Fiscal decentralization at subnational government in vietnam

3 How is the impact of decentralized expenditure and the degree of autonomy of sub - national government in general, that of provincial and district budget to governance 4 What are th

INTRODUCTION

1.Rationale of the research

Fiscal decentralization theory focus on explaining the impact of fiscal

decentralization and confirming that morefiscal decentralization given to the lowest

level of government will promote economic growth, decrease inequality through

efficient resource allocation and good governance However, meanwhile many

studies have concluded in favor of the positive impact of fiscal decentralization on

economic growth, studieson the impact of fiscal decentralization on governance

come up withvarious results

In Vietnam, moststudies on the topic mainly analyze the current status of the

contents of fiscal decentralization in order to propose solutions to improve the fiscal

decentralization system as perspecific period of time Besides, studies on the impact

of fiscal decentralization center around on the impact of fiscal decentralization to

economic growth and poverty reduction, yet there is few research on the impact

fiscal decentralization to governance

Based on theoretical and practical context of fiscaldecentralization in

Vietnam, authors choose the theme "Fiscal decentralization at subnational

government in Vietnam" as the subject of his doctoral thesis

3 Aim of the research

The research aims to propose policy recommendations to fiscal

decentralization at sub - national government level of Vietnam As such, the thesis

will answer the following questions: (1) The principle of deciding expenditure

responsibilities and revenue among levels of government (2) Current status of

decentralized budget for sub - national governments in Vietnam? (3) How is the

impact of decentralized expenditure and the degree of autonomy of sub - national

government in general, that of provincial and district budget to governance (4)

What are the solutions to improve fiscal decentralization system for sub - national

government in Vietnam that helps to promote efficient allocation of public

resources, ensure equity and improve quality of governance?

4 Object and Scope of the research

The object of the research is the content of fiscal decentralization and the impact of fiscal decentralization to sub- national governance in 62 provinces during the period from 2004 to 2012

5 Methodology

The thesis analyzes related regimes and policies to clarify the characteristics, advantages and disadvantages of fiscal decentralization at sub-national government

in Vietnam

Thesis using panel data regression (OLS, RE, FE) and dynamic panel data regression (system GMM) to assess the impact of fiscal decentralization at sub-national governmentto governance in Vietnam

In combination with the result of quantitative analysis of the impact of fiscal decentralization to local governance, the thesis will use them as the basis to give recommendations on the policy of fiscal decentralization in Vietnam

6 Major findings

The thesis sucessfully clarifies the impact of fiscal decentralization at sub- national government to governace as in the case of three – levels sub-national government: (1) Confirm that the contents of fiscal decentralization has different effects on each aspect of local governance (2) support fiscal decentralization theory : increased decentralization to lower government levelregarding providing goods and services will help to allocate resources more efficiently even in case of 3- level sub-nation government (3) Find out that outcomes of the effect of fiscal decentralization on local governance depend on the management of decentralized budgetfor each expenditure responsibility, the ability of upper government to control overlower government and capacities of the decentralized government

In practice, the thesis has contributed some following new proposals: (1) Decentralizethe providing of public goods and services of local to district authorities does not require the advantage of scale Transfer spending tasks that district governmentimplements inefficiently to the provincial government Distribute expenditure responsibilities in detail for sub - nation government (2) Adjustthe method of sharing value-added tax (VAT), corporate income tax (CIT) between central budget and local budgets Change royalties and taxes for

environmental protection to shared revenues among sub - government levels

Develop a compulsory list of revenue that sub - national governments must comply

and an open list of revenue for local government to choose from and decide which

revenues or tax rates to collect (3) Determine the scope of sub - national

government’s debt and debt limits should be built based on ability to pay debts

7 Structure of the Thesis

The thesis is structured into 5 main chapters together with the introduction

and conclusion part, including: (1) Chapter 1: Research topic overview (2) Chapter

2: Theory of Fiscal decentralization (3) Chapter3: Current situation of fiscal

decentralization in Vietnam (4) Chapter 4: Empirical Analysis of the impact of

fiscal decentralization at sub- nationalgovernment to governance in Vietnam (5)

Chapter5: Recommendations on fiscal decentralization atsub- national government

in Vietnam

CHAPTER 1: RESEARCH OVERVIEW

1.1 Internation researches on fiscal decentralization

1.1.1.Fiscal decentralization and social welfare, macreeconomic stability,

governance

Overall, fiscal decentralization is supportedas it can improve the

performance of the public sector, enhance social and economic welfare At first,

studies focused mainly on the theoretical aspects of economic growth when

implementing fiscal decentralization as in the researches of Kim (1995), Davoodi,

Xie and Zou (1995), Martinez-Vazquez & McNab (1998, 2001a, 2003), Zhang &

Zou (1997, 1998), Lin and Liu (2000), McNab (2001), Desai (2003), Lee (2003)

Although findings are inconsistent at some points but most studies show that fiscal

decentralization makes positive impact to economic growth [6] Secondly, fical

decentralization assists macroeconomic stability by reducing information costs,

operating costs in providing services, and promote the development of private

sector if fiscal discipline is strictly followed (Shah, 2006a) Thirdly, fiscal

decentralization has a positive influence on local governance As such, the studies

focused on the relationship between fiscal decentralization and corruption (Fishman

and Gatti, 2000; Gurgur and Shah, 2002), or regardedfiscal decentralization as an

indicator of public administration management from the aspect of public participation, and transparency of the provisions of law (Huther and Shah, 1998; DeMello and bärenstein, 2001) The conclusions on this relationship is quite diverse

There are also many studies suggest that fiscal decentralization may harm economic efficiency, economic stability and good governance According Prud 'homme (1994), in many cases, fiscal decentralization is merely a process of extending democracy in politics, but it does not help to improve economic efficiency Musgrave (1959) believes that it would be very difficult to enforce a uniform fiscal policy in the context of strong fiscal decentralization Tanzi (2005) found a correlation between decentralization and corruption

1.1.2 Empirical research on fiscal decentralization and governance

Reseach using multinational data:(1) De Mello and Bernstein’ research

(2001) demonstrated the impact of fiscal decentralization to governance (2) Study

of Fisman and Gatti (2002) applied morecontrolling variables thanDe Mello and Bernstein(2001) shows the decentralization of expenditure responsibility has a

strong impact on curbing corruption (3)Estache and Sinha (1995) also said that

decentralization of expenditures increased infrastructure investment at sub national

government (4) Study of Khaleghian (2003) proved fiscal decentralization strongly impact the results ofprevention in low-income countries (5)Treisman (2002) added

variable of autonomy to measure fiscal decentralization and came to conclusion that fiscal decentralization had a positive relationship with infrastructure and inversely

with education and health (6) The study of Huther and Shah (1998) shows that

fiscal decentralization madepositive impact on the quality of state governance - as measured by: the participation of citizens, accountability, social justice, improving

economic management and reducing corruption (7)Dreher (2006) estimates the

impact of fiscal decentralization to the quality of governance using control variables such as GDP, population, and intellecture This study concluded that fiscal decentralization has strong impact in low-income countries but there is no clear

evidence in the high-income countries (8)Research of Andreas P Kyriacou (2001)

extended measuring governance quality by adding such variables as corruption,

administrative quality and more control variables, which also show a positive

impact of fiscal decentralization to governance (9) The research of Adam et al

(2008) which uses two-step approach also shows a positive impact of fiscal

decentralization on the efficiency of governance

The studies in certain country : (1) Lorenzo Boetti, Massimiliano Piacenza

and Gilberto Turati’ research (2009) using data of 262 communes of Turin (Italy)

draws following conclusions: (i) the more autonomy over budget the less inefficient

the spendings are (ii) strict regulations on revenue and expenditure lead to increase

in spending efficiency (iii) Accountability in implementing the budget has a major

impact on the efficiency of sub national government’s spending (2) Research by

Adam B Elhiraika (2007) in South Africa, using data of 9 provinces during the

period of 1996-2005, shows that decentralization of revenue have different impacts

on education and health

1.2 Research on fiscal decentralization in Vietnam

Studies on the topic in Vietnam normally analyze the strengths and

weaknesses of the fiscal decentralization system at certain content and propose

policy recommendations to improve the system: Clay G Wescott (2003), report of

Adam Fforde (2003), the study of Jorge Matinez - Vazquez (2004), Le Chi Mai

(2006), Bui Thi Mai Hoai (2009), Vu ThanhTuAnh (2011), To ThienHien (2012),

Le ToanThang (2013) Other studies focus on evaluating the impact of fiscal

decentralization to economic growth, poverty reduction as in the study of:

LivBjornestad (2009), Nguyen Hoang Phuong (2008), Pham The Anh( 2008),

Nguyen Phi Lan (2009), Hoang ThiChinh Thon et al (2010), Mai Dinh Lam (2012)

These studies claim that more fiscal decentralization is given to sub - natioanl

governemt will promote economic growth and reduce poverty Regarding the

impact of fiscal decentralized to governance, according to the author's knowledge,

to date there has been only one study by Tran ThanhThuy (2012) examine the

impact of decentralization to good governance, in whichthe only variable related to

decentralization is used is the proportion of sub - national government’s revenue as

of total state budget revenues

1.3 "The blank" in research on fiscal decentralization Firstly, the preceding studies show trends of impact of fiscal

decentralization ongovernance is varied depending on the context of the study, thusit is impossible to applyextrapolate method for the impact of which for a

specific country Secondly, the studies show the impact of decentralization of

revenue and expenditure responsibility to governance follows same direction However, this can not be confirmed in the case that sub-national government is

organized into multiple levels Thirdly, due to the lack of research on the impact of

fiscal decentralization to governance, recommendations on fiscal decentralization policy is considered insufficient

CHƯƠNG 2: THEORY ON FISCAL DECENTRALIZATION 2.1 Nature, form, and content of fiscal decentralization

2.1.1 The nature of fiscal decentralization

Decentralization refers to the transfer of rights and responsibilities of central government to lower authorities Fiscal decentralization is the transfer of power and responsibility related to budget management of central government to sub - nation government Accordingly, the term fiscal decentralization emphasizes the allocation

of revenues, expenditures responsiblity and the division of duties and powers (including the power of management and decision-making) among levels of government in managing the budget

2.1.2 Forms of fiscal decentralization

From the perspective of authority, fiscal decentralization can be in the form

of: (1) Deconcentration: a form in which the administrative offices located in local area but they are branches of central government (2) Delegation: means the central

government transfer authority and responsibility of monitoring operating activities for sub - national government However, the sub - national government is not fully independent yet subjectedto the guidance and control of central government to

perform the tasks (3)Devolution: implies creatingan independent unit fromthe

central government

2.1.3 Contents of fiscal decentralization

Based on the relationship among levels of government, fiscal

decentralization comprises three following contents: (1) Determine the authority

and responsibilities of levels of authorities in issuing the policy, regulations,

standards and norms of state budget revenue and expenditure (2) Allocation of

revenues and expenditure responsibilitiesamong levels of authorities (3) Determine

the authority and responsibilities of levels of authorities in the budget cycle

Based on the activities of the state budget system, the contents of fiscal

decentralization include: (1) the decentralization of expenditure responsibilities; (2)

decentralization of budget revenues; (3) Transfer of funds between levels of

government; (4) Provisions on the possibility of borrowing and debt issuance of sub

-nation government

2.2.The views on fiscal decentralization

2.2.1 Musgrave– Oates’ theory on fiscal decentralization [67, 74, 83]

Richard Musgrave (1959) suggest that the government has three functions:

macroeconomic stabilization, redistribution of income, efficient allocation of

resources The division of functions of government that allows to narrow the scope

and determine of responsibilities of spending and revenue among authorities

Wallace Oates (1972) has developed this approach to fiscal decentralization theory

budget Highlights of Oates theory is thatfiscal decentralization increases social

welfare Musgrave – Oates’ theory of fiscal decentralization delimit the the

functions of government as follows: (1) Macroeconomic stability should be

assigned to the central government; (2) Re-distribution of income should be

allocated to central government; (3) efficient allocation of resources should be

allocated to sub-national government

2.2.2 Fiscal decentralization based on public choice theory

2.2.2.1Tiebout’s theory [2, 87, 97]

Charles M Tiebout (1956) found: due to lack of purchasing factor and

competitivenessthe public sector can not provide public goods at optimal level like

in the commodity markets As such Tiebout claims that if public good is offered by

local government, competition will arise and ensure supply at optimal level because the public can "vote by their feet" through moving to other locations Tiebout model helps to form the principles to orientate the assignment of public goods delivery among levels of authorities

2.2.2.2 Brennan-Buchanan’ theory [54, 84, 87]

Contrary to the theory on government of Musgrave- Oates, Brennan-Buchanan (1980) develop Leviathan government model assumingthat the organization is greedy and selfish In this model, maximizing the benefits can be achieved if local government isauthorized to own economic resources This model does not show the principle of allocating revenue and expenditure responsibility but give explanation on the impact of revenue and expenditure to improving efficiency and social welfare

2.3 Measuring fiscal decentralization

The indicators that are often used to measure fiscal decentralization including: the proportion of spending and income of sub - national government as

of total state budget or GDP Although widely used, but these indicators have limitations: they do not show the degree of autonomy of sub - national government

in implementing expenditure and revenue [12, 63, 79, 100]

2.4 Impact of fiscal decentralization to governance

2.4.1 Governance and measuring governance

Governance is a concept related to all aspects of the exercising power through formal and informal institutions to manage resources allocated to the state The quality of governance is determined by the impact of government activities on living quality of people (Huther and Shah, 1998)

In Vietnam, the Provincial Competitiveness Index (PCI) is popular, reflecting the quality of governance CPI is measured by 10 indicators [41] Another indicator hasjust been used to evaluate the effectiveness of governance and public administration at provincial level in Vietnam is PAPI The index is developed from the view of citizen on effective governance and public administration of the provinces [42]

2.4.2 Positive and negative impact of fiscal decentralization to governance

2.4.2.1 Positive impact of fiscal decentralization to governance

First, sub nation goverment has the advantage of meeting the needs of the

people, especially when the supply of public goods and services funded by taxes

collected from local people (Oates, 1999) Second, decentralization improves the

behavior and attitudes of the local politicians (Bardhan and Mookherjee, 2006)

Third, fiscal decentralization promote participation, social justice, improving

economic management, increase transparency, reduce corruption, institutional

quality and governance quality (Huther and Shah 1998; Kyriacou and Sagales,

2009; Mello, L., 2011) Fourth, fiscal decentralization is a condition for

sub-national government to make effort, even break through in implementation of

central policies to achieve their goals

2.4.2.2 Negative impact of fiscal decentralization to governance

First, tax competition can drive tax, or public spending below the social

optimum (Oates, 1999; Cai and Treisman, 2005) [69] Second, Oates (1985)

suggest that more fiscal decentralization will lead to budget losses due to increasing

the size and cost of government administration Third, corruption may increase

when implementing fiscal decentralization, especially for developing countries

(Tanzi, 1995b)

CHAPTER 3: CURRENT SITUATION OF FISCAL

DECENTRALIZATION IN VIETNAM

3.1 Economic context and budget revenue and expenditure for the period of

2004-2013

The growth rate of GDP during the period of 2004 - 2013 reached 6.23%

Inflation rose: in 2010, the inflation rate was 11.75%, 18.13%, and 6.81% in 2011,

2012 and 2013 respectively.Regarding state budget revenue, the average size of

2004-2013 period reached 24.9% as of GDP Regarding expenditure, average total

expenditures per period 2004-2013 reached 29.0% of GDP Regarding the structure,

state budget expenditures have changed towards gradually reducing the size of

development investment expenditure, while increasing recurrent expenditure

3.2 Situation of fiscal decentralization at sub - nation government in Vietnam

3.2.1 Decentralization of expenditure responsiblities for sub- national government

3.2.1.1 Scope of expenditure responsibility of local budget

Scope of expenditure responsibility of local budget:

Local budget’s primary responsibilities tend to increase in the provision of basic public services Considering the proportion of recurrent expenditure by sector

as of total local budgets also showed that education, health, economics and business administration accounted the largest propotion

Decentralization of expenditure responsibilities among provincial, district and commune level

Fiscal decentralization among levels of sub-nation government has some

following features: First, decentralization to district and commune levels is

limited: Although the level of the budget decentralized to the district level tend to

rise but expenditure spent by district budget accounts for only almost 40% of total local budget, while that of provincial budget still accounts for over 60% of local budgets Decentralization of expenditure responsibilities among different levels of

local budget are as follows: (1) decentralization of investment expenditure funded byin balanced investment capital : in the poor provinces, this responsibility is mainly allocated to provincial government; in those of higher economic development level

the responsibility is assingned clearly and in detail for 3 levels (2) decentralization

of investment expenditure funded by targeted capital is normally assigned to

provincial level to support general program (3) decentralization of recurrent expenditure for healthcare : 15 provinces allocate this responsibility for the provincial budget, and not for district and commune budget; 16 others decentralized

to the provincial and district; 26 decentralizefor all three levels, and 6 allocate to

provinces and communes, but not for districts (4) decentralization of recurrent expenditure for education and training : the majority of localsallocate the responsibility of expenditure forsecondary schools; elementary schools, preschools,

and cultural enrichment to district level High schools and regular education are

allocated to the provincial level There are many other locals sharing the

responsibility between the provincial and district levels for general education and

elementary enrichment In particular, Tra Vinh is the local that does not

decentralizethe responsibility to district and commune level Expenditure for

training is allocated to provincial level; vocational training centers and other forms

of training, short-term training are allocated to district level, commune is not

decentralized for these tasks Second, the transparency and clarity in allocating

expenditure responsibilities remain low The unclear delineation of expenditure

responsibilities among budget levels reduces spending efficiency, especially when

different levels of sub national government must work together in a jointly task

3.2.1.2 Autonomy of sub - national governments in spending management

and decision making

Autonomy ofsub-national governmentsfeatured by: (1)sub-national

government has authority to issue regulations and policies on expenditure yet only

responsible for enforcement and compliance (2)In terms of recurrent expenditure,

sub national governments are active in budget planning, implementing, but still

dependent on their upper authorities (3) For developmentexpenditure, the local is

activein allocation of capital plan and assign capital to each project

Decentralization of spending responsibility at all levels of sub-national

governmentis facing some problems: (1) there is a mismatch between spending and

revenue This is a consequence of the increased decentralization of spending tasks

while revenues remain centralized (2) Regional inequality or horizontal fiscal

imbalances is high.(3) Lack of clarity and overlapping in the allocation of

expenditure responsibilities among provincial level, district level and communes

That an expenditure responsibility is assigned simultaneously for all three levels is

common (4) Decentralization of some expenditure responsibilities is considered

inefficient: such expenditures as social sercurity, price subsidies, population and

family planning, training (5) Lack of mechanisms to ensure efficient allocation of

budget among sub national governments

3.2.2 Decentralization of revenue for sub - national government

3.2.2.1 Decentralization of revenue for sub national government

Allocation of revenue for local budget

Revenue that local budget has 100% entitlement:(1)Revenues from housing

and land : total revenue from land as of total local revenues as decentralized experienced a decline from 26.7% in 2004 to 19.2% in 2012 due to dependence on land use fee revenues This revenue is allocated to threelevels of sub - national government: provinces, districts and communes but more is allocated to the

provincial and district budgets (2)Registration fees: accounted for an everage of

about 4% local revenue, are often decentralized 100% for district budget, in which

the land registration fee is usually delegated to the commune budget (3)Excise:

proportion of this revenue is less than 1% of local revenues: communal budget entitle to excise paid by individuals, households; district and provincial levels

enjoyed the rest (4)Fees and charges: there are more than 290 types of fees and

charges with revenues account up to more than 4% of total local revenues and are decentralization among levels of sub national governmentin form of which level collect the revenue will enjoy that

Allocation of revenue among provincial, district and communal budget

In terms of scale, locals tend to increase revenues for the provincial budget and reducing the district budget revenues by reducing revenues that provincial and district budgets enjoy 100% entitlement, increasingshared revenues between the provincial budget and district budgets The localities have high levels of autonomy over budget tend to focus local revenues for the provincial government, while those

of low level of autonomy over budget and are dependent on central budget tend to decentralize more revenue for lower authority

As per assessing the degree of autonomy over revenue, it shows that the degree has decreased markedly over the years The degree of autonomy over revenueof provincial budget tends to improve while that of district budget tend to decrease over the years Even for locality of which district budget of highest

autonomy level over revenue can guarantee only about 60% of district expenditure

like in Ben Tre, Tay Ninh, Binh Duong, Da Nang

3.2.2.2 Decision making of sub - national government over revenues

Similar to decentralization of expenditure, in principle levels of sub -

national government do not have the authority to enact policy on collection of

revenue which is different from those enacted by central government But in fact,

local government may affect a certain number of revenues Therefore, in the future,

it suggested that more power to decide on revenue should be given to sub - nation

government by allowing sub - nation government to decide the tax rate within the

frame specified by the central government

3.2.3 Transfer forsub - nation government

The scale of transfer for sub-national government tends to rise over the

years, reflecting the target of support for local to ensure the implementation of the

tasks assigned Similar to local budgets, the district budget and commune budget

depend mainly on balancing transfer from the provincial budget and tend to rise in

recent years Current transfer system reveals some issues to consider: (1) the district

and commune levels are too dependent on transfer (2) The system of criteria and

norms for the current budget allocation is still based on the targets and inputs

(3)Poor localities heavily rely on targeted transfer to meet the investment needs in

the area

3.2.4 Local debt

Borrowing of sub-national government shows some problems as follows: (1)

existing mechanisms of asking and giving in borrowing In many cases, the use of

debt is for debt repayment of outstanding capital investment (2) The scope of debt

does not cover all debt obligations of the local (3) Unclear regulation about who is

the ultimate person responsible for the debts (4) Many local established investment

development funds in order to mobilize resources for development investment (5)

According to law, source of repayment of principal and interest is guaranteed from

the provincial budget, which creates a risk of future deficits by spending part of the

current budget to pay debt

3.3 General assessment of fiscal decentralization in Vietnam Fiscal decentralization of local budget has the following advantages: First, fiscal decentralization for sub-national government is getting stronger Second,

provincial government has been granted greater autonomy over managing local

budget Third, the assignment of expenditure responsibilities, especially the duty to

provide public services to local budgets of all levels have ensured the conformity of

the "scope of benefits" of that provided by sub-national government Fourth,

decentralization of local revenues for all sub-national government levels is relatively consistent with the characteristics and nature of revenues in terms of

efficiency standards and fairness Fifth, transfer among budget levels helps to improve vertically and horizontally budget imbalance Sixth,fiscal decentralization

has strengthened fiscal discipline

The limitations of fiscal decentralization: First, vertically and horizontally fiscal imbalance happen at all budget levels Second,decentralization of expenditure responsibility is still overlapping, lack of transparency and ineffective Third,the

allocation of some revenue to local budgets is not suitable to the characteristics and

nature of revenues Fourth, transferamong local budgets of all levels increased dependence on the upper government’s budget Fifth, the scope of debt has not

covered all debt repayment obligations of sub-national government although borrowing limits are established but not yet linked to fiscal discipline of local budgets

CHAPTER 4: EMPIRICAL ANALYSIS OF THE IMPACTS OF

ATSUB-NATIONAL GOVERNMENT IN VIETNAM

4.1 Empirical hypotheses

The theory of fiscal decentralization showed that assignment of expenditure responsibilities for sub-national government will improve the effectiveness of public good and service delivery, improve transparency, reduce corruption and

transaction cost Therefore, the first hypothesis (1a) is that allocation of

expenditure responsibilities for province and lower levels has positive impact on the effectiveness of public good and service delivery ; the second hypothesis (1b):

allocation of expenditure responsibilities for province and lower levels has negative

impact on the institutional performance of provinces ; and the third (1c): allocation

of expenditure responsibilities for province and lower levels has negative impact

on the efficiency of public administration of provinces

Increasing revenue autonomy at province and local levels in order to

fundexpenditure responsibilities will create the motivation for the province and

lower government to distribute effectively the public service, to save administrative

cost and to reduce corruption (Oates, 1999; Fisman and Gatti, 2002) However, in

the perspective of developing countries, fiscal decentralization has negative impact

and destroy the monitoring and controlling of central government over province

goverment resulting in the increase of corruption (Tanzi,1995b) In particularly, the

motivation of corruption will be higher in poorer areas Therefore, the next

hyphothesis (2a): fiscal automomy has positive impact on delivery of public service

of provinces ; hyphothesis (2b): fiscal automomy has positive impact on

institutional performance of provinces ; hyphothesis (2c): fiscal automomy has

positive impact on efficiency of public adminisstration of provinces

4.2 Data and variables

4.2.1 Data

The data for fiscal decentralization is budget disclosure of final account of 62

provinces from 2004 to 2012 The data is available in ministry of finance website

GDP, private capital, population and infant mortality rate by province are

from the data disclosure of General Statistics Office of Viet Nam and from 2004 -

2012 The population and infant mortality rate are in 2005 and from 2007-2012

Provincial governance indicators are from Provincial Competitive Index

(PCI Vietnam) from 2006 -2012 PCI Vietnam is built from the result of survey of

approximate 10000 companies and enterprises

4.2.2 Variables of fiscal decentralization at province and lower levels

As per IMF’s measure of fiscal decentralization, we use the proportion of

sub-national expenditure, provincial expenditure, and district expenditure in term of

GDP to reflect fiscal decentralization according to province and lower levels We also use the rate of sub- national decentralized revenue by the sub- national recurrent expenditure, the rate of provincial decentralized revenue in term of provincial current expenditure, the rate of district decentralized revenue by district expenditure to measure the revenue autonomy or the vertical fiscal imbalance of provinces and lower levels

4.2.3.Variables of governance of province or sub-national government

Labor training and infant mortality rate measure the effectiveness of public service delivery;the higher value of labor training is the better quality of public service supply; the lower value of infant mortality rate is the better quality of health service provision The institutional performance is measured by informal charges, transparency, land access and security of tenure The efficiency of public administration is measured by time costs All the above proxies are the indicators of PCI Vietnam from 2006-2012 The higher of proxy value is the higher quality of governance

4.2.4 Control variables

We use the provincial population variable to control the characteristics of population scale of provinces; the population data is the statistics data of job and population of General Statistics Office from 2005-2012 We also use the proactive indicator of PCI Vietnam as a controlvariable to reflect the difference of the creating and autonomy of province leaders According to FismanandGatti (2000),

we controlled the literacy of people in a province by the rate of literacy of the population above 15 years old We also used the variable of the rate of private capital by GDP in order to expressing the openness of provincial economy

4.3 Research methodology

4.3.1.Regression panel data:

In regression panel data, 3 models are used popularly as Pooled Ordinary Least Squares (POLS), Fixed Effect Model (FE), and Random Effect Model

According to Mello andBarenstein (2001), Kyriacou (2001) we started by a simple

regression model as follows:

ݕ௜,௧ = ߙ଴ + ߙଵܺ௜,௧ + ߙଶܼ௜,௧ + ݑ௜,௧ (4.1)

Variable ݕ௜,௧is quality of governance of province i;variableܺ௜,௧reflect fiscal

decentralization of province i; Variable ܼ௜,௧is controlled variables;ݑ௜,௧is errors

We use Lagrangian Multiplier - LM test to choose the suitable model

between POLS and RE (Breuschand Pagan, 1980) Then, if the result of LM test

rejects Ho, RE model will be chosen and we continued to use Hausman testfor

comparing between RE model and FE model (Hausman, 1978).The most suitable

model is tested by Wald test for heteroskedasticity in case of FE model and tested

by Wooldridge test for autocorrelation in panel data The model will be adjusted by

robust standard errors method if it got heteroskedasticity or autocorrelation

4.3.2.Regression dynamic panel data:

The disadvantage of the POLS, RE, FE regression models is not handled on

the endogenous.In order to solve this problem, Arellano and Bond (1991) proposed

to use the model of Difference Generalized Method of Moment (Difference

GMM).However, Blundell and Bond (1998) showed that the Difference GMM is

not effective and the instrument variables are not strong enough when dependent

variable got high correlation value between the current value and its value of

previous period and the period number is small.Blundell and Bond (1998) expanded

the model GMM (1991) by using 2 models at the same time: (i) the model uses the

lagged variables of the difference in order 1 of independent variables as instrument

variables; (ii) the difference model use instrument variables by the lagged variables

of explanation variables (including the lagged variables of dependent

variables).Therefore, the model GMM (1998) or the model system GMM will be

used

We also use Sargan and Hansen test for the overidentification and the

Arellano - Bond test for the autocorrelation of the system GMM

4.4 Empirical results

4.4.1 Descriptive statistics

Descriptive statistics data showed that: sub-national expenditure, provincial expenditure, district expenditure account for 32%, 15%, and 13% asof provincial GDP respectively.Vertical fiscal imbalance reflects the revenue autonomy and the vertical fiscal imbalance at sub - nation, province and district level is 0.92, 0.58 and 0.3 respectively These figures show that the central government strongly transfers their revenue to sub - national governments but the transfer from province to lower levels is quite lower

Exception of the correlation of fiscal decentralized variables, the correlation coefficients between independent variables are quite low (< 0.5) and the regression model will not get the multicollinearity However, the correlation coefficients of fiscal decentralized variablesare high (>0.8) shows the multicolinearitywill happen

in regression model of these variables In order to solve this problem, we will explore the impact of each fiscal decentralized variable on governance variables In addition, we use the variance inflation factor test for the multicollinearity The results are smaller than 10 reflecting that the multicollinearitywill not happen in the using data (Kennedy, 1992)

4.4.2 Regression results

4.4.2.1 The impact of fiscal decentralization on public service delivery

The impact on Labor Training: (1) decentralization of revenue and

expenditure of sub- national government, revenue decentralization at district level, the active of province governmenthave positive impact on the quality of labor training service; (2) population does not clearly impact on labor training service; (3)

the rest variables have no impact on labor training service.The impact on Health Service: (1)decentralization of revenue and expenditure of sub- national

government, expenditure decentralization at district level have negative impact on infant mortality rate showed the positive impact of fiscal decentralization on the quality of health service; (2) the dynamic model confirmed the impact of literacy

variable on governance variables

4.4.2.2 The impact of fiscal decentralization on the institutional performance

The impact on Informal Charges: the decentralization of revenue and

expenditure of sub- national government have negative impact on the informal

charges However, the dynamic model has not showed the evidence of the impact of

provincial and district fiscal decentralization on the informal charges as the results

of OLS, RE, FE models In addition, the dynamic model also confirms the negative

impact of private capital on informal charges This result explained for the changes

of the informal charge indicator from the demand side: the more private capital

invests, the more opportunity of informal charge happens The impact on Land

Access and Security of Tenure: the fiscal decentralization of sub- national

government, the expenditure decentralization at provincial level and population has

negative impact on the indicator of Land Access and Security of Tenure; the

proactive of province has positive impact on this indicator The dynamic model

showed the expenditure decentralization at district levels and private capital has

negative impact on the indicator of Land Access and Security of Tenure The

impact on the transparency: the revenue decentralization of sub- national

government and at district level has positive impact on the transparency The results

confirm that the sub-national government and lower levels assigned revenue sources

will stimulate sub-national government to improve the transparency in their

activities

4.4.2.2 The impact of fiscal decentralization on the efficiency of public

administration

The active of the government and the literacy have positive impact on the

time costs However, the OLS model showed the provincial expenditure

decentralization has negative impact and the provincial revenue decentralization has

positive impact on the time cost; but the revenue decentralization at levels have

positive impact on the time cost in the dynamic model

4.5 The results of empirical analysis

4.5.1 Conclusions of empirical analysis

The first, expenditure decentralization has positive impact on public service

delivery of sub-national government However, the impact of district expenditure decentralization on labor training service is different from the impact on health

service.The second, expenditure decentralization at sub-national levels have

negative impacts on the institutional performance that proxies by variables of Informal Charge and Land Access and Securityof Tenure; in the meantime, there is

no evidence for the its impact on the variable of transparency The third, the fiscal

autonomy has positive impact on public service delivery; it has negative impact on variables of the Informal Charges and Land Access and Securityof Tenure, but it

has positive impact on the variable of the transparency.The fourth, the quality of

the research will be improved when the governance indicators are measured from the citizen's view and future researches are suggested to solve this implication

4.5.2 The contributions of fiscal decentralization literature

The research result contributes to clarify the impact of fiscal decentralization

in case of the sub-national government including 3 levels as follows:

(1) Finds the contents of fiscal decentralization have the different impacts on the aspects of the governance of sub-national government

(2) Confirms the effectiveness of resource allocation of sub-national government is improved when we assign public service responsibilities for lower levels in case of the sub-national government including 3 levels

(3) Finds the results of the impact of fiscal decentralization on the governance are dependent on which expenditure responsibility is assign for sub- national levels, on the controls of the superior or higher level government, and the capacity of decentralized government

4.5.3 The policy implication for fiscal decentralization in Vietnam

(1) Transfers more expenditure assignment for lower government in public service delivery (2) Improves the vertical fiscal imbalance of levels of sub- national