FISCAL DECENTRALIZATION AND EDUCATION IN THEORITICAL OVERVIEW

"By Beny Trias Oktora Master of Arts in International Development and Cooperation at Korea University Graduate School of International Studies"

1. Conception of Decentralization in Economics

Foundation of fiscal decentralization theory lays on economic efficiency of allocation of public goods and services or efficiency of resource allocation, welfare gains of locally provision of public goods and services and mobility citizen to choose their preferences that are offered by local government as implied in Tiebout Model. Fiscal decentralization is one of sub divisions of public finance that discusses specifically on how a state manages their revenue and expenditure in accordance with basic principle of economics. Fiscal decentralization is mostly related with allocation function of economics. In first section of Chapter Two, this thesis focuses on discussion of theory of fiscal decentralization

United States is considered as origin of fiscal decentralization theory or fiscal federalism theory. In United States, public finance has reached advance development that indicated from innovation to grant local government to administer its local fiscal policy. Gramlich (1993) simply stated that in the name budget deficit, federal government has to decentralize its fiscal. By cutting state and local grants to local government federal government can reduce its deficit.
Tiebout (1961) in An Economic Theory of Fiscal Decentralization emphasized theory of fiscal decentralization from economic perspectives which are from supply and demand. Tiebout neutralizes political elements however decentralize in fiscal involves multilevel government. In supply side Tiebout discussed how to provide public goods and services efficiency. Branches of government central government and local government proceed as “production site” of public goods and services. Within centralism framework, central government assumes that their citizen had the same preferences of public goods and services so that central government provides public goods and services in uniformity modes. Regardless, there are some variations that exist in whole country.

By showing simple calculation of providing public goods inside municipal, Tiebout clearly described that in scope of municipal uniformity of public goods can meet criteria of economic efficiency. Important elements of Tiebout`s calculation, fire services and police patrol service, per square miles area of municipal and cost of provision public goods that in “U” shape.
Oates (2005) in Toward a Second-Generation Theory of Fiscal Federalism reviewed evolution of fiscal decentralization which has been evolved. First generation of fiscal decentralization theory explored from supply side which means how efficient of local public good being provided by local government. Whereas second generation put local government as private company that should compete with other local government in attracting citizen by providing best local public good or preferences. The assumption is citizen is mobile to find better preferences of them. As author observed from some articles that also become this thesis references, decentralization has space to explore either in political science or economics. Oates (2005) emphasized the second generation of fiscal decentralization utilized each aspects or each parts that contained in decentralization to explore that are never studied before.

Another Charles M. Tiebout works that explored of local expenditure conception is worthy to discuss in this thesis. In A Pure Theory of Local Expenditures, Tiebout (1956) explains theory of local expenditure that had different interpretation even though theory of local expenditure had “up-branch” from theory of public finance. Tiebout answers analysis of Musgrave and Samuelson on federal expenditure that does not have any similarity to apply in local expenditure. Tiebout article is explaining expenditure of public goods that conducted by local government.

Precisely, Tiebout describes theory of local expenditure that is modified from Musgrave-Samuelson theory on of federal expenditure. Conception of Musgrave and Samuelson on public goods is ”collective consumption goods” so that there is no subtraction from other individual because state has already provided for all individuals. The central point of federal expenditure or unitary state expenditure is public goods. Both two different governments should produce public goods that match for citizen needs. The problem is how to determine citizen preferences or Samuelson describes citizen preferences of public goods as consumers of state or government that had rights as voters so author called them consumer voters. And the other one is at the same time consumer voters had rights of provision of public goods and also they are object of charging tax. If government had succeeded of discovering consumers voters preferences so government can easily provide public goods for consumers voters.

Difference between central and local provision of public goods is the central level the preferences of the consumer-voter are given, and the government tries to adjust to the pattern of these preferences, whereas at the local level a mixture of governments have their revenue and expenditure patterns more or less set. Given these revenue and expenditure patterns, the consumer-voter moves to that community whose local government best satisfies his set of preferences. To develop model of local expenditure, Tiebout develops several assumptions:

1. Consumer-voters are fully mobile and will move to that community where their preference patterns, which are set, are best satisfied.
2. Consumer-voters are assumed to have full knowledge of differences among revenue and expenditure patterns and to react to these differences.
3. There are a large number of communities in which the consumer-voters may choose to live.
4. Restrictions due employment opportunities are not considered. It may be assumed that persons are living on dividend income.
5. The public services supplied exhibit no external economies or diseconomies between communities.
6. The assumption that some factor is fixed explains why it is not possible for the community in question to double its size by growth. The factor may be the limited land area of a suburban community, combined with a set of zoning laws against apartment buildings. It may be the local beach, whose capacity is limited. Anything of this nature will provide a restraint.
7. The last assumption is that communities below the optimum size seek to attract new residents to lower average costs. Those above optimum size do just the opposite. Those at an optimum try to keep their populations constant.

Because of consumer voters are fully mobile to determine their preference simply they search the best preferences of those local governmentsscheme of expenditure. Public goods in provision by local government are different from central government.
Rubinchik-Pessach (2005) in “Can decentralization be beneficial?” is trying to identify the benefit of decentralization by building a model to present that decentralization, even though some argued about unitary or centralized government is cheaper and the coordination or asymmetry problem exist, is valuable to implement. Most of discussion in decentralization especially in fiscal federalism or fiscal decentralization is focus on asymmetry in policy or lack of information access that resulted poor of public goods provision because of delaying in public goods and services delivering. However, Rubinchik-Pessach argues that even when asymmetry on policy tools and lack of information access, fiscal decentralization fetches benefit or welfare to citizen.

Most argument to legalize and support theory of fiscal decentralization or decentralization itself is because bringing government closer to its citizen by forming local government or using author term other tier of government (put another tier of government in hierarchy of government) with devolving authority will improve of provision of public goods in local or regional area. Furthermore, government that closer to its people naturally will know better its people preferences of public goods and the result is fulfilling citizen needs accomplished. To prove that argument, Rubinchik-Pessach builds a model.

Why decentralization can generate benefit according to Rubinchik-Pessach`s model? Rubinchik-Pessach argues that the reason decentralization is beneficial is transparent: under the hierarchy each level filters correctly the corresponding type of public project, as the design of recognition rules for the central government is independent of the presence of good local projects. This is not true for the unitary government that is pushed for to choose a rule that is optimal bonaverage for both local and global projects.

Rubinchik-Pessach`s model offers an economic foundation for the subsistence of a hierarchical government, which may do better than the unitary one, even in the absence of the vertical asymmetric information within the hierarchy and even though the absolute advantage of the (upper) central government in making decisions with respect to public goods is evident. Rubinchik-Pessach argues that it is contrary to the accepted view, if the issues to be tackled by the government are mainly of local importance, although generating some spillovers, central government alone is sufficient and there is no need to decentralize.

The hierarchy induces specialization of each level on the corresponding issues, thus, enhancing the overall welfare, which can justify possible costs associated with the additional level of government. Therefore, the main argument does not come from an assumed not efficient of a central government whereas lay on the idea of specialization.

Rubinchik-Pessach from another angle has captured that even without any asymmetry and globally opened information in provision of public goods, fiscal decentralization is beneficial in transparency and specialization. Each tier of government in term of doing filters of public project could select appropriate public project. Mixed of author skills in formulating ideas in decentralization and mathematical format skills has produced very different analysis of the beneficial analysis of fiscal decentralization. Cleverly author has been formulated the core principle of decentralization both in political field and economical field i.e. public finance. For example, in Rubinchik-Pessach`s model clearly author stated political systems in unitary and hierarchal government and how the flow of public finance goes in financing public project for sake of citizen. Rubinchik-Pessach`s article, is her dissertation, has given significant contribution in the study of decentralization and fiscal federalism or fiscal decentralization. Rubinchik-Pessach`s idea and argument are inclusive because author can combine theory of decentralization in domain of political science and theory of decentralization in public finance or economic science.

Arzaghi and Henderson (2003) in Why Countries are Fiscally Decentralizing had simple goal which is going to search and find the reason why countries are more decentralize in fiscal using 48 countries that lay in all continents as sample during 25 years since 1960 to 1995. To attain that goal Arzaghi and Henderson (2003) developed model and investigated what factors or determinants that force countries to put into practice fiscal decentralization.
Using those data, Arzaghi and Henderson (2003) compare the data of degree centralization and decentralization year by year. The result was developed countries and countries in the former Soviet bloc and Latin America have the lowest centralization while Middle Eastern and Sub-Saharan African countries have the highest centralization.

To explaining why, Arzaghi and Henderson (2003) developed model which are consists of two models. In their model, Arzaghi and Henderson (2003) consider a country that comprises two regions. They name region 1, the coastal region, and region 2, the hinterland region. The country has a unitary government, which is located in the coastal region. The unitary government provides local public goods to both regions. Arzaghi and Henderson (2003) ask under what conditions the hinterland region would want to separate. By separation, hinterland region is willing to set up and finance a regional government in order to provide its own local public goods. Central government does not stay away from providing a set of national public goods to both regions, although, Arzaghi and Henderson (2003) formally ignore this possibility for the sake of simplicity.

In modeling separation, they deem three scenarios. In the base scenario, the population and individual incomes are fixed in each region and governmental institutions are perfectly democratic. As the first extension, they diverge from the base scenario by considering limited or partial democracy. The officials or political leaders in the central or local governmentsare elected by majority rule but those officials or political leaders bring some leverage over the importance of publicly provided goods. Furthermore, when the second extension, they consider briefly a case with a mobile population. In this case, both population and income in the two regions are endogenous.

Arzaghi and Henderson (2003) formulate hypothesis to develop their analysis. Under imperfect population mobility, with perfect democracy suggested that separation, or adoption of formal federal structures is promoted by (1) income growth, (2) population growth, (3) higher spatial decay of local public services provided to the hinterland by the central government, (4) relative income growth in the hinterland region, (5) relative population growth in the hinterland and (6) lower costs of government for the hinterland region. With imperfect democracy had the same hypotheses, plus the additional one that separation is promoted by there being a greater degree of local democratic culture, compared to the nation.

Then they tested the hypothesis. Empirically, federalism can be defined as an institutional all-or-nothing concept: does a country adopt a federal constitution, or does a county adopt a federal constitution (de jure) and also hold local elections (de facto)? They also built up an index of the degree of institutional federalism, or decentralization, for a more nuanced institutional approach. The institutional approach has the advantage of being directly connected to their model, which asks when countries will adopt a working federal constitution.

Though, in many different circumstances such as cultural, historical, and political, institutional structures are very rigid and change may involve almost unworkable political processes. In contrast fiscal structures are much more flexible and can be altered through legislative initiatives or even executive order. Thus, a unitary country which would have great difficulty altering its constitution in a twenty-year time framework can still decentralize fiscally and become de facto federalized. They also examine fiscal decentralization where there is a continuum of outcomes described by the degree of fiscal centralization, as measured by the central government’s share of government consumption. For this, they think of the forces towards separation, as the forces that move a country across that continuum, from a high degree of fiscal centralization to a low one.

In their article, they develop and test a simple model of fiscal decentralization within a country. They hypothesize that basic economic and demographic variables explain most of the observed variation in the degree of institutional and fiscal decentralization across countries. They find that income per capita, population, land area, and the degree of population concentration in the largest (typically capital) city in a country have large effects on the degree of decentralization. Income and population growth lead to decentralization, as does population decentralization. The degree of centralization in countries is largely explained by simple economic and demographic growth processes. However, at the end, they argue that institutions affect fiscal outcomes. Having a federal constitution encourages both effective federalism and fiscal decentralization. Also, national democratization leads to potentially regional representation and the development of regional demands for greater fiscal autonomy. They have developed refined argument to explain why there are tendencies that most government to fiscally decentralization.

2. Fiscal Decentralization and Provision of Public Goods and Services in Developing Countries

There are a lot of discussion related to fiscal decentralization that emphasizes the efficiency of allocation of resources or public goods provision and developing countries. Most of empirical works for example Tiebout (1956 and 1961) stressed on the efficiency of resource allocation and efficiency of public goods provision in some developing countries that are eager to put into practice fiscal decentralization in order to achieve of economic efficiency. Most of them are located in Latin America, South East Asia and transition countries explicitly ex-communist countries in Eastern Europe and ex-soviet region. This thesis presents articles of fiscal decentralization and provision of public goods and services that are closely related to Indonesia precisely in south east Asia and generally developing countries and excludes other regions which had different background.

Larry Schroeder (2003) in Fiscal Decentralization in South East Asia is reviewing the implementation of fiscal decentralization in four countries in south east Asia which are Indonesia, Malaysia, Philippine and Thailand and making lists of points that are using to assess the feature of development fiscal decentralization in that region (south east Asia). In the region, Philippine is considered leading in fiscal decentralization implementation followed by Indonesia and Thailand in the midst and Malaysia still not interesting in involving deeper in fiscal decentralization.

The checklists are developed by Schroeder for review as follows.

1. Constitutional or statutory basis for decentralization.
This benchmark emphasizes how important legal basis for fiscal decentralization so that guarantee of autonomous of local government in operating fiscal decentralization.
2. Genuine political commitment.
Moreover, the real support from political field is needed and it consider as important as legal basis. Because without any commitment from political field, local government autonomy is useless.
3. Explicit service assignments.
Division of responsibility after autonomy is obviously necessary. What sectors that belong to central government responsibility and what sectors that local government should manage so that there are no overlaps between central government affairs and local government affairs.
4. Granting local revenue raising authority with explicit autonomy.
As there are responsibilities for each tiers of government, there should be room for local government to search sources to finance those responsibilities through tax. Yet it is not became constraints for macroeconomics stability in the country.
5. Local decision-making powers provided.
Moreover, power in budget decision making is useful in undertaking provision of public goods and services especially to meet efficient allocation of resources in local region.
6. Capacity to carry out these powers.
Personnel of local government in carrying autonomy power in frame of fiscal decentralization should be ready in managing relatively large authority. If not, local government should put investment to improve their personnel skills.
7. Creation of an intergovernmental transfer program.
Intergovernmental transfer program is needed to be improved in manner of its simplicity and transparency as extensive of fiscal decentralization so that its system would be understood by all.
8. Public participation encouraged.
People participation to ensure local government is looking for their preferences is ultimate checklists. Without people voice that represent their preferences of public goods and services provision, theoretically foundation of fiscal decentralization is not working.
Based on that review lists, Schroeder reviewed Indonesia in second place below Philippine. There are some burdens such as relatively low personnel skills in some local government to carry extensive assignment of fiscal decentralization and degree of dependency from central government on financing local government responsibility. Nevertheless, Schroeder argued that based on diversity that Indonesia had, fiscal decentralization had potential advantages for Indonesia e.g. on what way Jakarta would know people of Papua preferences unless there is representative elected government either executive or legislative that presumably know better of people of Papua preferences.

Smoke (2001) in Fiscal Decentralization in Developing Countries: A Review of Current Concepts and Practice highlights some findings. This thesis found interesting facts that discovered not only benefits but also negative of fiscal decentralization especially for developing countries and these facts are quite different with Prud`homme (1994) of On Danger of Fical Decentralization that emphasized only the dangers or demerits of fiscal decentralization not only for developing countries but also for developed countries.

Wave of decentralization in developing countries is relatively huge. Despite of its benefit or positive impact are still debatable especially on outcomes of most important of public policy, many developing countries are eager to implement decentralization policy. There are variations of opinions which are conclude that decentralization generates benefit, second conclude that decentralization is adding another problem to developing countries and the large numbers of them are ambiguous to conclude whether decentralization is benefited or demerit to developing countries.

Faguet and Sánchez (2006) in Decentralization’s Effects on Educational Outcomes in Bolivia and Colombia are focus on two countries which are Bolivia and Colombia that overall had similar condition in implementing decentralization policy background. Most developing countries are highly centralized and none of them held general election for governor and mayor. Central government explicitly pointed that president directly chosen governor or mayor (in Indonesia most governor and mayor are active or retired military during authoritarian or military regime held by President Suharto).

They are using quantitative to approach the relationship between decentralization and education. Most of the data are from government agency. Firstly authors examined using descriptive statistic analysis. In Bolivia, they found that decentralization seems to have changed the sectoral uses of investment and its distribution across space. Central government highly invested in economic driven sector such as infrastructure and energy. Whereas local government highly invested in social sector and the highest was in education, followed by rural development and water and irrigation. In distribution, local government was doing properly normal distribution whereas central government not.

Moreover, they made regression between government spending both central and local government and literacy rate. The line was negative relation between central government spending of education and literacy rate; it means that central government spending was negatively effect to literacy rate. While local government spending of education has had positive regression.

In Columbia, Faguet and Sánchez (2006) were doing the same thing which was descriptive statistic analysis. The result quite different where central government also spending relatively significant spending in social sector health and education and also did local government. Lastly, in Columbia they made comparison before and after decentralization of school enrollment and the result there was tendency that school enrollment has increased after decentralization. However using descriptive statistic analysis authors made not clear measurement for both cases.

To discover the facts clearly, Faguet and Sánchez (2006) formulated model to test the significance of decentralization to education. Because they argued that descriptive statistic analysis is not appropriately discovering the facts. In both countries, they found that significantly decentralization was fostering education. Decentralization was driven local government to more responsive for local needs. In Bolivia, public spending after decentralization was shifted from economic dimension expenditure to social expenditure such education and health. In Columbia, it was significantly decentralization made more children goes to school. Decentralization was empowered for those municipals that considered as poor municipal. So they can afford appropriately education using only their budget. Before decentralization, small and rural municipalities were being ignored by central government. So when decentralization was occurred those small and rural municipalities conducted relatively responsive to local needs. However, author did not find any discussion how small and rural municipalities could afford their education budget. If their condition is rural and in remote area it is probably that their economic activities are considerably poor and do not have sufficient sources. They should describe what kind of policy of decentralization in both countries.

Faguet and Sánchez (2006) concluded that the importance of systems before decentralization occurred which contributed significantly to the successful of decentralization. There was marking of central government policy conducted before decentralization implemented. If before decentralization central government did allocate resources to local government appropriately it would be better initial investment for local government in decentralization period.

Compared with Indonesia, poor provinces and municipals they received relatively low transfer fund from central government because there is not enough resources to share. It is different with provinces and municipals whose had sufficient natural resources and considered as tourism site, trading point and another services activities they had sufficient amount of fund that transfer from central government.

Economics is science that mostly focuses on how to allocate resources within a country. Recent development on how efficiently to allocate public goods is to apply decentralization either in economically or in politically. Currently, development of decentralization is to match with public preferences or market friendly as authors (Lindaman and Thurmaier (2002)) implied in their articles when they classified decentralization in three waves of decentralization. And Lindaman and Thurmaier (2002) in Beyond Efficiency and Economy: an Examination of Basic Needs and Fiscal Decentralization are trying to assess the effect economic aspects of decentralization i.e. fiscal decentralization on fulfilling human basic needs.

To capture the relationships, they are using quantitative methodology. They form hypothesis: higher achievement of basic needs objectives is, in part, a function of increased fiscal decentralization, increased national income, and the structure of government. It means that increase of fiscal decentralization, national income and structure of government will foster higher achievement of basic needs. Dependent variable of the model is human development index. Then independent variables are consist of three measurement of fiscal decentralization. Firstly, "DecenWAS" is the ratio of subnational expenditures, divided by total government expenditures less national social security and military expenditures. "DecenBHL," which follows from Bahl and Nath and is the ratio of subna- tional revenues-including intergovernmental grants from the central govern- ment-divided by total government revenues less international aid grants to the central government. And the third, "DecenREV" is the ratio of sub-national revenues-excluding intergovernmental grants from the central gov- ernment, divided by total government revenues less international aid grants to the central government.

Finally authors` model is HDI = B0 + B,(DECEN) + B2(ColHistS) + B3(ColHistF) + B4(ColHistB) + B,(Ebloc) + e, where HDI is the human development index; DECEN is either the DecenWAS, DecenBHL, or DecenREV decentralization term, scaled as a percentage, for 1985, 1990, and 1995; ColHistS is a dummy variable for Spanish colonial history; ColHistF is a dummy variable for French colonial history; ColHistB is a dummy variable for British colonial history; and Ebloc is a dummy variable for countries of the former Soviet Union and Eastern bloc.

The result of their regression is significantly and positively supports their argument that fiscal decentralization fostered high achievement in meeting basic human needs in term of high literacy rate. Their results provide evidence of a significant relationship between greater fiscal decentralization and better performance on measures of basic needs such as health and education. Authors argued that increasing fiscal decentralization appears to increase achievements in basic needs. Furthermore they pronounced that fiscal decentralization in terms of own-source revenue has the strongest positive effects on basic needs.

Their model provides a foundation for further research into the question of whether decentralization benefits basic-needs objectives in some quantifiable way. Because they argued that decentralization benefits are intangible or otherwise unquantifiable and their study does not address that issue. Certainly much more research must be done to verify the evidence presented in this study. More studies conduct could verify whether decentralization is an exercise in administrative reorganization or a means to a better life for the citizens of developing nations.

They has developed model that can explain the efficiency and allocative of fiscal decentralization in the sense that delivering of public goods and or public services. Fairly, they do not mention that their effort is not ultimately concluded that fiscal decentralization necessarily is empirically beneficially proved otherwise they implied that to capture real benefit of fiscal decentralization through some quantifiable way.

Kristiansen and Pratikno (2002) in Decentralising Education in Indonesia explored decentralization from political perspective and narrow down especially in decentralizing education in Indonesia. Initially, massive decentralization policy in Indonesia has started from political perspective and was followed by decentralizing in public finance function. Development of education in Indonesia started after its independence and during military regime that was conducted by President Suharto, there were highly intensified concern to develop education especially for elementary education that known as six years education program. However, when financial crisis occurred in 1997 there was enormous demand to reform of Indonesia political system from highly centralized to decentralize.

To explore the topic, they used mix data which were consists of qualitative and quantitative. To study and find empirical result, they chosen four districts or cities or municipals that based on national criteria of one district consists of population on average 500.000 (they were preferable to use district rather city or municipal). In each districts, there are approximately 10 sub-districts which are furthermore divided into administrative villages and sub-villages. Their selections have had different or variation of income per capita, level of urbanization and centrality within national context. Those districts are Kabupaten Bantul, Kota Mataram, Kabupaten Kutai Kertanegara and Kabupaten Ngada.

They employed quantitative and qualitative method to analyze the topic. In three districts which were Bantul, Mataram and Kutai Kertanegara, they conducted focus group discussion that involved various groups with also various different backgrounds (qualitative method). Those groups were legislative members, executives bodies of government, members of NGOs and ordinary parents. In quantitative method, they conducted household surveys in selected sub-districts of each districts using typical criteria of each sub-district. In each of sub-district, they divided into village (desa) and sub-village (dusun). Overall they covered 538 households (in sub-village, they conducted random sampling that not covered all households in each of sub-village and they set up minimum amount households in sub-village involved in households surveyed.

Moreover, they added secondary data from BPS (Biro Pusat Statistik) / Central Statistic Bureau of Indonesia for instance Susesnas / National Census and Survey. However, they doubted its availability and reliability of such data from government agency. And they admitted that their own data also had limitation as follows. In focus group discussion or depth-interviewed, respondents had tendency not telling the truth. School enrollment data probably had over-reported. Indonesia even though in relative free condition after authoritarian regime still had excused criticizing government. They conducted research in initial path of decentralization policy. Respondents judgment and opinion on quality of education should be analyzed carefully. One necessary step was during the interview none of government officer was presence.

Their findings and discussions consisted in three folds which are the administrative aspects and service delivery from the perspective of service providers (using qualitative data), changes in the quality of schooling as perceived by parents, mostly applying data from the surveys, changes in household costs of education, and analyzing school accessibility in terms of geographical distance and financial costs (using mainly quantitative data). In administration and service delivery, bureaucrats, executive and legislative, they were in optimistic mode in collaborating education policy. Those four districts conducted various policies in fostering education. However, outside those parties (bureaucrats, executive and legislative), principals, and teachers as main actor in the field had ambiguous opinion on education policy. In term of education budget, most of district has allocated significant amount of budget range from 30-40% and only Kutai Kertanegara has allocated 4.5% of its budget for education expenditure. And Kutai Kertanegara had significant amount of budget among 4 districts because Kutai Kertanegara had huge revenue shares from natural resources. Press and NGOs were groups that voiced opposite ways. These groups were pressure-group in local area that voiced transparency and accountability in managing education policy.

Uniquely, in quality of education most parents said that education quality was improved after decentralization but most of them also cannot point what indicator to point quality of education. For authors it was unclear evidence. Costs of education were considered had tendency to increase year after year and also after decentralization. Most accessibility to school in four districts study was low. Because of this reason, there was excess to pull out their children from school and help their parents to work (in Bantul especially in sand mining area).

They concluded that there were excess or negative impacts of decentralization policy especially in education. Firstly the devolution of power was not followed by transparency and accountability that could hinder the education policy itself. Secondly most people in those four districts implied that the quality of education was increased but they cannot point what indicator of quality of education. Third, the cost for obtaining education has increased each year. Fourthly accessibility to school was enormous for rural area and remote area people had face obstacles to access school especially in Kabupaten Ngada in Nusa Tenggara Timur province (one of the poor province in Indonesia). They offered policy recommendations which are increasing of government budget in education expenditure and considered as investment not subsidy. Secondly central government spending for education especially in poor province is still needed and together with awareness of transparency and accountability. And the last, they thought it is needed of political education to increase awareness of transparency and accountability in managing of education in local region.

They had conducted quite fair research by spelling its research limitations and not depended on government data. And also they conducted depth discussion in focus group discussion that could obtain real facts in the field. However in policy recommendations are little bit bias because they tend to blaming central government. Author of this thesis argued that decentralization policy means that each of parties central government and local government had their own obligation to managed their own affairs. It is fair to recommend that central government should involve in education mostly in poor province not all provinces.

Bayhaqi (2004) in Decentralization in Indonesia: The Possible Impact on Education (Schooling) and Human Resource Development for Local Regions examined conception of decentralization from political perspective. Author agrees that decentralization is not automatically established well in Indonesia without any supporting derivative policy in finance and political reformed also.
Bayhaqi (2004) considerably discussed widely across some sector that could avoid if he focus on his discussion on decentralization and education. He put some indicator that comparing between Indonesia and its neighboring countries on GDP per capita, number of working forces, education sharing in GDP and so on. Even author also discussed about FDI.

Bayhaqi (2004) described four major problems that could be triggered by decentralization policy as follows. One of the main problems of decentralization, from the perspective of the society, is that it could affect public service delivery, like education and health, especially in the local regions that only have limited resources. It is also because the central government has not made clear regulations regarding the minimum standard of service for public services. The second problem, from the perspective of business, is that decentralization could increase the cost of doing business. Because the local government is being pushed to increase their local revenues, the local governments are reacting by increasing their taxes and levies. Adding more kinds of tax could discourage investment that could further hinder the development of the local governments in the long run. The response is understandable, since local governments and local legislative are narrow-minded in their policy preference. The third problem, from the perspective of policy implementation, it is apprehensive that the local governments do not have sufficient human resources to put into practice their policies. Local governments are eagerly to put more preference towards its local staff which makes sense. However the problem is how to increase the capacity of the local staff in order to increase the capacity of the local government. Fourthly, central government is still having vested interests in controlling the local governments. And most local governments are still dependant on the transfer from central government (either in the form of DAU, DAK or Revenue Sharing), during transitional period most local government policies would still be ‘centralized’ to some extent.

He was using descriptive statistic analysis to approach the relationship between decentralization and education or human resources development. He put some data on education to examine the effect of decentralization on education. But Author observed that those data was too widely and not focused on specific education data. Overall the methodology was not clear to discover the real condition of decentralization and education in Indonesia. Bayhaqi (2004) implicitly had good idea to discuss the topic but he structured the topic inappropriately.

In conclusion part, Bayhaqi (2004) was not clearly answered the main topic of his article. Though he provided solution to education and not discussed properly on decentralization and education. He stated that “ One possible solution in securing education finance probably is by encouraging public participation. The fact that the share of education expenses is relatively low, parents could be encouraged to provide more resources in financing their children’s education. However, parents might be reluctant to do so if they feel that education (especially primary and secondary) is under the responsibility of the state. Local government also could be reacted negatively to increasing public participation in education financing by channeling the budget to other unproductive areas.”

At some points, Bayhaqi (2004) did not understand accurately especially in how local government finance its education budget such teacher salary and so on. Central government does not leave automatically local government as decentralization occurred. There is one responsibility when local government is received authority to manage their affairs locally it is axiom “money follow function”. So when local government had authority politically from election central government also transfer fund to local government to finance its affairs that in decentralization regime became local government responsibility such as education, health, transportation and so on. So “the real picture” is not “spooky” as author described, decentralization is not pure independently to manage locally affairs. There is still exist coordination between central government and local government.

Though points in describing problems, Bayhaqi (2004) has given new perspective to let know author of this thesis that decentralization could be possible generated some problems for Indonesia especially in education. So from these points, author can explore more and catch main problem and provides new idea for other studies.

3. Doubts about Fiscal Decentralization

In searching of empirical work that objectively criticized of theory fiscal decentralization, this thesis found one article that fairly observed and doubted theory of fiscal decentralization. Prud'homme (1994) in On the Dangers of Decentralization has done deeply critics the implementation of fiscal decentralization that had strong root in developed country i.e. United States not automatically fit for developing nations to implement.

Central point of theory of fiscal decentralization lays on efficiency of allocation of resources. To fulfill efficiency of allocation, allocation of resources should follow people preferences in which are different in each municipal. Prud'homme (1994) questioned how local leader that elected by people could automatically understand and know people preferences. And the most important is people vote not in line with party program that is offered during campaign. Prud'homme (1994) argued that people vote in basis of loyalty to party or figure not with program that party or political figure is offered during the campaign.

The reality in developing countries (most of them) is focusing on how to attain basic needs that could be harmful to their own people if government that has obligation to guarantee of provision of basic needs has changed the tool to decentralize system. That means below administrative decentralization, local government is fulfilling of its people preferences. The problem that is recognized by Prud'homme (1994) is the mechanism of letting know the preferences of people in developing countries through local election is hardly to conclude that local general election for electing local executive governor or mayor and member of local representative will be the mechanism in finding people preferences.

Another important point that is expressed by Prud`homme is fiscal decentralization makes burden for macroeconomics stabilization in a country. We can visualize easily: when every single province and municipal had autonomy to manage its local public finance that means they (province government and municipal government) can tax to increase revenue and can borrow money to finance their expenditure, it will be harmful for macroeconomics stabilization.
In the Dilemma of Fiscal Federalism: Grants and Fiscal Performance around the World, Rodden (2002) assessed the risk of being fiscally decentralized especially in macroeconomics. Combinations of autonomy of managing expenditure and autonomy of borrowing after fiscal decentralization would harm macroeconomics stability in a country. And it is possible to create another crisis for a country. Rodden (2002) named it as dilemma of being fiscally decentralized.
In Indonesia, local government can borrow fund to finance its budget to some resources whether issuing local bond or directly borrow from donor countries. However local government should fulfill some requirements that regulated by law so that this borrowing activity not harms macroeconomics stability. The regulation is lesson learnt from what happened in Brazil when many local governmentsborrowed fund that is out of limit of their fiscal capability and created crisis.

4. Overview of Education in Developing Countries

Education is a part of development that is emphasized by many economics theory. Simply credo of economics theory mentions that workforce composition that consists of higher level education would foster development in a country. Therefore, many developing countries are investing significant of fund from their budget to improve their education to foster their development. However, there are still lack

Chilcott (1987) A Critique of Recent Models for the Improvement of Education in Developing Countries are reviewing model of improving education output in developing countries. Twofold models of improvement of education in developing countries are explanatory model and process model.

Explanatory model is focusing on mixing of formal education of western style and informal education. Informal education appears influential because it is embedded from real condition of developing countries that consist of different tribal and different culture. Whereas process model emphasized on highly implementing value of western education in massively reforms regardless there are different background of developing countries and western value.

Glewwe and Kremer (2005) in Schools, Teachers, and Education Outcomes in Developing Countries highlight twofold approaches in improving education outcomes in developing countries which are to overcome education output in developing countries it needs significant amount of fund whether from countries budget or from aid and the opponent emphasizes on reform the systems and curriculum of education in developing countries.

Major issues in improving education in developing countries are whether to increase quantity of school physically or to increase quality of education means that improvement of education curriculum, quality of teacher and providing books. First attempt to improve education in developing countries is increasing number of school physically that also including access infrastructure to school and location of the school. Increasing number of school intends to span number of pupils in rural area to participate in school. At this stage, some problems are arisen. Before thinking of and going to school, most of pupils must solve some constraints e.g. helping their parents at work and cost of attending school such as uniform. Also there is attempt to combine with meal at school. Fulfillment of basic needs in developing countries especially in rural area namely food is more important than education as observed by Glewwe and Kremer (2005).
Second attempt is by improving from quality aspect of education which means that improving pupils test score. There are a lot of choices in order to achieve quality of education in developing countries. Those are by increasing ratios of teacher and students, providing meal supplement, providing textbooks, up grading school tools for example computer and radio for mathematic and improving school environment for example placing school in serene environment.

To sum, improvement of education quantitatively and qualitatively in developing countries especially in rural area are dealing with enormous constraints as emphasized by Glewwe and Kremer (2005). Author argues that for rural area in Indonesia basic need fulfillment, availability of teachers and access to school are most important problems. Crux of provision of education in developing countries is not always relating with education itself but also relating to other aspects outside education.