Background Of Decentralization In Cambodia

Published Date: 02 Nov 2017

CHAPTER I

INTRODUCTION

Improving the quality of life of people is the overarching task of all governments of whatever types. It cannot be achieved once and for all, requiring constant attention and efforts. To this end, mustering necessary resources is a key challenge especially for the sub-national governments. While decentralization supposedly brings more resources to the sub-national governments, it is not still clear enough, whether it leads to a better quality of life at the local level. This thesis will address this crucial issue in the context of Cambodia’s decentralization and her efforts to economically develop.

Many countries around the world now have implemented a decentralization policy to delegate more service responsibilities and authority to raise resources from the national level to the sub-national levels with the expectation that resources made available to the sub-national jurisdictions will be used in effective and efficient ways to increase the local economic development. According to Oates (1993), decentralization is an effective means which has contributed a lot to the increase of economic efficiency in the provision of public goods and services. Local governments will provide better services and developmental efforts to the preferences of their residents than the central ones, because they are closer to the people and their services are purportedly based on the local needs, situation and geographic differences. He also pointed out that services provided by local governments include local infrastructure, education, and environment.

Decentralization has been appeared after the economy of one country has been growing high; that is, when the economic growth of one country goes up at a significant level, the government will consider how to allocate the revenue earned effectively to ensure that the service deliveries will be close to the local people and be responsive on time to the needs of local people (Balh & Linn, 1992). No longer that the central government is able to troubleshoot with economic issues alone, particularly the developing countries in Asia. The economic planning, which only stipulates by the central government, will not effectively respond to the world various economic problems which has required the involvement of local government (Smoke, 2001).

Whereas, decentralization reform in Cambodia, based on the Administrative Management for Capital, Province, Municipality, District and Khan (Organic Law), means that the central government has given the ownership to the councils of local governments on function, power and resources to be responsive the needs of local residents, especially the councils must be accountable to their own local people. Royal Government of Cambodia (RGC) pursued decentralization policy is to promote sustainable development and improve local economic foundation and democracy at all local government levels through 1) priority needs of local people will be responsive through the elected council’s development planning 2) the service delivery from the public sector will be closer to the local residents and 3) resources and responsibilities allocated to local government will be channeled a reliable mechanism (RGC, 2010-2019).

Many other literatures have been studied on the link between decentralization with the economic growth/development. However, the results have been shown differently. Some have found the positive links between decentralization and economic growth (Akai & Sakata, 2000; Wall Oats, 1993; Iimi, 2005 and Stansel, 2005), while others found negative relations (Zhang & Zou, 1998; Davoodi & Zou, 1998 and Xie et al., 1999). They have their own reasons to defense their arguments.

One of the significant reasons causes to the empirically different result due to the different selection of proxy indicators of decentralization, unit of study and the different system in transferring budget. Most of studies have used unit of study at country, region, metropolitan and province level (Faridi, 2011). Therefore, there is not consistent result yet till now between the relations between this decentralization, we are still in doubt that whether the decentralization will have a positive influence on local economic development.

Decentralization reform in Cambodia has officially started since 2002 at CS level, which is the lowest tier of local government, and resources have been transferred from the central government to elected councils to enable the councils manage those resources with their local residents to develop their own territory. The budget has been directly transferred from the central governments to CS. The transferred budget for CS, known as Commune/Sankgat Fund (CSF), is divided into 2 components/transfer which are administrative and development component. Based on the article 12 of Sub-decree # 16 on the establishment of CSF in 2002, the administrative component should not be allocated exceed 1/3, while developmental component should not be allocated less than 2/3 of the total budget. The fiscal budget for CS which is allocated from central government is gradually increased. According to the Achievement Report of CS from 2007-2011, the budget allocated for CS is 1.5% of National Current Revenue (NCR) in 2002 and goes up to 2.8% of NCR in 2011. Given these intergovernmental transfers to the sub-national governments with the objectives of local capacity building and economic development, it may be the right time to ask what impact the decentralization reform efforts have on local economic development.

There are few studies showing the positive impact of decentralization in Cambodia on the local economic development and those studies were not done in an empirical study as many other scholars above did; that is, there is no strong empirical study to support that this decentralization reform has influenced to the local economic development/growth in Cambodia.

The objective of my study paper at this time will examine the links between decentralization and local economic development in the context of Cambodia to inform the policy maker whether the decentralization will contribute to local economic development, and provide policy recommendation to improve the decentralization reform in Cambodia.

CHAPTER II

BACKGROUND OF DECENTRALIZATION IN CAMBODIA

Based on the national constitution of Cambodia, Chapter 11, article 145-146, The Kingdom of Cambodia is split into the capital city and 23 provinces. The capital city consists of Khans, and a Khan, of Sangkats. The province is composed of municipalities and districts (Srok). A municipality is divided into Sangkats. There are communes and Sangkats (CSs) under districts. The management and administration of these sub-national legal entities are stipulated in the Organic Law. Thus, communes and Sangkats are the lowest administrative arm of RGC.

Table 2.1 Cambodian Territory

Central Government

Provinces

Capital

Municipality

District

Khan

Sangkats

Communes

Sangkats

Sangkats

Source: Law on Administration and Management of Capital Province Municipality and District

2.1 Decentralization before the Civil War

Decentralization in Cambodia has a long history dating back to the French colony period (from 1853 to 1953). Based on the Royal decree, known as a rule or regulation issued by King in 1925, the commune was established as a legal entity and commune chiefs were elected with council in every four years. Commune councils had the responsibilities to manage its administration and finance to implement its work (Cheng, 2011).

After getting independence from France, Cambodia fell into Civil War again in 1970 till 1991. During the civil war, local governments had no clear roles and responsibilities in terms of contribution to poverty reduction and economic development. Their tasks mainly focus on the security and soldier selection (Cheng, 2011). In 1991, all Cambodian conflict parties had agreed to the cease-fire and signed for a peace accord. All parties agreed to have a national election with support from the United Nations in 1993, except Khmer Rouge Party, later on became the rebel of the country. In 1993, a new regime was established, which is the Kingdom of Cambodia. It has a monarchy ruling according to the constitution. Members of the Parliament began being elected, and the first legislative RGC (1993-1998) was also instituted. The priority task of the RGC at that time was to rebuild and rehabilitate the war-stricken national infrastructure to grow the economy and reduce poverty. With this first-term of new RGC, the centralized system was dominant; all decisions were made by the central government even the local development. However, later on, there was a pilot project in terms of decentralization in some provinces for local economic development and poverty reduction (see detail below at Decentralization after the civil war)

2.2 Decentralization Effort after the Civil War

2.2.1 Decentralization at Communes and Sangkats

According to the Andersen (2004), there was a pilot project called Cambodia Reintegration and Rehabilitation (CARERE), Phase I from 1994-1995 to promote the living standard of the Cambodian poor people in rural areas. The province, which is one of the sub-national governmental tiers, was selected to involve in the project to support to communes to promote local good governance at the local community through a participative mechanism in financing and planning. The implementation of this project was regarded as the re-commencement of decentralization in Cambodia.

Andersen (2004) continued that with the implementation experience of CARERE phase I, RGC established the Seila Program in 1996 which was to be supported by CARERE Phase II from 1996 through 2000. This program was executed to stimulate decentralization and de-concentration reforms in Cambodia; it was operated as a mechanism to effectively distribute and bring together the aid from development partners. The Seila Program’s main goal was focusing on improving good governance and poverty reduction. To achieve this goal, it has come up with an implement strategy which includes 1) increases the development in the rural areas through the administrative management improvement by decentralization and de-concentration at provinces and communes and 2) promotes the public participation to ensure that the priority needs of local people will satisfy and to effectively mobilize government budgets and grants for local development at the commune and province levels. It means that budgeting, financing and planning will be structured in an effective harmonized way to avoid duplication of developmental efforts.

There was a significant change in the decentralization efforts in 2001, which fell on the period of implementing Phase II of Seila Program led by Seila Task Force. The National Parliament of Cambodia (NPC) has adopted the LAMCS in 2001. Based on LAMCS, the CS council election occurred, and those council persons were from different parties in Cambodia. The council members of CSs are elected by the local residents in the respective jurisdiction every five years. The first election was held in 2002. The council of a CS has two functions: 1) serving local affair. This function requires the CS council to work for the best interests for its own local residents; that is, the CS council is expected to spend its effort to promote economic status and the social welfare of their local residents. Moreover, it is also obliged to take care about security and public order in its jurisdiction. 2) serving as an agent on behalf of the central government, implementing delegated tasks. Also, the councils are required to prepare development plans with the resident involvement in the whole process. Such developmental plans need to ensure that the actual development projects respond to the local preferences. In Chapter 7 of LAMCS, CS must have their own financial resources and properties. The necessary financial resources are to be from national allocation budgets and from other resources which include tax revenue, non-tax revenue and other service fees.

Although the LAMCS stipulates that CSs can have the financial resources from taxes, they have obtained resources only from the national budget allocations (CSF) and contribution from local people and other NGOs. They are not able to collect tax yet. Regarding local service delivery, CSs are working now as delegates of the national government. Their main responsibility of service providing till now is just only the work of civil registration (birth, marriage and death) on behalf of the Ministry of Interior, and the voter registration for the National Election Committee (NEC). Other key functions are not yet decentralized to the local governments (Romeo & Spyckerelle, 2003). This practice still exists till now.

However, with the decentralization strategy, CSs are substantially autonomous; it can manage its resources with participation from the residents for local development. The budget allocation for the CSs has gradually increased since its initiation in 2002. Comparing to the national current revenue, the allocated budget for CSs has steadily moved from 1.5% in 2002 to 2.80% in 2011.

Table 2.2 Allocated Fund for Communes/Sangkats through CSF in USD

Year

Commune/Sangkat Fund

Ration to National Current Revenue

Administration

Development

Total

2002

3,462,043

4,209,935

7,671,978

1.50%

2003

4,847,199

8,007,683

12,854,882

2.00%

2004

4,833,333

9,666,667

14,500,000

2.50%

2005

5,178,000

11,347,000

16,525,000

2.52%

2006

6,161,300

13,292,200

19,453,500

2.54%

2007

8,128,500

14,259,000

22,387,500

2.56%

2008

8,209,055

15,047,195

23,256,250

2.70%

2009

8,581,025

19,753,975

28,335,000

2.75%

2010

9,657,310

27,486,190

37,143,500

2.80%

2011

13,150,500

26,699,500

39,850,000

2.80%

Total

72,208,265

149,769,345

221,977,610

Source: Annex of Achievement Report of Commune/Sangkat Councils 2007-2011

Simultaneously, given that the newly independent administrative institution lacks capacity and experience, a supporting mechanism has been launched to help CSs councils to ensure that they comply with the law and regulations. The support provided is not only from central government, but also from the other tiers of sub-national government such as the capital city, the provinces, municipalities, Khan, and districts.

According to the Sub-decree #58 ANK by RGC, 2001 (known as an Anukret, which is a sort of rule or regulation issued by the Prime Minister), the new Department of Local Administration (DoLA) has been established under the supervision of the Ministry of Interior to support the CS work implementation. Following the Anukret #58, the Ministry of Interior has issued a declaration known as a Prakas, which is a rule or regulation issued by a Minister, 2002, the Structure, Roles and Responsibilities of DoLA. DoLA has roles and responsibilities to monitor and evaluate CS’s work implementation to ensure the legality of CSs’ operations, especially to make sure that the decentralization policy at CS is fully applied. Also, the department has a role as the mediator between CSs with other agents of RGC and civil societies like NGOs, and coordinates conflict resolution among the CS councils and between councils and other institutions. Another main responsibility of this department is to conduct training needs assessment of CSs and prepare training curricula for CS councils’ capacity development.

Moreover, the Provincial Rural Development Committee (PRDC) was also established by the Prakas #292 of Seila Task Force (Phase II of the Seila Program) in 2002. Based on the Prakas, the above committee is chaired by provincial governor. One deputy governor and director of the provincial department of rural development are deputy chairs, whereas the provincial department of planning is a permanent member of the committee. Also, this committee includes all the directors of the technical department and district governors as members. This committee has the responsibility to facilitate all kinds of development in the provinces. More importantly, it has the responsibility to support CSs in the implementation of their regular works that include the preparation and execution of CS Development Plan (CDP) and CS Investment Plan (CIP), monitoring and evaluation of CSs in complying with decentralization policy. The Committee also cooperates with other relevant governmental institutions and civil societies to build capacity for CS councils/staff, to assist in financial management, and to ensure that the CS project information is properly recorded in the database.

2.2.2 Decentralization and De-concentration Reform at Province Capital Municipality Khan and District.

Before the constitutional amendment in 2008, Cambodia is divided into Cities and Provinces. Then the city is divided into Khans and a Khan, to Sangkats. A province is divided into districts, into communes. Overall, there are 20 provinces, 4 cities, 14 Khans, 171 districts, 1510 communes and 111 Sangkats. With the process of decentralization reform for a period of time at the CS level, RGC has seen that the local governments are weak as a whole in the effective and efficient development and service delivery to local people, which required an administrative change at all levels of the local government. At the bottom level, CS councils’ capacity was still limited in implementing their responsibilities (RGC, 2005). On the other hand, its revenue was only the allocated budget from the central government, even though the law stipulates that they are able to get revenues from tax collection

At the Khan/district level, there was not an efficient mechanism for service provision from the line offices (line ministry institutions at Khans/districts). The lowest local service providing institutions from ministries are offices at Khans/districts, but these entities have just only a small budget allocated from provinces/capital city in execution their administrative work without any investment budget (RGC, 2005).

On the other hand, City/provincial levels, the governor assumes the responsibility of ensuring public order in the jurisdiction and coordinates all developmental efforts of line departments in a way to make them consistent with the RGC policies and development goals. However, at this tier, provincial line departments carry out the policies of central government, accountable to the ministries at the national level, not to the provincial governors. Accordingly, the Capital city and provinces are not able to effectively manage their budget, planning, and personnel in delivering the service satisfactory to the residents (RGC, 2005). In short, there is a scarcely local government with substantial autonomy, the RCG want a unified administration at sub-national level. According to the Organic Law, a unified administration (administrative jurisdiction) is the system of intergovernmental relations where councils at the sub-national level maintain authority to manage their finance, revenue, staff, and assets. They are to be equipped with the ability to effectively coordinate public service delivery and developments in the jurisdiction in responding to the needs of their local residents. With the issues described above, RGC has seriously committed decentralization and de-concentration (D&D) reform at all levels of sub-national government and has adopted the Strategic Framework for Decentralization and De-concentration in 2005 as guidance for D&D reform in Cambodia.

With the introduction of the D&D strategic framework in 2005, Seila Program was phased out by the end of 2006, and a new inter-ministerial committee, named as the National Committee for the Management of Decentralization and De-concentration Reforms (NCDD), was set up to lead the D&D reforms. Following the D&D strategic reforms of 2005, NCDD has led to the constitutional amendment of Cambodia on the Articles 145 and 146 (see above). Subsequently, the Organic Law for the administrative management of the capital, provinces, municipalities, districts and Khans was also adopted in 2008 for establishing a unified administration system at sub-national levels and continued support of the D&D reform in Cambodia. At the same time, NCDD has completed its mandate; then a new of National Committee for Sub-National Democratic Development has replaced the old NCDD. This new committee with the same abbreviation of NCDD is still now operating to lead the D&D reforms.

With the administrative management of the capital city/province, municipalities/Khan /district, the sub-national councils in the capital city, provinces, municipalities, districts and Khans were indirectly elected by commune councils for the first legislative term in 2008 (and since then, every 5 years). Under each council of the capital/province, municipality/district/Khan, there is a board of governor (governor and deputy governors appointed by the central government). The governors are accountable for both councils and RGC. As such, councilors as the representatives of local people make decisions on development planning and budgeting in their respective jurisdictions, whereas governors will implement the decisions made by the councilors.

In addition, the organizational structure in the capital/province, municipality/district/Khan was also amended in 2010 to ensure unified administration at the sub-national level; that is, the administration among the three tiers of government will be cooperatively and effectively promote sustainable development in their own jurisdictions. They will work effectively for social and economic development, because the resources of each tier will not be used in a duplicated manner to respond the needs of local people both present and future.

The role of PRDC no longer exists; it was integrated into the new unified administration system. Based on the guidebook issued by the Secretariat of NCDD in 2010 about the integration of the current organizational structure and staff to the new organizational structure, two divisions among the five of the provinces which are planning & investment division and financial division have been working closely to support the local governments at the municipality/district/Khan and CS levels such as supporting administrative work, cooperation and development planning, project feasibility study and project proposals to be included in the capital/provincial development, especially on the financial operation. On the other hand, at the municipality/district/Khan level, a new office called Office of Planning and Supporting Coummune/Sankat was established to closely support decentralization practices at communes/Sangkats.

Implementation of the Organic Law and democratic development (D&D policy) at the sub-national levels is now full-fledged. RGC and development partners have worked together to create a National Program for Sub-National Democratic Development (NP-SNDD) for 10 years (2010-2019), which was approved by Council Ministers on 28 May 2010. With this plan being implemented, the sub-national government at the three tiers is expected to assume more autonomy. The functions will be decentralized from the central ministries to the sub-national governments to make service delivery closer to the local people. Especially, the local governments will be able to have the autonomy to earn and manage their revenues for local economic development with participation from their own people.

2.3 Decentralization Effort at Commune Level to Economic Development

We could see that RGC has a strong commitment to enable this reform to be successful with the expectation that it will contribute to poverty reduction and local economic development for the people. Since 2002 up to now, we could see that only the commune/Sangkat level has been implementing in a more decentralized manner, whereas others is still more centralized as they have just started the decentralization reform. So what the commune/Sankat has done to promote economic development?

According to LAMCS, after councilors elected, they are required to prepare a five-year commune/Sangkat Development Plan (CDP) for the development in their own jurisdiction. Based on the guideline on the CS development plan and CS investment plan, to prepare a development plan with the limited amount of budget, the councilors are required to consider with some of the dimensions such as priority needs of the own local people (meeting with local people) which include infrastructure and non-infrastructure project investments, and what to achieve in the next five years, what services the councilors want from the Ministry’s line departments at sub-national levels. The plan is required to cover with 5 main components which are economic, social, natural resources and environment, administrative & security and gender. After getting the draft of CDP, CS councils will have a public meeting with their own people for the comments, and other stakeholders such as NGOs, representation from the capital/province, private sector, etc. After reviewing the draft with above participants, councilors will have a meeting with capital/provincial governors and capital/provincial line departments and related institutions to get their reviews on draft CDP. After this review, the councilors will have a meeting to approve their CDP. Once CDP is approved for five years, then Councils need to prepare Commune Investment Plan (CIP) to be more detail in project investments with the available budget on an annual basis to execute the CDP. CIP is done with participation from the local residents to identify what projects should be done with the estimated cost, then this information will be recorded for District Initiative Workshop (DIW). DIW is the mechanism to mobilize the resources from various sources to support the project investments (with a limited budget, CS need to find from the other sources) which have been recorded in CIP, and to guide all the CS project investments will comply with the policies of national and provincial level. The participants of DIW are capital/provincial governor, Ministry’s line department directors at capital/provinces, district/Khan/municipality governors, NGOs, donors, and other supporter. Those will support to the CS projects which are considered to be consistent with their organizational vision, mission and planning.

Referring to project information database (PID) which records all the investments by CSs from 2009-2012, there are many project types that have been implemented by CSs in the areas of the agriculture, community fisheries, community water resource management, community forestry, domestic sanitation, education, electricity, environmental management, health, irrigation, market infrastructure, rural domestic water supply, rural drainage and flood protection, rural transportation, social intervention, tourism development, urban drainage and flood protection and urban transportation. Among of these project types since 2009-2012, CSs highly invested in rural transportation projects, which is 81% of total investment expenditures, whereas urban transportation is only 5%. The average budget spent for the project investment is 4,750 USD. These projects have contributed to poverty elimination and local economic development.

Table 2.3: Types of Projects Invested by Commune/Sangkats 2009-2012

Project Type

Freq.

Percent

Cum.

Agriculture

448

2.75

2.75

Community Fisheries

85

0.52

3.28

Community Water Resource management

3

0.02

3.3

Community forestry

78

0.48

3.78

Domestic Sanitation

18

0.11

3.89

Education

60

0.37

4.26

Electricity

2

0.01

4.27

Environmental Management

48

0.3

4.56

Health

2

0.01

4.57

Irrigation

1,149

7.07

11.64

Market infrastructure

1

0.01

11.65

Rural Domestic Water Supplies

177

1.09

12.73

Rural Drainage and Flood Protection

76

0.47

13.2

Rural Transport

13,137

80.78

93.98

Social Interventions

2

0.01

93.99

Tourism development

30

0.18

94.18

Urban Drainage and Flood Protection

102

0.63

94.8

Urban transport

845

5.2

100

Total

16,263

100

Source: http://db.ncdd.gov.kh/pid/home/index.castle

In the evaluation done by Romeo & Spyckerelle (2003), they have described that 60% of small-project investments implemented by CSs in 2003-2004 was on rural transportation and followed by 16.3% of other projects was on canal & irrigation construction and rehabilitation. With this result, he argued that the two categories above "reveal the strong local preference for economic infrastructure." (p. iii)

CHAPTER III

LITERATURE REVIEWS

Many scholars have spent their time to study about the correlation between the central and local governments. They have studied more specifically about how these two interact to improve the life of the people in the country with better quality and welfare; that is, how the relations of these two will contribute to economic growth. To this end, the decentralization policy has come the object for many scholars to study, how it is associated to the economic growth/development. Decentralization is classified into three dimensions that are administrative, fiscal and political decentralization (Iimi, 2005). Fiscal decentralization (FD) is how much the central government transfer resources to local governments; whereas, administrative decentralization refers to what extent local government obtain autonomy from the central government. Political decentralization means to what extent the local government has political functions as a representation of local people (Schneider, 2003). Fiscal decentralization can also be the revenue and expenditure assignment to sub-national governments (Kim, 2013).

At the same time, economic growth/development measure which have been used by many scholars include GDP per capita (Iimi, 2005; Rodriguez-Pose & Kroijer, 2009 and Stansel, 2005). Also, population growth is also used as the measures for the economic growth/development. (Stansel, 2005).

3.1 Relations between Decentralization and Economic Development.

In general, it is believed that decentralization will provide positive impacts on the society; that is, economy and quality of local people lives may be improved through decentralization. This argument is supported by many scholars. Oates (1993), the fiscal decentralization has a positive impact on economic growth in term of economic efficiency. Oates has elaborated that the public expenditure by sub-national governments that satisfy the preferences of regional and local people has enhanced economic development than that of the national government.

Also, Stansel (2005) argued that the decentralization is really kind of catalyst to promote economy. He has conducted the studied on the relationship of government structure to economic growth, focusing on decentralization by using a data set of 314 metropolitan areas from 1960 to 1990 in the United State of America. In the model for analysis, two dependent variables were used to measure the economic growth, which are population growth and per capita income. The reason the he used the population growth as a proxy for local economic growth is that the more population growth, the more attractive of the metropolitan to people and investments for business. On the other hand, the measure for local decentralization is the number of local governments as they are the service providers to the local people. In addition, the number of school systems per 100,000 residents is also considered as one of measures to local decentralization, "since the quality of local schools tends to have a strong impact on residential choice" (p. 60). Also, some control variables are included into the model to run the OLS regression, which are population growth from 1950-1960, population in 1960, real per capita GDP in 1959, unemployment rate in 1960, share of the manufacturing sector to total employment, population whose age is 25 or older in1960 with 16 or more years of school and state dummy variables. The result of the regression has demonstrated that the more population increase, the growth of the economy goes up. Interestingly, a positive sign is found on the coefficient of a number of local government variable.

Moreover, the two proxy indicators for decentralization have a negative impact on the economy. Those variables include sub-national expenditure and transfers to local government; whereas, sub-national tax also has inverse relations with economic growth, but only at the initial, then it gradually move from negative to positive relation after one year. This finding means that the sub-national government has its own revenue will promote economic growth (Rodriguez-Pose & Kroijer, 2009). This finding is based on the study the level of fiscal decentralization and economic growth by using panel data from 1990-2004 in 16 Central and Eastern European (CEE) countries which considered as developing countries. In this study, three indicators were used as proxies of fiscal decentralization which includes sub-national expenditures, sub-national tax revenues, and transfers from the central government. The fixed effect is utilized in running the model y= α + β1x + β2z + ɛ

Where y is GDP per capita, x is a cluster of control variables, namely growth of population, GDP per capita in the initial period, proportion of investment to GDP, growth deflator, computer numbers per 1,000 inhabitants and human capital (secondary school enrollment) and z is fiscal decentralization measurement.

Furthermore, Iimi (2005) confirmed that any forms of decentralization will effect on economic growth. He has used data across both developed and developing countries over the period of 1997-2001. The model which is used by Davoodi & Zou (1998) in analyzing the impact of decentralization on economic growth is applied, and he adds the variable of political freedom.

"gi = θ0 +θ1τi +θ2 FDi + θ3 PFi + θ4FDi * PFi + θ5 ni + θ6Xi ɛi " (p. 453) where g: is the growth rate, τi : tax rate, FDi: Sub-national government share of total government spending, ni : Population growth rate, Xi : control variable group, PFi: measures of political freedom.

With this model, the dependent variable refers to GDP per capita over the five years. The result shows that the fiscal decentralization (sub-national expenditure) has significantly correlated with economic growth – GDP per capita positively. However, there is a negative correlation between political freedom and growth.

Actually, there have been a few studies have been done as well to examine the relations between the decentralization and economic development in Cambodia by Andersen (2004) and Romeo & Spyckerelle (2003).

Andersen (2004) has argued "The investments show a high rate of economic return" (p. 2). This assertion was based on the study of Socio-Economic Impact of Local Development Fund/Local Planning Process and Review of Provincial Investments in Seila 1997-2001 by (Tracy-White et al, 2001&2002). This study used the difference in difference method, comparing the situation of the beneficiaries before and after the commune/Sangkat project developments based on their survey responses as qualitative indicators. For example, road projects have contributed to reducing time and fare of the journey. It is a 40% reduction in travel time and 30 % in the fare. Moreover, school projects have increased the number of schools, which led to the increase of school teachers. It also reduces average commuting time for students to and from school.

In addition, Romeo & Spyckerelle (2003) has observed in a case study on Local Governance on Pro-Poor Service Delivery that budget transferred for CS from 2002-2003 was a strong contribution to budget for CS project developments in economic infrastructure, especially for small infrastructure projects targeting roads, bridges, canals, and irrigation systems. About 80% of the total transfer to commune/Sangkat nationwide was spent on local infrastructure projects.

In contrast, some scholars have found the negative relations or no relation between the two variables (Davoodi & Zou; 1998, Woller & Phillips, 1998 and Xie, et al. 1999 and Zhang & Zou, 1998).

Woller & Phillips (1998) has revealed that no strong correlation between fiscal decentralization and economic growth. This argument is based on his study within 23 less developed countries (LDCs) on the link of fiscal decentralization and economic growth/development from 1974-1991.

Zhang & Zuo (1998) have analyzed the trend of fiscal budget between national and local government on economic growth by using provincial panel data from 1978 to 1992. In this empirical study, they had selected some variables to examine the relationship between two variables which is provincial income growth rate as the dependent variable. At the same time, they select the independent variables which include investment and labor, fiscal decentralization in terms of spending, the total expenditure of the province and central government. Other control variables are also included into the model such as the tax rate, foreign trade and the inflation rate. In analyzing by Least Square Dummy Variable (LSDV) regression, the result has shown that there is a negative sign on coefficient and significant of fiscal decentralization; that is, there is an inverse link between fiscal decentralization and the growth of provincial income.

In short, the study done by Zhang & Zou (1998) has indicated that there is a negative link between fiscal decentralization and the provincial economic growth. This result has absolutely made scholars or experts surprised, because they are supposed to be associated with together.

Besides, Davoodi & Zou (1998) have claimed that there is an inverse link between fiscal decentralization and growth in developing economic countries whereas in developed countries, there is no correlation. Data set from 1970-1986 of 46 countries have been used for this study. The study concentrates on the how efficient of fiscal decentralization contribute to the economic growth. Fiscal decentralization at this time refers to the expenditures by sub-national government as part of total government expenditure. In analyzing, they have estimated the Ordinary Least Square (OLS) regression with control variables and without control variable. Independent variables which have been used for this regression includes "average tax rate, average population growth rate, initial human capital, initial per capita GDP, Invest Share of GDP" (p.248)

Moreover, (Xie et al., 1999) has mentioned that "further decentralization in public spending may be harmful for the growth." (p. 228). The unit of their study was the three tiers of government of the United State of America from 1948-1994. They have used OLS regression study the relation between the fiscal decentralization and economic growth. To measure fiscal decentralization, they have used the share of expenditure using by each level; however, they have included only two shares into the regression which are local and state spending share as independent variables, whereas per capita growth rate is dependent variables. Result regression has demonstrated that local spending share coefficient has a negative sign which reflects that fiscal decentralization has a negative relation to economic growth. However, the t-statistics are not significant. At the same time, At the same time, the positive sign on state spending share with growth is found, but not significant.

However, the reason why those scholars such as Zhang & Zuo (1998) and Xie et al. (1999) found negative relations is due to the selected data for the studies. They have used data in a time that China and The United States obtaining the high economic growth, so at this time, the high tier of government would have the main role for the economic development at the early stage. Also, Davoodi & Zou (1998) has used cross-country data; therefore, it is difficult to examine the relations between fiscal decentralization and economic growth as they have a different history, cultures and the institutional management in their own countries Akia & Sakata (2002).

Moreover, the more reason that negative relations are found between fiscal decentralization and economic growth, because, in their model, they have not controlled on variable of political freedom. When we talk about decentralization, we need to know that there are three dimensions of decentralization such as political, fiscal and administration decentralization. Secondly, it is concerning the limited capacity of local governments. This means that decentralization needs time to realize the theorized positive impact on economic growth. Lastly, it is about migration such that people can move freely from one area to another; that is, if they live in one area in a short period of time, the measures to respond to the needs of people may become not consistent (Iimi, 2005).

3.2 Hypothesis:

With the brief review of the empirical studies above, a hypothesis is stated that it is positively associated between decentralization and economic growth/development. Even, there are a few studies on the impact of decentralization on local economic growth/development in Cambodia, but those have not used the model applied by many scholars, especially they have not used the variable of GDP per capita, which is a main proxy indicator for economic growth/development to examine the impact of the decentralization. Actually, GDP per capita at CS level is not available till now in Cambodia. Therefore, this study will choose some proxy indicators for economic development instead of GDP per capita. Those are the number of small businesses, the number of motor vehicles, infant mortality rate, and poverty rate in CS. Whereas, decentralization measure in Cambodia for now will be based on the fiscal decentralization; that is, it refers to the CSF which is the budget transfer from the central government to 1621 CSs.

The hypothesis of this thesis can be mentioned; in other words, that fiscal decentralization in Cambodia will provide a positive impact to economic growth; that is, the CSF will have a positive relation to the numbers of small business and the number of motor vehicles, whereas there is a negative correlation between CSF and infant mortality rate and poverty rate.