WPS6075 Policy Research Working Paper 6075 General Purpose Central-provincial-local Transfers (DAU) in Indonesia From Gap Filling to Ensuring Fair Access to Essential Public Services for All Anwar Shah Riatu Qibthiyyah Astrid Dita The World Bank Poverty Reduction and Economic Management Unit World Bank Office, Jakarta June 2012 Policy Research Working Paper 6075 Abstract Indonesia has come a long way from centralized presents illustrative simulations of alternative programs governance to decentralized local governance, and and compares these with the existing Dana Alokasi today Indonesia ranks among the most decentralized Umum allocations. developing countries. The Government of Indonesia The paper concludes that super complexity leads to lack is revisiting all aspects of local governance to make of transparency, inequity, and uncertainty in allocation. appropriate legal and institutional adjustments based on Simpler alternatives are available that have the potential lessons leaarned during the past decade. An important to address autonomy and equity objectives while also area of this re-examination and possible reform is the enhancing efficiency and citizen-based accountability. central financing of subnational expenditures. The system Such alternatives would represent a move away from the of intergovernmental finance represents one of the most complex gap-filling approach to simple output-based complex systems ever implemented by any government transfers to finance operating expenditures. Capital grants in the world. The system is primarily focused on a gap- would deal with infrastructure deficiencies. And the filling approach to provincial-local finance in an objective alternatives would institute fiscal capacity equalization manner to ensure revenue adequacy and local autonomy as a residual program with an explicit standard to ensure but without accountability to local residents for service that all local jurisdictions have adequate means to delivery performance. This paper takes a closer look at deliver reasonably comparable levels of public services at Dana Alokasi Umum—the most dominant program of reasonably comparable levels of tax burdens across the unconditional central transfers to finance provincial-local country. government expenditures in Indonesia. The paper also This paper is a product of the Poverty Reduction and Economic Management Unit, World Bank Office, Jakarta. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at shah.anwar@gmail.com. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team General Purpose Central-provincial-local Transfers (DAU) in Indonesia: From Gap Filling to Ensuring Fair Access to Essential Public Services for All Anwar Shah (World Bank, SWUFE,China and ADB) , Riatu Qibthiyyah ( University of Indonesia) and Astrid Dita (University of Indonesia) Key words: grants, fiscal transfers, fairness, equalization, public services, provincial- local finances, decentralization, results based accountability, incentives, Indonesia reforms Sector Boards: EP, PSG, PR, ARD, URBAN, INFRASTRUCTURE, SOCIAL, SDN, HDN General Purpose Central-provincial-local Transfers (DAU) in Indonesia: From Gap Filling to Ensuring Fair Access to Essential Public Services for All Anwar Shah 1(Center for Public Economics, China, World Bank and the Asian Development Bank), Riatu Qibthiyyah (University of Indonesia) and Astrid Dita (University of Indonesia) 17 April 2012 1. Introduction Provincial and local government reforms continue to dominate the national policy agenda today in Indonesia inspite of the sweeping legislative and administrative changes affecting the organization, functions and finance of sub-national government undetaken since 1999. During the past decade, Indonesia has come a long way from a centralized governance to decentralized local governance. By the big bang in 2001, Indonesia has leap-frogged to the community of decentralized nations, and today Indonesia ranks among the most decentralized developing countries (see Ivanyna and Shah, 2011; see Figure 1). With this transformation, the Central Government still accounts for 91% of revenue collection and 64% of direct spending (2011 figures). Provinces and local governments account for 9% of revenue collection and 36% of expenditures. Provincial governments are relatively less important than local governments and command only 9% of national expenditures. Local government expenditure share (27%) in Indonesia compares favorably with most developing and industrial countries (see Figure 2). 1 This paper was presented at the World Bank/GOI-MOF Conference on ‘Alternative Visions for Decentralization in Indonesia’ held at Jakarta from March 12-113, 2012 and the Asian Development Bank/GOI-MOF workshop on Intergovernmental Finance in Indonesia from April 4-5, 2012 in Bali. The authors are grateful to Pak Dr. Marwanto Harjowiryono, Pak Dr. Heru Subiyantoro, Pak Putut, Ibu Erny Murniasih, Pak Aditya Nuryuslam of the Ministry of Finance , Indonesia, Wismana Adi Suryabrata, BAPPENAS, Raksaka Mahi, University of Indonesia, Gutat Setyo Tamtomo Yudo Baroto, MOHA, Juan Luis Gomez, ADB and Blane Lewis, University of Singapore, Shubham Chauduri, William Wallace, Anna Gueorguieve, Pak Daan , World Bank and participants of the Bali workshop for helpful comments. The views presented here are those of the authors alone and may not be attributed to the institutions they represent. Comments may please be addressed to Anwar Shah (shah.anwar@gmail.com). . 2 Figure 1. Local Government Share of Total Expenditures (2005) Figure 2. Expenditure, Employment and Revenue Collection Shares by Order of Government Source: Ministry of Finance and Ministry of Home Affairs, Indonesia Indonesia quite remarkably achieved this status without experiencing any service delivery disruptions even during the early stages of this rapid transformation. Today, the Government of Indonesia is revisiting all aspects of local governance to make appropriate legal and institutional adjustments based upon lessons learned during the past decade. An important area ripe for this re-examination and possible reform is the central financing of subnational expenditures. The system of intergovernmental finance in vogue today represents one of the most complex system ever implemented by any government in the world. The system is primarily focused on a gap-filling approach to provincial-local finance in an objective manner to ensure revenue adequacy and local autonomy, but without accountability to local residents for service delivery performance. This is done through a great degree of academic rigor using highly complex procedures with the objective of providing precise justice and more importantly to keep politics at bay. Do these complex programs serve their explicitly stated objectives? This 3 paper takes a closer look at the most dominant program of central transfers to finance provincial-local government expenditures in Indonesia. The paper concludes that super complexity leads to a lack of transparency, inequity and uncertainty in allocation. Simpler alternatives are available that have the potential to address equity objectives while also enhancing efficiency and citizen-based accountability. Such alternatives would represent a move away from complex gap filling and special allocation approaches to simple output based transfers to finance operating expenditures, complemented by capital grants to deal with infrastructure deficiencies and instituting fiscal capacity equalization as a residual program with an explicit standard to ensure that all local jurisdictions have adequate means to deliver reasonably comparable levels of public services at reasonably comparable levels of tax burdens across the country. The paper argues that such an alternative system of intergovernmental finance would preserve autonomy, while enhancing equity, simplicity, objectivity, transparency and accountability. The rest of the paper is organized as follows. Section 2 provides a bird’s eye view of central transfers to sub-national governments. Section 3 presents a critical review of general allocation transfers (Dana Alokasi Umum—DAU). Section 4 presents simulations of alternative approaches to such transfers. Section 5 discusses the merits and demerits of a popular proposal to set up an independent grants commission. A final section presents concluding remarks and presents a forward-looking view about the reform of intergovernmental finances in Indonesia. 2. Central Transfers to Finance Provincial-Local Public Services in Indonesia – A Review Central transfers to provincial-local governments in Indonesia—a brief synopsis Central transfers are the most important source of revenues for sub-national governments in Indonesia. These financed 90% of sub-national governments, 54% of provincial, 86% of cities and 93% of districts expenditures in 2010. Major transfers (Balance Grants or Dana Perimbangan) to finance provincial and local expenditures are as follows. Table 1. Central-Provincial/Local Transfers in Indonesia (2010) Transfer Share of total transfer in Share of sub-national expenditures in 2010 2010 Tax sharing 25% 20% Gap filling (DAU) 56% 46% Special Allocation Grant 6% 5% (DAK) Other specific purpose 13% 10% All 100% 90% (Provinces: 54%; Cities: 86% and Districts: 93%) Source: Ministry of Finance, Indonesia Tax by tax sharing (Dana Bagi Hasil—DBH). Central Government collects taxes on personal income, property, and renewable aand non-renewable natural resources and returns by origin a pre-defined share of the revenues to the originating jurisdiction. These transfers accounted for 25% of total central ransfers in 2010 and financed 20% of sub-national expenditures. 4 Transfers to deal with vertical and horizontal fiscal gaps. Central government provides a basic allocation for wages and salaries and a fiscal gap transfer (Dana Alokasi Umum or DAU) if a jurisdiction’s revenues fall short of calculated expenditure needs using macro indicators. These transfers accounted for 56% of total central transfers and financed 46% of sub-national expenditures. Specific Purpose Grants. These grants include the Special Allocation Grant (Dana Alokasi Khusus or DAK), Special Autonomy grants for Aceh, Papua and Papua Barat, Adjustment Fund compensation, and Special Incentives Grants (Dana Insentif Daerah or DID) and Hibah. DAK is intended to influence local government spending on areas of national priority. It accounts for 6% of central transfers and finances 5% of sub-national expenditures. Adjustment Fund compensation (Dana Penyesuaian or DP) provides special ad-hoc assistance e.g. for school operational assistance (BOS), allowances for certified teachers etc. Special Autonomy grants (Dana Otonomi Khusus or DOK) are intended to provide special and preferential support to Aceh and Papua provinces. DID is a small grant program accounting for less than 1% of total transfers and are granted to better performing provinces and cities on public financial management, tax effort, having higher HDI relative to fiscal capacity, higher economic growth, higher reductions in poverty, unemployment and inflation. Hibah transfers are primarily financed by external assistance and are intended to finance sub-national infrastructure and social development expenditures. Specific purpose transfers in total accounted for 19% of central transfers in 2010 and financed 15% of sub-national expenditures (see Qibthiyyah, 2011). The following paragraphs present a review of the general-purpose transfer (DAU). 3. The General Purpose Gap Filling Transfer—Dana Alokasi Umum (DAU): An Introductory Overview The general-purpose unconditional transfers, DAU, constitute the dominant sources of revenues for provincial and local governments in Indonesia. As part of the DAU transfers, the Central Government of Indonesia provides a basic allocation for wages and salaries and a fiscal gap transferif a jurisdiction’s revenues fall short of calculated expenditure needs using macro indicators. These transfers accounted for 56% of total central transfers and financed 46% of sub-national expenditures in 2010. These transfers according to Law 33 (2004) are intended to balance revenue means with expenditure needs for sub-national governments providing central financing in ―proportionate, democratic, fair and transparent manner‖ by taking into account ―local potential (fiscal capacity) and conditions and local needs‖. Total pool of these transfers is arbitrarily set at of 26% of central revenues net of tax sharing transfers in 2011. The 20% of the total pool is allocated to provinces and the remaining 80% to all cities and districts. The DAU provides a basic allocation to cover wages of provinces cities and districts. The remaining funds are allocated by formula that determines fiscal gap based upon the differences between fiscal needs and fiscal capacity. Formula factors for both provinces and cities are the same but receive differential weights due to the peculiar application to DAU allocation of the weighted coefficient of variation- the so called Williamson’s Index. Fiscal capacity of a province is determined by summing up 50% of own source revenues, 80% of non-resource tax sharing and 95% of resource and mining tax sharing. Fiscal capacity of a city or district government on the other hand is based upon 93% of own source revenues, 100% of non-resource tax revenue sharing and 63% of resources and mining tax revenue sharing. 5 The weights for individual revenue sources to determine fiscal capacity varies from year to year as weights are picked up to achieve a given numerical value for the Williamson’s index for each year. Table 2 provides these choices for numerical values of the index for the past few years. Table 2. Williamson’s Index as Equalization Standard in Indonesia: 2005 -2011 2005 2006 2007 2008 2009 2010 2011 Provinces 0.941 0.769 0.975 0.793 0.802 0.836 0.801 Cities and 0.630 0.678 0.699 0.710 0.690 0.718 0.694 Districts Source: Ministry of Finance, Government of Indonesia Fiscal needs of provinces and cities/districts are determined separately for each of this group by developing a composite index based upon relative population, relative area, relative construction price index, inverse of human development index (comprising arbitrary weights for life expectancy, literacy rate, mean years of schooling and purchasing power adjusted relative real GRDP per capita) and inverse of relative nominal per capita GRDP. The weights for the above mentioned factors vary for provinces and districts/cities and over time for each group based upon the specified value to be achieved for the Williamson’s index (see Table 3). The resulting indexes are multiplied by the average aggregate spending for the past year to arrive at numerical values of the expenditure need component. DAU allocation for each jurisdiction is then determined as follows: The DAU is a gross program and compensates a jurisdiction for excess needs but does not tax regions with excess fiscal capacity. The jurisdictions displaying negative fiscal gap (surplus fiscal capacity) e.g. Jakarta metropolitan region receive only the basic allocation and the negative fiscal gap is ignored. Table 3. Williamson’s Index Determined Weights for Fiscal Capacity and Expenditure Need Factors in DAU allocations Province DAU 2006 DAU 2007 DAU 2008 DAU 2009 DAU 2010 DAU 2011 DAU variable weight Fiscal Need Variables Population 30.00% 30.00% 30.00% 30.00% 30.00% 30.00% Area 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Construction price 30.00% 30.00% 30.00% 30.00% 30.00% 30.00% index Inverse of human 10.00% 10.00% 10.00% 10.00% 10.00% 10.00% development index Index of Inverse per 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% capita GRDP Fiscal Capacity Variables Own source revenue 50.00% 50.00% 50.00% 50.00% 50.00% 50.00% Tax revenue sharing 100.00% 75.00% 75.00% 95.00% 73.00% 80.00% Resource revenue 100.00% 50.00% 41.25% 70.00% 95.00% 95.00% sharing 6 Cities/Districts DAU 2006 DAU 2007 DAU 2008 DAU 2009 DAU 2010 DAU 2011 DAU variable weight Fiscal Need Variables Index of population 30.00% 30.00% 30.00% 30.00% 30.00% 30.00% Index of area 15.00% 15.00% 15.00% 15.00% 13.25% 13.50% Index of construction 30.00% 30.00% 30.00% 30.00% 30.00% 30.00% price index Index of inverse of HDI 10.00% 10.00% 10.00% 10.00% 11.00% 10.00% Index of inverse per 15.00% 15.00% 15.00% 15.00% 15.75% 16.50% capita GRDP Fiscal Capacity Variables Own source revenue 100.00% 75.00% 75.00% 70.00% 93.00% 93.00% Tax revenue sharing 100.00% 75.00% 75.00% 73.25% 100.00% 80.00% Resource revenue 100.00% 50.00% 50.00% 100.00% 100.00% 63.00% sharing Source: Ministry of Finance, Government of Indonesia DAU – An Evaluation In the interest of tax harmonization, limiting differentials in sub-national fiscal capacities and lowering tax administration costs, Indonesia has adopted a highly centralized tax system with the Central Government raising 91% of total revenues. This creates a large vertical fiscal gap (almost 90%) that is filled by revenue sharing and transfers. Revenue sharing by origin while reducing vertical fiscal gap accentuates horizontal fiscal inequalities. DAU is the foremost program to bridge horizontal inequities. It is an objective formula based transfer program that partially compensates for civil service wages and partially tries to limit the differentials the fiscal capacity across jurisdictions by focusing on reducing the variations in regional allocation of transfers as measured by the weighted coefficient of variation. This results in reducing overall inequality in fiscal capacities and some redistribution of income across provinces and cities and districts as shown by Eckardt and Shah (2007). However, the current program has a number of important limitations. One size fits all approach leads to fiscal inequity.The foremost concern is that the program equalizes jurisdictions with widely dissimilar responsibilities and characteristics. This is especially true when you group metropolitan areas, cities of varying population sizes and rural municipalities or districts of varying geographical areas as done under the current program. This violates the most fundamental dictum of transfers that ―one size does not fit all‖. The Constitutional Court of Indonesia in South Sulawesi case had earlier ruled that, ―Uniform treatment of different entities causes injustice‖. Indeed it would be a travesty of justice to consider that a small town with a tiny population such as Kabupaten Puncak has the similar fiscal needs and capacities as a large city like Kota Bandung or for that matter a small district in terms of area, Tangaran has similar revenue means and expenditure needs as a large district of Tangerang. The existing program ignores the fiscal capacity or fiscal need differentials of various size and class of municipalities and assumes that they all have equal per capita needs and revenue means other than from those factors explicitly considered in the formula. If one examines local finance in other countries, there are wide justifiable variations in per capita revenues and spending across various size and urban/rural class of municipalities in view of the diversity of needs and preferences and responsibilities. Equalizing unequals leads to injustice 7 for all. One cannot possibly have the same standard and access and diversity of services in a small remote district as opposed to a large city. A complex and opaque standard of equalization. A second important concern has to do with the choice of the Williamson’s index as the equalization standard. Most industrial countries adopt a simple, transparent but an explicit standard to reach a broad political and social consensus on overall amount of equalization payments (see Table 4). This is important because equalization programs can have important efficiency and equity tradeoffs. An excessive standard of equalization can lead to adverse impacts on growth just as too little equalization can create potential for succession. While equalization standards vary in terms of their relative emphasis on fiscal capacity versus fiscal need equalization, all bear some affinity to providing reasonably comparable levels of public services at reasonably comparable levels of tax burdens across all jurisdictions to ensure a common political and economic union. The central focus of an equalization program is to help disadvantaged jurisdictions have comparable public service standards to allow them to integrate with the wider economy. Indonesia is unique in selecting complex statistical criterion—the weighted coefficient of variation or the Williamson’s index— as the equalization standard. This choice is unfortunate as it introduces complexity and lack of transparency in the allocation criteria. Further the Williamson’s index has relatively greater sensitivity to outliers. For example, it is possible to redistribute income among the two top quintiles and have a lower value of the Williamson’s index while there may be no significant redistribution to the poorest quintile. Its use in determining factor weights is particularly worrisome as multiple distributions of component weights can yield the same index. The use of this index in determining factor weight introduces uncertainty and inequity in allocations as without any material changes in need and capacity factors, changes in weights alters the allocation of transfers across jurisdictions. In a developing country context, complexity is sometimes cited as a way to hold the politics at bay as policy makers may not fully comprehend the limitations of a complex design and may hold fire. The Indonesia program cannot be justified on this basis as Table 3 shows that while policy makers may not understand the working of such an index, they have not refrained from forcing the choice of higher variations in inequality in a subsequent year if the resulting allocations lead to more satisfactory outcomes for their jurisdictions of interest. Table 4. The Choice of An Equalization Standard: Indonesia in the International Context Standard of None (general National Average Complex statistical Equalization revenue sharing) (Per Capita) criteria Determines grant pool Determines allocation UK and most Australia, China, Indonesia (weighted to jurisdictions developing countries Russia, Switzerland Coefficient of e.g. Brazil, India, Variation – the so- Thailand called Williams’ Index) Determines both the Canada, Germany, pool and the alloca- Finland, Denmark, tion Sweden An erroneous view of fiscal capacity. A third concern has to do with how fiscal capacity is measured. Various sources of revenues are given arbitrary and differential weights for provinces and cities/districts and revenues from specific purpose transfers (DAK) are excluded. This gives an erroneous view of fiscal capacities of various jurisdictions. 8 Discouraging local tax efforts. The use of actual revenues as opposed to potential revenues creates disincentive effects for own tax effort. Jurisdiction may not have incentive to improve collection of own revenues as any increase in own tax effort at the margin is mostly offset by decrease in DAU entitlements (see Table 5). There is some evidence to show that such reduced tax effort is indeed happening. Table 5. Share of Own Revenue to GRDP 2008 and 2010 Type of Regions Correlation TE 2010 with DAU 2010 Provinces -0.1385 Districts -0.2075 Cities -0.1928 Notes: TE is tax effort refers to share of Own Revenue to GRDP. Source: calculated from Ministry of Finance, Indonesia From Figure 3 and 4, majority of jurisdictions experienced lower tax effort in 2010 in comparison to 2008. All provinces have lower tax effort in 2010 in comparison to 2008, as shown in Figure 3. For some of the provinces, such as Kalimantan Timur, Papua, and Riau, decline in share of own source revenue to GRDP is quite high, and it may have arisen from a change and higher weight of fiscal capacity for resource sharing taxing in DAU formula. In addition to low growth of own source revenues, revenues from resource sharing in resource rich provinces is also taxed more than other type of revenues. Figure 3. Proportion of Provincial Own Source Revenue to Gross Regional Domestic Product in 2008 and 2010 A similar pattern of a decline in tax effort also occurred at districts and cities. Figure 4 shows that on average, share of own source revenue to GRDP have declined in 2010 from 2008 levels in both districts and cities. 74 percent of districts and 84 percent of cities have lower tax effort in 2010 in comparison to 2008. 9 Figure 4. Average Proportion of Own-Source Revenue to GRDP 2008 and 2010: Provinces (Provinsi), Districts (Kabupaten), and Cities (Kota) Undermining agreements with special autonomy regions. The Government of Indonesia (GOI) has entered into special arrangements with Aceh, Papua and Papua West and allows them a greater share of resource revenues through the tax sharing system. The DAU offsets a large part of those gains by including 95% of those gains as increases in fiscal capacity for the provinces and 63% for cities and districts. Thus the GOI taxes back most of the gains to resource rich regions through the operation of the DAU formula. Incentives for local governments to serve as employment creation agencies to the neglect of their role as service providers. The basic allocation provides financing for public sector wages. This creates incentives for padding up local rolls. Such perverse behavior in Indonesia is circumvented by central controls over local recruitment and staffing. But this takes away local autonomy for hiring and firing and setting terms of employment of local employees. It also ties local governments to the personnel policies of the central government taking away any incentives they might have to experiment with new public management paradigms to improve accountability for servicer delivery performance e.g. through contracting out or partnership arrangements within and beyond government agencies. In short, wages compensation creates an incentives and accountability regime that works against good local governance. Inappropriate indicators of fiscal need. Beyond basic allocation, formula based expenditure needs determination as done in Indonesia has important limitations. Regional per capita income is used twice as a need factor—real per capita GDP adjusted for purchasing power parity in the formation of the Human Development Index and nominal per capita GDP more directly. Regional per capita income is an imperfect measure of fiscal capacity but not a very useful measure of fiscal need. The inclusion of resources and mining based GDP in both concepts of income inflates the fiscal capacity of resource rich local jurisdictions although significant portions of these incomes may accrue to foreigners and non-residents. It also undermines special autonomy agreements with resource rich provinces. Further local jurisdictions may have limited and partial access to taxing these bases as is the case in Indonesia. Expenditure need determination uses both fiscal capacity and fiscal need factors that work at cross purposes. Further other than population and area, indicators used have little or only remote relationship to service needs. There is also multiple hierarchy of arbitrariness in determining relative factor weights. First the HDI index uses arbitrary weights for life expectancy, literacy rate, and mean years of schooling and per capita GDP. A second degree of arbitrariness arises from the 10 application of the Williamson’s index as multiple distributions of relative weights can lead to the same value of the index. The use of the Williamson’s index in determining formula factor weights also causes complexity, non-transparency, uncertainty and inequity in individual allocations. Individual jurisdictions entitlements can change from year to year and relative to others for no apparent justification. Non-neutrality to amalgamation and incorporation decisions.The expenditure need determination component formula is also non-neutral as to the amalgamation and incorporation decisions. Using composite index indicator and referencing the transfer not to per capita but to average of total fiscal need or fiscal capacity may contribute to non-neutrality to fragmentation of jurisdictions. Amalgamation of existing jurisdictions leads to lower central transfers for amalgamating jurisdictions and break up of existing jurisdictions benefits all in terms of higher per capita central transfers (see Table 6 from Marwanto Harjowiryono, 2011). No wonder, three new provinces have been created and the number of cities/districts mushroomed from 336 in 2001 to 502 in 2010. Table 6. Existing DAU Allocation Rewards Jurisdictional Fragmentation Province Number of Number of Total DAU for Total DAU for % change Districts/Cities Districts/Cities Districts/Cities Districts/Cities in DAU: in 2001 in 2011 in 2001 in 2011 2001-2011 (billion Rp) (billion Rp) Kalteng 6 14 0.9 5.5 528% Yogyakarta 5 5 0.9 2.7 216% Source: Marwanto Harjowiryono (2011) Inequity for large urban and rural jurisdictions. Finally and most importantly expenditure need determination uses a one size fits all approach and assumes per capita fiscal needs of large cities are similar to those of a small town or a rural district when it uses aggregate average spending for all local governments as a reference point. This creates tremendous injustice for large urban and large rural areas. Complex, non-tranparent and inequitable allocations. In sum, the gap filling approach is unnecessarily complex, non-transparent and uses a macro approach that is not well grounded in the local realities to ensure inter-jurisdictional equity. These manna from heaven transfers also create an incentive and accountability structure that is not conducive to responsible, responsive, fair and accountable local governance. Simpler alternatives as outlined in the following paragraphs have the potential to enhance efficiency and equity of such transfers mechanisms. 4. DAU Simplification Reform Alternatives In the following three alternative options for the reform of DAU are presented. Common elements of these three alternatives are:  One size does not fit all. Requires grouping (clustering) of local governments by population size, area and class of local governments.  The adopted formula must have a sunset clause of five years but interim changes in the formula should not be allowed.  The formula must have ceilings and floors to keep yearly entitlements stable and predictable. 11  National average (by size and class) gap filling or equalization standard to replace Williamson’s index. Pool can be adjusted by affordability and allocation determined by standard.  Gap filling or equalization by size and class of local governments and possibly equal per capita grants to villages.  Fiscal capacity measurement based upon potential revenues plus tax shares and other transfers and 50% of resource revenues.  Fiscal needs measurement to discontinue wages compensation (basic allocation), discontinue use of HDI and the Williamson’s index and other indexes. Instead consider need measurement based upon service/client population for each service category. Alternative 1: A Simpler Yet More Objective Gap Filling Approach To simplify DAU and to ground it in local comparative context, one requires meaningful grouping of local governments. One alternative in this regard is to have the following classes or clusters. Please note that the use of sophisticated statistical techniques in clustering as recently advocated by the World Bank and the Indonesian academics simply adds more complexity without having any clear advantages to a simpler and transparent approach advocated here. Table 7. Characteristics of Each Cluster: Province, C1-C4, and D1-D4 Cluster No. Total Min. Max. Group Min. Max. Group Per Per capita of popu- pop. pop. ave- area area average capita 2009 juris- lation (mil.) (mil.) rage (km2) (km2) area 2009 expendi- dicti- (mil.) pop. (km2) own ture ons revenues (IDR) (IDR) Provinces 32 228 0.76 43.1 7.12 3,133 319,036 59,696 164,181 666,348 Cities C1 11 19.2 1.02 2.63 1.74 119 683 302 148,971 853,093 C2 13 8.8 0.52 0.88 0.68 39 2,938 589 168,587 1,405,204 C3 58 12.4 0.10 0.50 0.21 16 2,400 352 189,369 2,493,247 C4 11 0.7 0.03 0.10 0.07 23 9,565 1,426 221,736 4,336,555 Districts D1 100 71.3 0.01 3.68 0.71 425 1,891 1,232 100,068 1,725,989 D2 99 52.1 0.01 4.09 0.52 1,901 3,802 2,710 94,200 2,345,628 D3 99 34.4 0.01 2.43 0.35 3,803 7,155 5,218 122,226 3,383,613 D4 100 26.1 0.02 1.54 0.26 7,248 49,958 19,603 202,034 5,431,698 Notes: Cluster classification: Provinces—one group: P1: Provinces excluding Jakarta. Cities (Kota)—4 groups: C1 with population over 1 million; C2 with population 500K to 1 million; C3: Population 100- 500 K; C4: Population under 100K. Districts (Kabupaten)—4 groups: D1: vast area—top quartile in area; D2: large area—2nd quartile in area; D3: medium area—3rd quartile in area; D4: small area—4th quartile in area. 12 Please note that urban kabupatens that are part of large metropolitan areas may be treated just like cities (kotas) and grouped with cities. Meanwhile, the fiscal capacity and fiscal needs would be calculated as follows: Fiscal capacity is defined to include potential revenues from own sources (PAD) if the jurisdiction applied group average tax effort to own bases plus tax shares and transfers (DBH Pajak) plus potential natural resource revenues (DBH SDA) plus other grants. A simple way to calculate potential own source revenues would be to apply national average effective tax rate to local non- resource GRDP (in this simulation, it is proxied by using non-oil and gas GRDP). Potential resource revenues will be similarly calculated by applying national average effective resource tax rate to local resource based GRDP only (in this simulation, it is proxied by using oil and gas GRDP). Due to instability of resource reveues, and higher public expenditure needs with resource exploitation and exhaustible nature of some resources only 50% of resource revenues will be counted towards revenue capacity of resource rich regions as done in Canada. Revenues from specific purpose grants will be fully included. Capital grants and loans earmarked to finance spefic projects to finance centrally determined infrastructure deficiencies will be excluded from such calculations. Fiscal need is calculated for a representative expenditure system of about 10 functions comprising most of local operating expenditures. This expenditure system will be dfferentiated by size class of local governments and will have service population indicators as determinants with weights based upon aggregate group local government expenditure for the specified function based on a 3 or 5 years group moving average. Table 8 provides illustrative expenditure functions, service indicators and weights based upon 2008 expenditures only. Since the data for public housing is unavailable, its weight is distributed to other field of services. The allocation for Alternative 1 is then determined by a simple gap-filling for the per capita fiscal gap, that is the difference of per capita fiscal capacity and per capita fiscal need. The per capita fiscal gap for each region is multiplied by the service population to give the fiscal gap or DAU allocation of the first alternative. Total allocation of DAU is, however, capped at 26% of net GOI revenues as prescribed in the law. This means a proportionate reduction of grant funds for all jurisdictions based upon the availability of funds factor. Table 8. DAU Fiscal Need Compensation: Alternatives—Provinces, Cities (C1…C4), Districts (D1…D4) Expenditure function Need Indicator Weight (%) Provinces Cities Districts General administration population (66.6%) and area 52.1 29.3 35.0 (33.3%) Law and order population 0.6 Law, order, and fire protection 1.0 1.0 Education school age population 7.3 30.4 20.0 Health weighted population of age group 6.0 9.2 17.5 with higher weights for age groups 0-5 and 65+ Social protection and welfare no. of unemployed 1.0 2.3 1.5 Housing no. of people in public housing or Annulled Annulled Annulled no. of units of public housing Roads and transportation km of roads 10.0 10.0 10.0 Agriculture and forestry area 6.0 13 Water and sewer no. of residential and 9.2 commercial/industrial units Other services population 17.0 8.6 Rural services area 15.0 Results from simulations of this alternative are presented in Table 9 for provinces, in Table 10 for cities and in Table 11 for districts. Figures 5-13 report the redistributive impact of various alternatives. These tables and figures show that on most indicators of fairness, Alternative 1 offers superior results for most jurisdictions. These superior results in fairness are achieved while having simpler, more meaningful and transparent allocation of resources. The suggested refinements of the existing DAU offers significant improvements over the existing program. These include simpler and more meaningful and easily understood indicators. The deteremination of allocation by group leads to equal treatment of equals. Factor weights are objective and would be stable as these are determined by taking moving average of aggregate spending by the group as a whole. There is more clear and transparent need equalization. Both pool and allocation are determined by formula. Total pool , however, can be constrained by affordability. Nevertheless, there is one major drawback of the proposed design— unconditonal fiscal gap compensation strengths autonomy but without accountability to local residents. A second alternative retains this drawback but moves the current program from gap filling to an equalization program. Table 9. Results for the Provinces Group under Various Simulation Scenarios* Current DAU Alternative 1 Alternative 2 Alternative 3 Formula No. of Observed Provinces 32 32 32 32 No. of Receiver 31 30 10 32 Average DAU allocation 622,547 979,405 693,084 905,339 (million IDR) Average per capita DAU 209,547 236,475 379,869 171,008 allocation (IDR) Maximum per capita DAU 796,794 956,711 1,347,027 306,176 allocation (IDR) Minimum per capita DAU 10,629 63,671 22,309 75,320 allocation (IDR) Max to min ratio of per 74.96 15.03 60.38 4.07 capita DAU allocation Maximum per capita 3,617,952 3,777,869 4,168,184 3,124,595 revenue after transfer (IDR) Minimum per capita 170,194 228,693 71,730 240,342 revenue after transfer (IDR) Max to min ratio of per 21.26 16.52 58.11 13.00 capita revenue after transfer Coef. of variation 1.031 1.102 1.447 0.944 Weighted coef. of variation 1.051 1.013 1.411 0.838 Relative mean deviation 0.310 0.338 0.431 0.288 14 Gini Index 0.423 0.428 0.546 0.371 Weighted Gini Index 0.381 0.332 0.435 0.293 Notes: *) Excluding non-receivers, inequality measures calculated for after transfer per capita revenue 15 Table 10. Results for the Cities Group under Various Simulation Scenarios* Current DAU Alternative 1 Alternative 2 Alternative 3 Formula No. of Observed Cities 93 93 93 93 No. of Receiver 91 92 59 93 Average DAU allocation (million) 318,295 346,842 62,647 397,776 Average per capita DAU allocation 1,404,176 1,226,520 234,929 1,372,161 Maximum per capita DAU 7,301,673 2,594,625 530,996 3,242,229 allocation (IDR) Minimum per capita DAU 147,101 232,243 18214 379,163 allocation (IDR) Max to min ratio of per capita 49.64 11.17 29.15 8.55 DAU allocation Maximum per capita revenue after 9,943,591 5,115,631 5,769,565 7,029,313 transfer (IDR) Minimum per capita revenue after 510,582 671,978 283,775 736,689 transfer (IDR) Max to min ratio of per capita 19.48 7.61 20.33 9.54 revenue after transfer Coef. of variation 0.611 0.437 0.742 0.467 Weighted coef. of variation 0.598 0.488 0.739 0.507 Relative mean deviation 0.208 0.154 0.228 0.160 Gini Index 0.299 0.230 0.327 0.235 Weighted Gini Index 0.291 0.249 0.312 0.250 No. of Observed C1 Group Region 11 11 11 11 No. of Receiver 11 11 5 11 Average DAU allocation (million) 599,780 772,744 171,417 926,020 Average per capita DAU allocation 352,450 449,996 114,966 537,202 Maximum per capita DAU 506,550 543,150 170,570 634,337 allocation (IDR) Minimum per capita DAU 153,056 314,452 60,271 379,163 allocation (IDR) Max to min ratio of per capita 3.31 1.73 2.83 1.67 DAU allocation Maximum per capita revenue after 1,233,261 1,346,061 985,273 1,473,621 transfer (IDR) Minimum per capita revenue after 510,582 671,978 288,977 736,689 transfer (IDR) Max to min ratio of per capita 2.42 2.00 3.41 2.00 revenue after transfer Coef. of variation 0.240 0.197 0.365 0.192 Weighted coef. of variation 0.233 0.198 0.364 0.195 Relative mean deviation 0.092 0.070 0.137 0.070 Gini Index 0.126 0.101 0.183 0.097 Weighted Gini Index 0.128 0.106 0.193 0.103 No. of Observed C2 Group Region 13 13 13 13 No. of Receiver 13 13 9 13 Average DAU allocation (million) 367,534 484,375 86,086 578,064 16 Average per capita DAU allocation 544,808 710,779 128,791 851,087 Maximum per capita DAU 810,767 887,157 322,956 1,053,690 allocation (IDR) Minimum per capita DAU 147,101 330,507 49,547 642,642 allocation (IDR) Max to min ratio of per capita 5.51 2.68 6.52 1.64 DAU allocation Maximum per capita revenue after 2,932,250 3,442,057 2,811,546 3,452,077 transfer (IDR) Minimum per capita revenue after 854,038 910,819 324,900 967,542 transfer (IDR) Max to min ratio of per capita 3.43 3.78 8.65 3.57 revenue after transfer Coef. of variation 0.395 0.423 0.730 0.400 Weighted coef. of variation 0.371 0.397 0.693 0.375 Relative mean deviation 0.135 0.148 0.260 0.147 Gini Index 0.189 0.199 0.346 0.195 Weighted Gini Index 0.181 0.193 0.339 0.190 No. of Observed C3 Group Region 58 58 58 58 No. of Receiver 56 57 39 58 Average DAU allocation (million) 272,053 271,306 50,566 303,549 Average per capita DAU allocation 1,387,206 1,285,945 269,172 1,455,062 Maximum per capita DAU 2,308,378 1,526,098 519,377 1,973,908 allocation (IDR) Minimum per capita DAU 361,909 232,243 58,569 914,632 allocation (IDR) Max to min ratio of per capita 6.38 6.57 8.87 2.16 DAU allocation Maximum per capita revenue after 5,769,565 5,115,631 5,769,565 7,029,313 transfer (IDR) Minimum per capita revenue after 927,534 924,709 283,775 1,644,595 transfer (IDR) Max to min ratio of per capita 6.22 5.53 20.33 4.27 revenue after transfer Coef. of variation 0.333 0.275 0.757 0.367 Weighted coef. of variation 0.345 0.282 0.750 0.350 Relative mean deviation 0.118 0.083 0.205 0.107 Gini Index 0.168 0.124 0.299 0.150 Weighted Gini Index 0.178 0.127 0.302 0.143 No. of Observed C4 Group Region 11 11 11 11 No. of Receiver 11 11 6 11 Average DAU allocation (million) 214,035 149,815 15,376 153,297 Average per capita DAU allocation 3,557,909 2,304,629 271,529 2,385,820 Maximum per capita DAU 7,301,673 2,594,625 530,996 3,242,229 allocation (IDR) Minimum per capita DAU 2,291,657 1,991,547 18,214 1,743,818 allocation (IDR) Max to min ratio of per capita 3.19 1.30 29.15 1.86 DAU allocation 17 Maximum per capita revenue after 9,943,591 4,956,495 2,660,132 5,114,642 transfer (IDR) Minimum per capita revenue after 3,004,333 2,882,973 712,676 2,464,611 transfer (IDR) Max to min ratio of per capita 3.31 1.72 3.73 2.08 revenue after transfer Coef. of variation 0.380 0.206 0.499 0.263 Weighted coef. of variation 0.307 0.185 0.489 0.248 Relative mean deviation 0.128 0.090 0.220 0.117 Gini Index 0.171 0.108 0.260 0.135 Weighted Gini Index 0.139 0.098 0.258 0.130 Notes: *) Excluding non-receivers, inequality measures calculated for after transfer per capita revenue 18 Table 11. Results for the Districts Group under Various Simulation Scenarios* Current DAU Alternative 1 Alternative 2 Alternative 3 Formula No. of Observed Districts 398 398 398 398 No. of Receiver 391 397 126 398 Average DAU allocation (million) 370,746 357,357 90,961 371,465 Average per capita DAU allocation 1,903,119 1,278,589 1,449,887 1,136,520 Maximum per capita DAU 21,675,276 21,037,218 34,552,332 5,095,879 allocation (IDR) Minimum per capita DAU 136,837 32,246 4,320 333,629 allocation (IDR) Max to min ratio of per capita 158.40 652.39 7,998 15.27 DAU allocation Maximum per capita revenue after 36,078,212 24,734,346 35,392,980 21,921,194 transfer (IDR) Minimum per capita revenue after 356,598 560,275 120,262 537,761 transfer (IDR) Max to min ratio of per capita 101.17 44.15 294.30 40.76 revenue after transfer Coef. of variation 1.346 1.091 1.889 0.966 Weighted coef. of variation 1.089 0.940 2.137 0.860 Relative mean deviation 0.370 0.324 0.491 0.303 Gini Index 0.500 0.440 0.639 0.412 Weighted Gini Index 0.383 0.337 0.538 0.329 No. of Observed D1 Group 100 100 100 100 No. of Receiver 100 100 31 100 Average DAU allocation (million) 437,288 386,566 84,570 392,470 Average per capita DAU allocation 1,172,551 806,197 1,110,100 714,346 Maximum per capita DAU 15,242,422 5,084,624 7,834,006 3,777,651 allocation (IDR) Minimum per capita DAU 171,044 292,594 14,443 333,629 allocation (IDR) Max to min ratio of per capita 89.11 17.38 542.41 11.32 DAU allocation Maximum per capita revenue after 22,708,516 12,550,717 15,300,098 10,197,022 transfer (IDR) Minimum per capita revenue after 356,598 560,275 120,262 537,761 transfer (IDR) Max to min ratio of per capita 63.68 22.40 127.22 18.96 revenue after transfer Coef. of variation 1.513 1.069 2.034 0.934 Weighted coef. of variation 0.686 0.495 1.289 0.432 Relative mean deviation 0.328 0.275 0.480 0.242 Gini Index 0.438 0.362 0.612 0.321 Weighted Gini Index 0.229 0.165 0.332 0.152 No. of Observed D2 Group 99 99 99 99 No. of Receiver 97 98 40 99 Average DAU allocation (million) 376,778 340,249 56,915 348,238 19 Average per capita DAU allocation 1,576,243 948,519 765,176 840,495 Maximum per capita DAU 21,675,276 3,966,180 5,555,236 3,184,663 allocation (IDR) Minimum per capita DAU 273,026 353,957 15,913 461,592 allocation (IDR) Max to min ratio of per capita 79.39 11.21 349.10 6.90 DAU allocation Maximum per capita revenue after 32,266,992 14,557,896 16,146,952 11,791,707 transfer (IDR) Minimum per capita revenue after 520,858 606,292 136,799 744,153 transfer (IDR) Max to min ratio of per capita 61.95 24.01 118.03 15.85 revenue after transfer Coef. of variation 1.477 0.962 1.695 0.848 Weighted coef. of variation 0.871 0.594 1.395 0.522 Relative mean deviation 0.315 0.254 0.433 0.228 Gini Index 0.445 0.355 0.574 0.322 Weighted Gini Index 0.312 0.218 0.418 0.192 No. of Observed D3 Group 99 99 99 99 No. of Receiver 99 99 31 99 Average DAU allocation (million) 331,227 292,833 55,243 306,006 Average per capita DAU allocation 2,051,840 1,185,130 1,205,038 1,061,774 Maximum per capita DAU 16,397,758 5,053,894 7,083,893 2,240,744 allocation (IDR) Minimum per capita DAU 340,037 49,799 22,213 440,941 allocation (IDR) Max to min ratio of per capita 48.22 101.49 318.91 5.08 DAU allocation Maximum per capita revenue after 36,078,212 24,734,346 26,764,344 21,921,194 transfer (IDR) Minimum per capita revenue after 547,832 797,903 169,158 648,736 transfer (IDR) Max to min ratio of per capita 65.86 31,00 158.22 33.79 revenue after transfer Coef. of variation 1.298 1.142 1.886 1.061 Weighted coef. of variation 1.013 0.686 1.666 0.608 Relative mean deviation 0.339 0.285 0.466 0.255 Gini Index 0.462 0.385 0.609 0.351 Weighted Gini Index 0.388 0.253 0.510 0.232 No. of Observed D4 Group 100 100 100 100 No. of Receiver 95 100 24 100 Average DAU allocation (million) 335,726 408,791 202,092 438,260 Average per capita DAU allocation 2,850,914 2,166,975 3,346,229 1,925,755 Maximum per capita DAU 19,862,532 21,037,218 34,552,332 5,095,879 allocation (IDR) Minimum per capita DAU 136,836 32,246 4,320 891,093 allocation (IDR) 20 Max to min ratio of per capita 145.15 652.39 7,998.02 5.72 DAU allocation Maximum per capita revenue after 27,022,350 21,877,864 35,392,980 12,559,045 transfer (IDR) Minimum per capita revenue after 551,144 1,244,735 194,263 1,124,271 transfer (IDR) Max to min ratio of per capita 49.03 17.58 182.19 11.17 revenue after transfer Coef. of variation 1.041 0.783 1.531 0.621 Weighted coef. of variation 0.963 0.725 1.732 0.606 Relative mean deviation 0.342 0.274 0.466 0.230 Gini Index 0.465 0.361 0.603 0.309 Weighted Gini Index 0.411 0.299 0.600 0.279 Notes: *) Excluding non-receivers, inequality measures calculated for after transfer per capita revenue 21 Figure 5. Lorenz Curves for Provinces Before and After Equalizations, Weighted by Population Figure 6. Lorenz Curves for C1 Regions Before and After Equalizations, Weighted by Population 22 Figure 7. Lorenz Curves for C2 Regions Before and After Equalizations, Weighted by Population Figure 8. Lorenz Curves for C3 Regions Before and After Equalizations, Weighted by Population 23 Figure 9. Lorenz Curves for C4 Regions Before and After Equalizations, Weighted by Population Figure 10. Lorenz Curves for D1 Regions Before and After Equalizations, Weighted by Population 24 Figure 11. Lorenz Curves for D2 Regions Before and After Equalizations, Weighted by Population Figure 12. Lorenz Curves for D3 Regions Before and After Equalizations, Weighted by Population 25 Figure 13. Lorenz Curves for D4 Regions Before and After Equalizations, Weighted by Population Alternative 2: Moving from a Gap Filling to a Comprehensive Fiscal Equalization Approach Calculation of fiscal capacity and expenditure needs would follow the same approach as under the gap filling approach described above. However surplus or deficiency of per capita fiscal capacity and per capita expenditure needs with reference to average or other explicit standard of equalization are calculated. The jurisdictions in net deficiency positions will receive equalization payments from the center equivalent to the net deficiencies calculated after taking into account net positions with respect to capacity and needs. The jurisdictions in net fiscal surplus position will not receive any equalization transfers. Since it is a central program, the surplus is not taxed or redistributed to poorer jurisdictions. This alternative has the clear advantage of having an explicit standard of equalization determine the total pool as well as distributions. We have also followed a simpler representative expenditure system to determine needs. The pursuit of greater rigor in calculating expenditure needs can result in a complex and controversial determination of expenditure need equalization, as has been the experience in Australia (see Shah, 2004). Simulations of results using this approach are presented in Tables 8-10 and the redistributive impacts are graphed in Figures 5-13. Using group average standard of equalization separately for the nine groups yields the overall pool of funds to be distributed as well as allocation among jurisdictions. Only 10 provinces, 59 cities 126 districts qualify to receive equalization payments. Of course such a program would serve as a residual program rather than a general revenue sharing mechanism. Since this would serve as a residual program, its overall impact on fairness has to be analyzed in conjunction with other program. In isolation as the taxing powers of local grants are quite limited, on the fiscal capacity equalization, the program will 26 redistribute small sums. While expenditure need equalization may redistribute more resources, as shown by calculations presented in Tables 9-11, overall redistributive impact of this program is relatively smaller than actual DAU and other alternatives presented here (see Figures 5-13) and as a result the program does not fare well on comparative indicators of fiscal equity. Alternative 3: Back to the Future - An Almost Ideal Approach: Fiscal Capacity Equalization Supplemented by Output-based Operating Grants for Merit Services Under this option, fiscal capacity equalization follows the same approach as under alternative #2. However expenditure need compensation is done through output-based operating transfers for merit public services only where allocation is based upon the share of service population and there is no conditionality on spending but instead customized conditions on service delivery performance in terms of service access and quality for individual jurisdictions and providers for continuation of the grant program. The design of such transfers are spelled out in Shah (2007, 2010, 2011). In addition to these operating transfers, there would also be a need to have a capital grant and capital market access program (the former for poorer and the latter for richer jurisdictions) for merit services based upon a planning view to deal with infrastructure deficiencies pertaining to a national minimum standard. Our simulation of this alternative combines a fiscal capacity equalization grant with output based grants for education, health, transportation and social welfare and protection only. We have not included capital grants to overcome infrastructure deficiencies as such grant program must be a separate grant program based upon a planning view and must not be formula based grant program available to all. This alternative further simplifies the determination of grant pool and allocations. But the most important advantage of this alternative is that it preserves local autonomy while enhancing accountability to local government residents. Tables 9-11 present simulations of allocations using this approach and Figures 5-13 depict redistributive impacts. The results combine fiscal capacity equalization using group average standard for each cluster/group and output based grants for education, health, infrastructure and social welfare and protection using relevant service populations. The results demonstrate clear superiority of this approach in establishing fair financing allocation to ensure fair and equitable access to merit services to all Indonesians regardless of their place of residence. In addition, this approach improves simplicity of allocation criteria, preserves local government autonomy against higher-level undue interference and enhances local government accountability to local residents. The approach is home grown as it embodies rich and successful past Indonesian experience with INPRES grants. 5. Does Indonesia Need a Grants Commission? For determining the system of grants, one finds four stylized types of models used in practice (see Shah, 2005, for a formal framework for evaluating such institutional arrangements). The first and the most commonly used practice is for the federal/central government alone to decide on it. This has the distinct disadvantage of biasing the system towards a centralized outcome whereas the grants are intended to facilitate decentralized decision-making. In India, the federal government is solely responsible for the Planning Commission transfers and the 27 centrally sponsored schemes. These transfers have strong input conditionality with potential to undermine state and local autonomy. The 1988 Brazilian constitution provided strong safeguard against federal intrusion by enshrining the transfers’ formulae factors in the constitution. These safeguards represent an extreme step as they undermine flexibility of fiscal arrangements to respond to changing economic circumstances. The second approach used in practice is to set up a quasi-independent body, such as a grants commission, whose purpose is to design and reform the system. These commissions can have a permanent presence as in South Africa and Australia or they can be brought into existence periodically to make recommendations for the next five years as done in India. These commissions have proven to be ineffective in some countries largely because many of the recommendations have been ignored by the government and not implemented as in South Africa. In other cases, while the government may have accepted and implemented all they recommend, they have been ineffective in reforming the system due to the constraints they have imposed on themselves as is considered to be the case in India. In some cases, these Commissions become too academic in their approaches and thereby contributing to the creation of an overly complex system of intergovernmental transfers as has been the case with the Commonwealth Grants Commission in Australia (Shah, 2004, 2007). The third approach found in practice is to use executive federalism or central-provincial-local committees or forums to negotiate the terms of the system. Such a system is used in Canada andGermany. In Germany, this system is enhanced by having state governments represented in Bundesrat—the upper house of the parliament. This system allows for explicit political input from the jurisdictions involved and attempts to develop a common consensus. The fourth approach is a variation on the third approach and uses an intergovernmental-cum- legislative-cum-civil society committee with equal representation from all constituent units but chaired by the federal/central government to negotiate changes in the existing arrangements. The so-called Finance Commission in Pakistan and the Indonesian DPOD represent this model. This approach has the advantage that all stakeholders—donor, recipients, civil society and experts are represented on the commission. Such an approach keeps the system simple and transparent. It is important that in such forums only the donors and recipients be the voting members (principals) with civil society members and experts to serve as observers (non-voting members) and provide feedback and technical assistance to the principals of these forums. An important disadvantage of this approach is that if unanimity rule is adopted, such bodies may be deadlocked forever as was witnessed in Pakistan in the 1990s. The Indonesian DPOD is an important forum for decisions on grant determination. The DPOD’ role in grant determination can be strengthened by requiring that only the cabinet ministers, governors and mayors can serve as members and vote on matters relating to fiscal transfers and all these decisions must require three-fourth majority of the DPOD. In conclusion, there appears to be no clear advantage in creating an independent grant commission in Indonesia. Grant determination role is better served by the intergovernmental forum such as the DPOD supported by a technical secretariat at the Ministry of Finance. 6. Concluding Remarks on Long-Term Reform Options for Central-Provincial-Local Fiscal Relations in Indonesia 28 During the past decade Indonesia has made a remarkable transformation from centralized rule to decentralized and democratic local governance. This transformation can be sustained if the intergovernmental finance create the right incentives and accountability regimes for responsive and accountable local governance. Prior to the 2000 reforms, Indonesia had an intergovernment finance system the so called INPRES (presidential instruction) grants system, that was simple, transparent and focused on results based accountability. This paper has called for Indonesia to return to its roots and implement reform options that represent a ―back to the future‖ approach—an approach that draws upon rich and successful Indonesian experiences that have often been cited in the public finance literature as examples of better practices in central transfers (see Shah, 1994, 1998 and Boadway aand Shah, 2009). To strengthen accuntable local governance, Indonesia needs to consider the following reform options:  Tax decentralization and tax base sharing. Tax base sharing is feasible for personal income taxes on residence principle. Tax decentralization may be feasible for royalties, fees, severance, production, output and property taxes, sin taxes (gambling, liquor and massage parlors) and local environmental taxes and charges.  Output based per capita operating (non-matching) grants for setting national minimum standards for merit services such as education, health and infrastructure. These grants should embody simple allocation criteria to local governments based on service population, e.g. school operating grant based upon school age population. Local governments would disburse these grants to all providers – government and non government as done in Canada, Brazil, Chile, Finland and Thailand. Continuity of finance can be assured by maintaining or improving upon existing standards of access and service quality. Such transfers will preserve local autonomy while enhancing simplicity, transparency and citizens’ based accountability for service delivery performance. Indonesia in the past had a measure of succces with a grant program (INPRES grants) that embodied at least some of these features.  Fiscal capacity equalization grants based upon national average standard for each cluster/group to enable all jurisdictions to provide reasonably comparable levels of public services at reasonably comparable levels of tax burdens. Adoption of fiscal capacity equalization grants would also simplify the use of indirect fiscal capacity equalization on specific grants.  Capital (with 10 to 90% matching) grants to fiscally disadvantaged jurisdictions to overcome infrastructure deficiencies in setting national minimum standards for merit services. These grants should be based upon a planning view of identified infrastructure deficiencies and should contain matching requirements that vary inversely with per capita fiscal capacity. These grants combined with ouput based operating grants will create a level playing field and enable poorer juridictions to integrate with the broader national economy and help reduce regional income and fiscal disparities.  Assistance for responsible capital market access to richer local jurisdictions. The above mentioned reforms will result in an intergovernmental finance system that is more transparent, objective, predictable and simpler with a sharper focus on objectives. These reforms may be considered as integral elements of any effort at fine tuning the existing fiscal system of multi-order governance in Indonesia. This paper has demonstrated the need for simpler and transparent allocation criteria for greater fairness and accountability. Complex systems compromise both fairness and accountability as has been happing under the current DAU system. 29 References Bishop, George and Anwar Shah, 2011. Sharing Petroleum Resources in Iraq: Obstacles or Foundations to Decentralization. In Jorge Martinez-Vazquez and Francois Vallancourt, eds. Decentralization in Developing Countries. Chapter 16: 549-594. Cheltenham, UK: Edward Elgar Press. Boadway, Robin and Anwar Shah, 2009. Fiscal Federalism: Principles and Practice of Multiorder Governance. New York and London: Cambridge University Press. Boadway, Robin and Anwar Shah, editors, 2007. Intergovernmental Fiscal. Transfers:Principles and Practice. Washington, DC: World Bank. Eckardt, Sebastian And Anwar Shah 2007. Local Government Organization and Finance: Indonesia. In Anwar Shah, ed. Local Governance in Developing Countries. Chapter 7:233-274. Washington, DC: World Bank. Fadliya and Ross H. McLeod, 2010. Fiscal Transfers to Regional Governments in Indonesia. Working paper no. 2010/14, ANU College of Asia and the Pacific. Harjowiryono, Marwanto, 2011. Lessons Learned From Indonesia’s Fiscal Decentralization. Paper presented at the International Conference on Fiscal Decentralization in Indonesia a Decade after the Big Bang. Ministry of Finance, Jakarta. Indonesia, Government of, Ministry of Finance, 2012. . Sistem Informasi Keuangan Daerah Lewis, Blane, 2002. Revenue Sharing and Grant Making in Indonesia. The First Two Years of FiscalDecentralization. In Intergovernmental Fiscal Transfers in Indonesia, ed. Paul Smoke. Lewis, Blane. 2001. The new Indonesian equalisation transfer. Bulletin of Indonesian Economic Studies, Vol. 37, No. 3, 2001: 325–43. Mochida, Nobuki 2008. Measuring Fiscal Needs: Japan’s Experience. In Kim, Junghun and Jorgen Lotz, eds. 2008. Measuring Local Government Expenditure Needs. The Copenhagen Workshop 2007. Chapter 7:166-187, Copenhagen, Denmark: The Danish Ministry of Social Welfare. Moisio, Antti, Heikki Loikkanen and Lasse Oulasvirta, 2010. Public Services at the Local Level – The Finnish Way. Government Institute for Economic Research, Helsinki, Finland. Qibthiyyah, Riatu 2011. Review of Incentives and Sanctions Linked Intergovernmental Transfers. Working paper ADB-INO.TA 7184-Local Government Finnace and Governance Reform, Jakarta, Indonesia. Shah, Anwar. 1983. The New Fiscal Federalism in Australia. , Expenditure Needs and Fiscal Equalization Grant Series Working Paper #1,, Ministry of Finance, Government of Canada, Ottawa. ---------- and Zia Qureshi et al. 1994a. Intergovernmental Fiscal Relations in Indonesia. Issues and Reform Options. World Bank Discussion Paper Series. No. 239. Washington, DC: World Bank. ----------. 1996. ―A Fiscal Need Approach to Equalization.‖ Canadian Public Policy 22 (2): 99–115. ----------. 1998. ―Indonesia and Pakistan: Fiscal Decentralization—An Elusive Goal?‖ In Fiscal Decentralization in Developing Countries, ed. Richard Bird and François Vaillancourt, 115– 51. Cambridge: Cambridge University Press. ----------. 2004. ―The Australian Horizontal Fiscal Equalization Program in the International Context.‖ Presentation at the Heads of the Australian Treasuries (HOTS) Forum, Canberra, September 22, and the Commonwealth Grants Commission, Canberra, September 23. 30 ----------. 2005. ―A Framework for Evaluating Alternate Institutional Arrangements for Fiscal Equalization Transfers.‖ World Bank Policy Research Working Paper 3785, Washington, DC. ----------and Chunli Shen. 2007. ―Fine Tuning the Intergovernmental Transfer System to Achieve A Harmonious Society and A Level Playing Field for Regional Development in China In Lou Ji-wei and Shuilin Wang eds. Public Finance in China. Washington, DC: World Bank. ---------- 2007. A Practitioner’s Guide to Intergovernmental Fiscal Transfers. Chapter 1, pp. 1-54, in .Intergovernmental Fiscal Transfers: Principles and Practice, edited by Robin Boadway and Anwar Shah, chapter 8, pp. 225-258. Washington, DC: World Bank. ----------, ed. 2007. The Practice of Fiscal Federalism: Comparative Perspectives. Montreal and Kingston: McGill-Queen’s University Press. ----------, 2008. Fiscal Need Equalization: Is it Worth Doing? Lessons from international practices. In Kim, Junghun and Jorgen Lotz, eds. 2008. Measuring Local Government Expenditure Needs. The Copenhagen Workshop 2007, chapter 1:35-60. Copenhagen, Denmark: The Danish Ministry of Social Welfare. ----------, 2010. Autonomy with Accountability: The Case for Performance Oriented Grants. In Kim, Junghun, Jorgen Lotz and Jorgen Mau, eds. 2010. General Grants Versus Earmarked Grants: Theory and Practice. The Copenhagen Workshop 2009, chapter 2:74-106. Copenhagen, Denmark: The Danish Ministry of Interior and Health. ----------, 2011. Autonomy with Equity and Accountability: Toward a More Transparent, Objective, Predictable and Simpler (TOPS) System of Central Financing of Provincial-Local Expenditures in Indonesia. Presented at the International Conference on ‘Decentralization in Indonesia: A Decade After the Big Bang’ held at Jakarta from September 11-12, 2011. Forthcoming in the Asian Development Bank book, Decentralization In Indonesia: A Decade After the Big Bang, edited by James Lamont. Shankar, Raja and Anwar Shah. 2003. ―Bridging the Economic Divide Within Countries: A Scorecard on the Performance of Regional Policies in Reducing Regional Income Disparities.‖ World Development 31(8):1421-1441. Sidik, Machfud. A New Perspective on Intergovernmental Fiscal Relations. Lessons from Indonesia’s Experience. Jakarta:Ripelge. 31 Annex Table A1. Result Summary of Alternative 1 versus Current Formula* Current Formula Alternative 1 Per Average Average Per Capita DAU Per Per Average DAU Average Per Capita DAU Capita DAU Per Allocation Capita Capita Allocation Per Allocation Difference Allocation Capita Fiscal Fiscal (million) Capita (Current – (million) DAU Need Capacity DAU Alter. 1) Allocation Max Min Allocation Max Min Province 622,547 209,547 796,794 10,629 576,351 171,593 979,405 236,474 956,711 63,671 (26,928) C1 599,780 352,450 506,550 153,056 975,775 151,810 772,744 449,996 543,150 314,452 (97,546) C2 367,534 544,808 810,767 147,101 1,520,083 218,612 484,375 710,779 887,157 330,507 (165,971) C3 272,053 1,387,206 2,308,378 361,909 2,539,058 264,076 271,306 1,285,945 1,526,098 232,243 101,261 C4 214,035 3,557,909 7,301,673 2,291,657 4,445,029 225,142 149,815 2,304,629 2,594,625 1,991,547 1,253,280 D1 437,288 1,172,551 15,242,422 171,044 1,563,318 87,132 386,566 806,197 5,084,624 292,594 366,354 D2 376,778 1,576,243 21,675,276 273,026 1,817,404 110,371 340,249 948,519 3,966,180 353,957 627,724 D3 331,227 2,051,840 16,397,758 340,037 2,317,179 147,148 292,833 1,185,130 5,053,894 49,799 866,710 D4 335,726 2,850,914 19,862,532 136,836 4,214,821 246,985 408,791 2,166,975 21,037,218 32,246 683,939 Indonesia 376,658 1,712,272 21,675,276 10,629 2,356,092 165,778 391,449 1,209,122 21,037,218 32,246 503,151 Notes: *) excluding non-receivers 32 Table A2. Result Summary of Alternative 2 versus Current Formula* Current Formula Alternative 2 Per Capita Average Average Per Capita DAU Per Per Average DAU Average Per Capita DAU Difference DAU Per Allocation Capita Capita Allocation Per Allocation (Current – Allocation Capita Fiscal Fiscal (million) Capita Alter. 2) (million) DAU Max Min Need Capacity DAU Max Min Allocation Allocation Province 622,547 209,547 796,794 10,629 576,351 171,593 693,084 379,869 1,347,027 22,309 (170,323) C1 599,780 352,450 506,550 153,056 975,775 151,810 171,417 114,966 170,570 60,271 237,484 C2 367,534 544,808 810,767 147,101 1,520,083 218,612 86,086 128,791 322,956 49,547 416,017 C3 272,053 1,387,206 2,308,378 361,909 2,539,058 264,076 50,566 269,172 519,377 58,569 1,118,034 C4 214,035 3,557,909 7,301,673 2,291,657 4,445,029 225,142 15,376 271,529 530,996 18,214 3,286,380 D1 437,288 1,172,551 15,242,422 171,044 1,563,318 87,132 84,570 1,110,100 7,834,006 14,443 62,451 D2 376,778 1,576,243 21,675,276 273,026 1,817,404 110,371 56,915 765,176 5,555,236 15,913 811,068 D3 331,227 2,051,840 16,397,758 340,037 2,317,179 147,148 55,243 1,205,038 7,083,893 22,213 846,802 D4 335,726 2,850,914 19,862,532 136,836 4,214,821 246,985 202,092 3,346,229 34,552,332 4,320 (495,315) Indonesia 376,658 1,712,272 21,675,276 10,629 2,356,092 165,778 113,272 1,027,412 34,552,332 4,320 684,860 Notes: *) excluding non-receivers 33 Table A3--. Result Summary of Alternative 3* Alternative 3 Per Capita Average Average Average Per Capita DAU Difference Per Capita Per Capita Allocation Based Allocation Based DAU Per Capita Allocation (Current Equalization Output- on Equalization on Output Allocation DAU Formula– Grant Based Grant (million) (million) (million) Allocation Max Min Alter. 3) Province 57,136 129,941 287,054 699,018 905,339 171,008 306,176 75,320 38,539 C1 62,463 508,810 98,728 881,143 926,020 537,202 634,337 379,163 (184,752) C2 89,302 782,393 57,373 533,931 578,064 851,087 1,053,690 642,642 (306,279) C3 139,798 1,332,136 28,595 278,406 303,549 1,455,062 1,973,908 914,632 (67,856) C4 98,465 2,341,063 6,531 150,328 153,297 2,385,820 3,242,229 1,743,818 1,172,089 D1 26,244 694,926 18,083 379,088 392,470 714,346 3,777,651 333,629 458,204 D2 55,546 792,804 25,294 326,522 348,238 840,495 3,184,663 461,592 735,748 D3 80,115 995,417 30,363 280,857 306,006 1,061,774 2,240,744 440,941 990,066 D4 151,540 1,799,976 36,388 408,058 438,260 1,925,755 5,095,879 891,093 925,159 Indonesia 86,010 1,050,604 43,442 374,089 408,809 1,119,346 5,095,879 75,320 592,926 Notes: *) excluding non-receivers 34