster the positive effects of public investment (de la Croix and Delavallade, 2009; Dort

et al., 2014) While a poor institution, especially corruption, waste is popular (Tanzi and Davoodi, 2002b)

Firstly, corruption impacts the public investment in both quantity and quality There is

a wide spread of corruption on over the world, especially in developing countries in the recent years From its detrimental impact on economic growth, corruption distorts decision of higher political, retards economic growth (Tanzi, 1998) Heath W (2000) stated that: “corruption is a cancer in society and is slowly eating away at the public sector” It reduces the return in public investment, lead to the economic development down following by Kneller R and Haque M E (2008) In conclusion, through the impact on the public investment, corruption retards the economic development

Vietnam is a developing country in Southeast Asia with over 90 million inhabitants

By political and economic reforms (Doi Moi) launched in 1986, Vietnam’s GDP per capita increase from around 100$ to over 2000$ by the end of 2014 within a quarter of a century In reducing poverty, Vietnam has a dramatic change in which the number of people dropped from over 50% in the early 1990s to 3% today (World Bank, 2015)

With the development of two big cities is Ha Noi and Ho Chi Minh, Vietnam inherited the contribution of 63 provinces in the process of its own economic development The GDP of

Ha Noi and Ho Chi Minh account for 10.3% and 20% of the GDP of whole country

2

respectively (Provincial statistic yearbook, 2013) GDP per capita is the highest at Ba Ria Vung Tau province (about 160 million VND in 2013) and Lai Chau province is the smallest (9 million VND in 2013) The rate in GDP per capita was also found different among 63 provinces/cities

The share of public investment or public expenditure in investment in Vietnam accounted for a high rate of government expenditure activities In detail, in the period 2017-

2013, average public investment expenditure of Vietnam accounted for 29% of budget (show

in Appendix 3) This average figure on the provincial scale is 25% (Bac Can province) and 22.5% (Cao Bang province) respectively in period 2008-2013 (source from Provincial

Statistical Yearbook, 2008-2013) However, the growth rate in GDP of Bac Can province and

Cao Bang province is quiet small, 5.24% and 4.78% in 2013 be compared with this rate of a whole country is 5.42% in 2013 The question is whether corruption could be a cause make differently in provincial economic growth through the impact of public investment?

In this research, we aim to investigate whether public expenditure in investment has a negative effect on provincial economic growth when there is a presence of corruption’s effect

or not

1.2 RESEARCH OBJECTIVE

This paper examines whether corruption has a negative effect on public investment and economic growth, and through this investment decelerates economic growth In other words, the study test whether bureaucracy or “grand” corruption results in increasing the share of public investment on GDP leading to lower economic growth rate In order to get that, this research find out more about corruption on its own expansion will affect public expenditure in investment, and whereby has direct and indirect effect to economic growth

3

H4: The lower corruption has a positive effect on private investment

H5: The foreign corruption has a negative effect on corruption

1.4 THESIS STRUCTURE

The thesis remains in five chapters Chapter 2 reviews literature about the relationship among public investment, economic growth and corruption, also some other factors effect to the development such as foreign investment and private investment Chapter 3 discusses the data characteristics and an estimation method employed in the study Chapter 4 presents empirical result on the scope of 63 provinces/cities in Viet Nam in period 2008-2013 Chapter

5 is the conclusion of this study, comment about the policy implication, also the limitation of this study whereby suggest some further studies

4

CHAPTER 2 LITERATURE REVIEW

This chapter presents the theory and empirical review about various links among public expenditure, corruption, private investment, foreign investment and economic growth

2.1 The Keynesian theory about government expenditure

Following Keynesian theory, fiscal policy can be used to intervene in the economy by government through the impact of public expenditure There is an increasing in government expenditure will lead to a risen in economic growth in the condition of an expansionary fiscal policy (cited in Odihiambo, 2015) Public expenditure stimulates consumer demand, arouses business investment, support employment and contribute to rising income That if government expenditure is not occurred, the economic growth will go down, have no building infrastructure, and protect civil rights

The public expenditure investment in public good is usually lower in return, need a long time to payback, but it is necessary for the social economic development The public spending in investment has been creating more productivity, a long term impact to the economy Beside, spending in budget is also instrument to boost private investment by its support in the quality in infrastructure, the environment investment, etc

However, if there is an overspending in budget (exceeded the optimal thresholds), it will prevent the economic growth This is similar with Richard Rahn (1986) (cited in Anh P.T 2008) That is there is an optimal level of government spending in economic growth maximizing The optimal level of government spending is between 15% and 25% of GDP Its efficiency is not only depending on what government spends but also the level of spending Tejvan Pettinger (2008)

5

Figure 2.1: Rahn Line (source: Tejvan Pettinger (2008))

2.2 Government expenditure in Neo-classical theory

In 1980s, there is a dramatically risen in some model of neoclassical about the role of public expenditure in economic growth In this time, many empirical researches were born Typically, it is begun with the endogenous growth model of Barro (1990) on the research about the role of public spending and tax on the economic growth process To build this model, Barro (1990) based on the maximum welfare behavior of all sectors in the economy included government region, households, and in production of private firm On the production sector aspect, public service is considered as inputs in the production of final good At the beginning, the productive government has a positive impact to growth and saving rates, but subsequently decline

After that, an endogenous growth had been developed by Futagami, Morita, and Shibata (1993) with private and public capital in dynamic analysis; Cashin (1995) with the influence of public investment, public transfer and distortionary taxation on the rate of growth;

or Glomm and Ravikumar (1997) with the growth effect of productive government spending Kneller et al (1999) find a strong support for Barro model (1990) that productive government expenditure enhances the steady-state growth rate Furthermore, the identifying of Ghosh and Roy (2004) to analyze the impact of element of public spending in both public capital and

Government spending as a share of GDP

6

public services on economic growth Government spends on both public service for short run and public capital for long run Pubic capital is such as roads, railways, airports and other forms of infrastructure or the long time projects Public service for short run is such as spending on policing and the maintenance of road, and it has more immediate effect

2.3 About corruption in public expenditure in investment

Corruption is “the single greatest obstacle to economic and social development” (cited

in Farooq et al, 2013) Tanzi (1998) discussed the relationship between corruption and government expenditure in term of investment project, procurement spending, and extra budgetary accounts Firstly, there is appearance of “commission” which some higher officials will be received when projects has been chosen Because the existence of discretion of high-level public official is truth, distorted decisions regarding on public investment can become more popular In fact, these decisions will effect to the efficiency in government’s project investment activity in some aspect such as: the size, composition, the usage in resources

In reality, public investment projects are very large but useless The procedure of this case was described as: when some enterprises will be chosen to undertaking the project, a

“commission” to the government officials is obligatory The “commissions” are often calculated as percentage of the total cost of the project So, when helping the enterprises win the bid or win a large project bureaucrat will receive a “commission” increasing together Consequently, some projects are much larger and complex than necessary Moreover, a list over-cost material was supported to increase the expenditure of project From private benefit

of political agent, some resources such as human, land, technique were allocated inefficiently Consequence, some projects are completed but never used Some are low quality and they need more re-investment

In summary, the presence of corruption in previous study will make the cost in investment exaggerated, whereby that public expenditure in investment is higher than normal

is no exception; reducing the efficiency of public investment also

7

2.4 Empirical review

2.4.1 Public spending in investment and economic growth

Being developed from Barro (1990), the endogenous model was continued to use by Cashin (1995) and many theorists On the level of public investment, Cashin (1995) found the positive effect of public investment on the rate of economic growth with the test of data from

23 developed countries in period 1971-1988 Devarajan, Swaroop and Zou (1996) using data from 43 developing countries from 1970 through 1990 year to focus a rise in the level of current expenditure1 has positive and dramatically effect on growth, but the effect of public expenditure in capital component was negative on per-capital growth This result was the same with Ghosh and Gregoriou (2007) Being analyzed in an endogenous growth framework for the panel data of 15 developing countries over 28 years, using GMM techniques, current (capital) government spending has positive (negative) and significant effect on the growth rate However, Bose (2007) cited that: “The share of government capital expenditure in GDP in

positive and significant correlated with growth but current expenditure is insignificant”

On the other hand, about the component expenditure in public investment in the link with growth rate of the economy was also discussed by many previous researchers Easterly and Rebelo (1993) found that “investment in transport and communication is consistently correlated with growth”, but in the condition of budget constrain, Bose (2007) shows that:

“these do not survive”, also find the government expenditure and investment in education are significant associate with growth

To discuss that whether in the developing countries public investment is consider as an engine of growth or not? A country with a big public capital invest in infrastructure have positive effect on accelerating economic growth in low-income countries “if governments do not behave in the past” (Warner, 2014) In fact the need of infrastructure and relate capital in

1 This defined was classified by International (Monetary Fund IMF) based on the type and purpose of expenditure Capital expenditure used for durable goods (more than one year) Current expenditure includes wages, other goods or services, interest payments, and subsidies (see Devarajan, Swaroop and Zou (1996))

8

developing countries is great, so public investment was accounted for a large share of total investment in the majority of developing countries (Khan, 1996)

2.4.2 Corruption and public spending in investment

Corruption has been defined by many different ways The reason for that originated from the various in its appearance and impact Treisman (2000) cited corruption is self-interest

of public official in public sector Lamsdorff (2001) in the condition of corruption, the

decision of higher official will be distort, the same with Ditella (1999), Tanzi (1998), Van Ryckeghem and Weder (2001) This lead to misallocate resources such as human talent,

distribution public capital, the size and technology of project will be Corruption may lead bureaucrats using unproductive public expenditure, or offer them opportunities for rent-seeking

The government spending tends to items which are easy to levy The level and composition public investment activities will be over-estimated too For example, in case of Mauro (1998) government spending in education is found to be reduced Lucrative opportunities only come from markets difficult to monitor such as technology project or immature markets for instance infrastructure projects With Tanzi and Davooodi (2002b), they find the higher corruption associated with the higher public investment by using two kinds corruption index Business International (BI)2 and Political Risk Services, Inc Developing a theory game, Dartanto (2010) cites that by the rent-seeking behavior, the volume of public investment tend to be larger This study has built the corruption index on its own local government in Indonesia

2

Business international (BI) available for 68 countries over the 1980-1983 range from 1(most corrupt) to 10 (least corrupt) indicate “the degree to which business transaction involve corruption or questionable payments”, ICRG international country risk guide cover 1982-95

is available for 42-128 countries higher indicate “high government officials are likely to demand special payment” and “illegal payments are generally expected throughout lower level

of government” range from 0(most corrupt) to 6 (least corrupt) detail in Kneefer and knack (1993) (cited in Mauro (1995))

9

2.4.3 Corruption and growth

One hand, corruption has a positive impact on growth Getting bribery, officials will work faster and more efficient (Leff, 1964; Hungtinton, 1968), or with Rock and Bonnett (2004), when they investigate a positive effect of corruption in the new large East Asian economies Paul (2010) in the case of Bangladesh growth, there is a positive association between corruption and growth This research supposed corruption greases the wheels of commercial Bangladesh’s regulation heavy system and unleashed private investments, but this thing only happened when unconnected between in bureaucracy and major public utilities in Bangladesh case However, most of the earlier empirical studies before discovered that corruption impedes economic growth Mauro (1995) finds corruption has a negative effect to private investment and make economic grow down, the same with Li, Xu and You (2000); Ugur (2013) and Dridi (2013) In case of Pakistan, Farooq et al (2013) use the time series data over the period of 1987-2009 plus applied structural break unit test find that corruption has a negative effect to economic growth in the long run

Beside a substantial empirical work on the negative impact of corruption on economic growth, there is some studies has been research inverse effect (Shabbir and Anwar, 2007; Saha and Gounder, 2013) Taking to reverse side, Treisman (2000) states: “more developed economies, and (probably) higher import were less corrupt” Rauch and Evans (2000) finding less corruption in a good condition of promoting the productivity of state officials In the decentralization of public expenditure, corruption will be reduced following by Fisman and Gatti (2002) In the case of Vietnam, Bai et al (2013) with an empirical analysis uses survey data collected from over 13,000 Vietnamese firms between 2006 and 2010 shows that economic growth is predicted to decrease corruption

2.4.4 Private investment in the relationship with growth, corruption and public investment

The question is: Does public investment has more impact on the growth than private investment? The answer is not clear Khan (1996) also finds the answer this question, with a large sample of 95 developing countries over the period 1970-90, the finding is that there is a

10

distinction between them, and private investment has a stronger effect on the growth than public investment does This result is the same with Ghani and Din (2006) in the context of Pakistan economy

When corruption still exists, private investment was impacted in both quantity and quality Following O’toole and Tarp (2014) that bribery (one of the categories of corruption) has negative and statistically effect on efficiency of domestic SMEs investment, and when facing with the prospect of paying bribes, it is possible lead to reduce the level of private investment

Baliamoune-Lutz and Ndikumana, (2008) uses unbalanced panel data from 33 Africa countries for the period 1982-2001 to explore the impact of public and private investment under the control of corruption on growth This research shows corruption has a positive effect

on public investment while it has a negative effect on private investment On the view of negative affect, corruption discourages private investment by increasing the costs of doing business and instability in expected returns Corruption acts as a tax on capital, but this tax is uncertain This thing delay investment activities For example, one standard deviation decrease

in the corruption index raises private investment 2.5 percent points (Pellegrini and Gerlagh, 2004) (cited in Baliamoune-Lutz and Ndikumana, 2008)

Pall (2010) exclaims that corruption discourage private invest by raising indirect production cost In the condition of corruption, government tend public fund toward unproductive activities, have large investment infrastructure in order to get more embezzlement

The crowding in effect of public investment on private invest has been considered (Nguyen Hoang Bao, 2014) The high level of public investment might attract more private investment In fact, public investment creates conditions for profitable private investment such as: reduce cost, enhances the infrastructure quality, support an attractive the investment environment, build a high skill labor, etc

11

2.4.5 Foreign investment in the relationship with growth and corruption

Openness to international trade and their role was mentioned in many fields such as: a positive effect on growth through the result of learning by doing channel, specialization and spillovers Lucas, (1988), the scale effect (Rivera-Batis & Romer, 1991) Another aspect concerning about financial flow: foreign aid, especially foreign direct investment together with knowledge and tech transfer (Moran et al., 2005) (cited in Nguyen Hoang Bao, 2014)

In term of economic growth, Wacziarg (2001) investigate the link between trade policy and economic growth in a panel of 57 countries from 1970 to 1989 This study suggests a positive impact of the openness on economic growth In Viet Nam case, the openness also states its role in the economic growth For example, Su Dinh Thanh and Nguyen Minh Tien (2014), estimate panel data of 43 out of 63 provinces/cities in Viet Nam in the period 1997-

2012, they find that foreign direct investment has a positive impacts on the provincial economic in the long term Tam Bang Vu (2008) shows that FDI contribute to Gross domestic product growth through enhancing on labor productivity by regressing GDP on the robust OLS estimation (OLSR), Least square dummy variable (LSDVR) and FGLS method

To approaching on the corruption aspect, Ades and Di Tella (1999) find that countries get more competition from foreign firm, from abroad trade can reduce corruption Larrain B and Tavares (2004) show the share of FDI relative with GDP is significant associated with lower corruption Shabbir an Anwar, (2007); Gounder Rukmani and Saha, shrabani (2013) the openness or globalization can reduce corruption The interactive relationship between economic development and openness of India’s economy leads to a reduction in corruption

13

CHAPTER 3 METHODOLOGY AND DATA

This chapter presents the methodology and data characteristics The Cobb-Douglass production function has been used to find the contribution of public spending in economic growth This part also mentions data source, some typically characteristic of variables, model specification, and estimation method

3.1 METHODOLOGY

A productive government expenditure and non-productive government expenditure are being applied in the production function of final goods following (Barro, 1990; Futagami et al., 1993; Cashin, 1995; Glomm and Ravikuma, 1997; Ghosh and Roy, 2004) Haque M E and Kneller R., (2008) argued that non-productive expenditure has no effect on the investment decision, and the growth rate unless be financed by some form of distortionary taxation A simple Cobb-Douglass version has been used and be extent with non-productive government expenditure as exogenous input

Where 0<⍺<1, L is labor; K is capital stock, Y is total output, and G is total government spending In the Cobb- Douglas production function, production technology is constant with scale of labor and capital stock Technology change is treated as exogenous The Cobb-Douglas production function assumes that elasticity of substitution is equal to one With assumption that L is constant, if there is an increasing in K and G, the production function will have a fixed performance That has an endogenous growth in the economy

Accessing to the production function:

Where y= Y/L; k=K/L; A is represents a constant technology parameter; and Gγ is a form of non-rival, non-excludable public good

14

If β = (1- α) there is sustainable endogenous growth that is determined in part by public

investment; If β < (1- α) a version of the neoclassical model in which expenditure affect the level but not the growth of output (assuming exogenous technical process)

The β term, that is the elasticity of output with respect to public investment When the society has corruption, the return to public investment is equal to:

β = γ (1-η)

Where η is increasing in the level of corruption; If η=0, then β=γ the absence of corruption; If η=1, then β=0 at the highest level of corruption, the returns to investment are bound at zero Differentiating the production function with respect to time yields the following:

̇ ̇ ̇ ̇

The change in output is followed by the change of K and G dependent on and When having more corruption, the effect of G on Y will reduce

3.2 DATA SOURCES AND CHARACTERISTICS

3.2.1 DATA SOURCES

The panel data of 63 provinces in Vietnam for the period 2008-2013 was collected on Provincial Statistical Yearbook, General Statistics Office (GSO) of Vietnam, the Provincial Competitiveness Index (PCI)3, and the World Bank (WB) A full list of our data sources was showed in Appendix 1

The expenditure in Vietnam State’s budget is various in the composition In the final settlement of account they are: expenditure in development investment, current expenditure

3

The PCI index is based on the experiences of nearly 8.093 domestic enterprises (2013) about

a quality execution and business environment through at 63 provinces/cities of Vietnam and the estimation of nearly 1.609 foreign firms This survey was done by Chamber of Commerce and Industry of Vietnam (VCCI), with support from United States Agency International Development (USAID)

15

and the others The development investment programs might reduce private cost in terms of human resource, infrastructure and hence, create more productivity for economic development process Current expenditure together with the others tends to national defense and security, state apparatus operation, state debts settlement In this study, we approach the data of public expenditure in investment to answer for research question

The investment capital followed by the capital resource has been divided in three kinds such as: State, non-state, and foreign sector The state capital is expense to increase and remain physical assets, in a given period It was related with the big project of national program The data of foreign capital investment in this research indicates for the openness level of each province Whether province gets more attraction in foreign investment, there is a growing in openness In this study, we employ State investment capital is public investment capital variable; non-State investment capital is private investment variable in model

3.2.2 MEASURE GDP FOR EACH PROVINCE

Because of having the over-estimate in the GDP data of each province, we used a method to get the consistence between the GDP of provinces data and the GDP of Vietnam Firstly, we calculated:

Where

GDP: Gross Domestic Product of Vietnam in year t, be get from GSO

GDPit: the sum of Gross Domestic Product of 63 provinces in year t, get from the Provincial statistic Office

If I < 1: there is an over-estimation in the data of GDP of each province

Secondly, we have one formula in GDPi as follow:

16

Where : The GDP of each province after re-calculating, and this figure will be used to calculate the other index such as: GDP per capita, or public investment in a relative to GDP in this research

3.2.3 CHOICE OF CORRUPTION VARIABLE

Corruption has been defined in many different ways Economists generally see corruption as a part of the problem of rent seeking (Tanzi 1998, Shleifer and Vishny 1993, Mauro 1995) (cite on Kwabena et al., 2001) Most of the corruption practices are illegal and difficult to measure Following Tanzi (1998), the act of corruption can be divided into different categories For example, bureaucracy or “petty”, cost-reducing or benefit enhancing, briber-initiated or bribee-initiated, coercive or collusive, ect Corruption can be a gift form or bribes It can be an illegal income in trade (good or services) when official choose the best price-quality from the seller, and government budget be distorted, according to Ablo and Reinikka (1998) It is important to distinguish or define corruption

In this research, corruption has been seen in a special case, in any transactions involved

to public official Corruption in public sector by the local contains two elements: the exaggeration of the cost of public investment project and use of inferior materials (Haque M.E and Kneller R 2008) That corruption inflates the public investment and reduces its return in the economy These lead to lower in the economic growth Tanzi and Davoodi (2002b) discussed about corruption as a “commission” in public project bidding Thereby, corruption distorts decisions of higher officials Thus, how can this study choose the corruption variable for the case of 63 provinces/cities in Vietnam?

Take advantage of the annual Vietnam Provincial Competitiveness Index (PCI) survey, this research use the informal charge (“chi phi khong chinh thuc”) to measure corruption in 63 provinces This data get from Chamber of Commerce and Industry of Vietnam (VCCI) in period 2008-2013 This variable measures a type of bribery during business registration and licensing To do this, PCI survey employ the unmatched count technique (UTC) This technique has been used widely by a list question to answer so-called sensitive question (Ahart and Sackett 2004; Coutts and Jann 2011) (cited in Malesky E.J et al, 2015) It is measure on a

17

scale of 0-10 with higher scores indicating lower corruption The informal cost or “chi phi khong chinh thuc” was constructed form a survey of domestic and foreign firms located in each province by the questionnaire A part of them tends to indicate for: i) unofficial business payment as a percent of total revenue of each firm; ii) “yes or no” response in embezzlement

of bureaucrat; iii) An actual effect to firms from bribery

In previous research of Nguyen Quoc Viet and Chu Thi Nhuong (2012), they also employ “chi phi khong chinh thuc” to indicate for the corruption in provinces This study aims

to find the impact of corruption to scale and quality in public investment based on an institution economics perspective

3.2.4 ENDOGENOUS AND EXOGENOUS VARIABLES

Based on the theoretical framework above, we propose some dependent and independent variables which are capture the impact of corruption on public investment and economic growth The dependent variable is log of GDP per capita (lggdpper), but this is a simultaneous equation, so the dependent variable in this equation can be an explanatory variable in other equations Hence, we have four endogenous variables: log of GDP per capita, informal cost, log of public spending in investment (devex) and log of private investment capital (LgIp)

Following Haque M.E and Kneller R (2008), we draw some exogenous variables for this model including: log of foreign investment capital (LgIf), log of population (Lgpop) Besides, including capital state investment (LgIg) in the equation four, it is considered as an exogenous variable, and it has the crowding in effect on private investment

Furthermore, the rich country has low economic growth The poor country has high economic growth Thus the initial growth might impact on the economic growth (Log of initial GDP per capita - Ligdp) This variable also effect to corruption If country has a high initial

is as same as with public investment; can a risen in economic growth reduce corruption?

The simultaneous equation has been used by Tavares and Wacziarg (2001) They use it for analyzing the effects of democracy to economic growth Nguyen Hoang Bao (2014) also approaches the model of simultaneous equation system to explore the direct and indirect effects of factors on economic growth in Viet Nam Following his research, this model 1) Overcome the problem of biased estimation in the single equation approach of any kind because the single equation approach greatly exaggerates and distorts the impact of exogenous variables on the endogenous variables concerned 2) Explain the economic mechanism through which the independent variables might impact on endogenous variables directly and indirectly

Following Haque M.E and Kneller R (2008), we begin focus the change in the GDP per capita growth by the impact of informal cost, public investment, and other factors as follows:

Lggdpperit=η0+η1(informalcostit)+η2(devexit)+η3(LgIpit)+η4(LgIfit)+η5(Ligdpit)+u1,it(1)

informalcostit = β0 +β 1(gdppercapit) + β2(LgIfit)+ β3 (Ligdpit)+ u2,it (2)

devexit=α0+α1(informalcostit)+α2(LgIfit)+α3(Lgpopit)+α4(Ligdpit)+ u3,it (3)

LgIpit = γ0 +γ1(informalcostit)+ γ2(LgIgit) + γ3 (Lgpopit)+ γ4 (Ligdpit)+u4,it (4)

Where i refers to province and t refers to year; Lggdpper is log of GDP per capita; informalcost is the informal cost index; devex is log of government expenditure in investment;

4 If the national income increases, government expenditure increase, following Wargner (1983, 1912) (cited in Odihiambo, 2015)

19

LgIg is log of state investment capital; LgIp is log of non-state investment capital; LgIf is log

of foreign investment capital; Lgpop is log of provincial population; gdppercap is GDP per capita of each province, Ligdp is log of initial GDP per capita

In fact, there is the causality between corruption and growth; the impact of corruption

on the level and composition public spending; corruption and foreign investment capital; corruption and non-state capital investment (this discussion was presented on the literature review) Furthermore, the crowding in effect between non-state capital investment and state capital investment is also necessary be considered

The growth in GDP per capita is accessed as equation 1:

Lggdpperit=η0+η1(informalcostit)+η2(devexit)+η3(LgIpit)+η4(LgIfit)+η5(Ligdpit)+u1,it(1)

The informal cost (higher = better), public spending in investment (devex), Non-state investment capital (LgIp) and foreign investment capital (LgIf) also has a positive impact on the GDP This effect still mix when private sector or public has more effect than another Furthermore, the government expenditure has been become an important role in the developed

of infrastructure, the stability in the institution, and solid foundation to develop economic About the foreign factor in the economy, the transfer in knowledge or skill working, technology, as same as the flow of FDI fund has been promoted, leading to economic growth

In equation 2:

informalcostit = β0 +β 1(gdppercapitait) + β2(LgIfit)+ β3 (Ligdpit)+ u2,it (2)

Corruption is not only impact to growth, but also be impact of growth That mean a risen in economic growth (GDP per capita), corruption will reduce (or informal cost increase),

or public officials tend to more transparence and cleaner In addition, when having the foreign factor in the economy, it made the economy more open to integrate, and reduce corruption The public spending in investment is expressed by equation 3:

devexit=α0+α1(informalcostit)+α2(LgIfit)+α3(Lgpopit)+α4(Ligdpit)+ u3,it (3)

20

We expect the coefficient α1 is negative and positive for α2, α3 The public spending is depending on the level of development in GDP per capita, the size of population and foreign investment capital (LgIf), while the informal cost has a negative effect on the public spending because once get more clean in bureaucratic system , the room of corruption activity is not enlarge and get more difficulties to activated It could reduce the inefficient in the investment activities of government

On the level of public expenditure, Rodrik (1998) argues that openness to international trade has a positive effect (stated in Haque M.E and Kneller R 2008) On the other hand, the correlation of country size and government size is negative, this thing are same with open to trade (Alesina and Wacziarg, 1998)

The non-state investment is presented as follow, equation 4:

LgIpit = γ0 +γ1 (informalcostit) + γ2 (LgIgit) + γ3 (Lgpopit)+ γ4 (Ligdpit)+u4,it (4)

We expected the public investment (LgIg) and informal cost (informalcost) has the positive effect on the private investment because of the crowding in effect of public investment, also the expansion in the investor’s belief when corruption is lower The size of population (Lgpop) is expected to have a positive sign

3.4 ESTIMATION METHOD

The simultaneous equation system in this research, some equations contain endogenous variables among the explanatory variable such as: Lggdpper, devex, informalcost, and LgIp These endogenous variables are the dependent variables of other equation in the model for example devex and informal cost; LgIp and informal cost This disturbance is correlated with endogenous variables or the assumption of OLS was violated

We will exam the effects of corruption in the relationship between public spending in investment and growth by using Three stage least squares method (TSLS) (Zellner and Theil (1962) In previous study, Haque and Kneller (2008) supposed that TSLS is preferable to Two stage least squares (2SLS), when they find the role of corruption in the impact of public investment in economic growth

21

CHAPTER 4 ANALYSIS RESULTS

This part presents result of this research It includes in three parts Part one indicates some preliminary analysis about descriptive statistic and correlation coefficients Secondly, the identification problem is discussed Finally, that is empirical result

4.1 PRELIMINARY STATISTICS

4.1.1 DESCRIPTIVE STATISTIC

There is a difference among 63 provinces/cities of Vietnam on economic growth The reason for that may be come from geographical and natural features, human resources, local culture, especially from the difference in institution

In Table 4.1, we present descriptive statistics of the panel data available for 63 provinces/cities in Viet Nam for period from 2008 to 2013

In 2008, growth rate in GDP of Vietnam is 5.67% Up to 2010, this figure is 6.42% and 6.24% for 2011 Together with the cause of uncertainty event in the macro economics and a bubble real estate market, the growth rate has been dropped to 5.4% (source from GSO, 2008-2013)

The economic growth of 63 provinces has increased in this period The average of GDP per capita is about 21.8 million VND and the rate growth of this is 5.21% Ba Ria Vung Tau is province has the highest value of GDP per capita at 160 million VND in 2013, and some provinces has the smallest in GDP per capita such as Lao Cai, Ha Giang, and Lai Chau