Lombok Timur Public Expenditure Review

8 downloads 142 Views 306KB Size Report
The focus of this report is Kabupaten Lombok Timor (hereafter Lotim) and is in large ..... 7 See PAD workshop Sumba Timur PRODA-NTB/NTT, May 2001 ...
Lombok Timur Public Expenditure Review

25 January, 2002

Drafting Team: Basilius Bengoteku Bert Hofman Kai Kaiser Paul McCarthy Stephen Mink (Team Leader) George Soraya Unggul Supriayitno Farida Zaituni

2

TABLE OF CONTENTS

Introduction

3

Lotim’s Budget, and Budget and Planning Process

5

Sectoral Budget Management

12

Education

33

Public Works : Roads

40

Public Works : Solid Waste Management

43

Agriculture

46

The Governance Agenda

54

3

LOMBOK TIMUR PUBLIC EXPENDITURE REVIEW INTRODUCTION

1. This report is one of three regional reports – the two others being West Java and North Sumatera - that examine the management of public budget resources in the context of decentralization. These reports will serve as inputs to an overall report, all to address three broad objectives: • To support regional governments in managing decentralization and improving their planning and budgeting processes; • To build a relationship with local stakeholders, and consults with them on ways to address key development challenges in the regions; and • To build an analytical foundation for the selection of regions in which the Bank could operate, and of the Bank’s activities in the regions. 2. The focus of this report is Kabupaten Lombok Timor (hereafter Lotim) and is in large measure based on wide ranging interviews in Lotim and at the NT Provincial level with civil servants, parliamentarians, NGO representatives and citizens (see Annex). Initial and tentative observations from this field work was presented (in Annex) and discussed in a workshop in Selong September 25, 2001. 3. Lotim faces enormous challenges to gain control over its public resources and their effective use to provide the building blocks for economic growth and improved social welfare of its citizens. It is a largely agricultural economy, but confronted by a relatively harsh climate and difficult natural resources. Concerted efforts to invest in irrigation over the past two decades has significantly reduced food insecurity, but above-average public expenditure on health and education, on a per capita basis, has yet to move Lotim’s indicators from well below national averages (Table 1). Over the 1990-99 period, Lotim was amongst the districts starting from low indicators that made the slowest progress on human development.

Table 1: Lotim Human Development Indicators In National Perspective Life Expectancy (years) 1996 1999 53.2 56.0

Adult Literacy Rate (%)

Mean Years of Schooling (years) 1996 1999 4.3 4.8

HDI

HDI Rank

1996 1999 1996 1999 1996 1999 Lombok 66.5 68.6 54.4 52.1 286 of 284 of Timor 291 294 NTB 54.9 57.8 68.0 72.8 4.6 5.2 56.7 54.2 26 of 26 26 of 26 Indonesia 64.4 66.2 85.5 88.4 6.3 6.7 67.7 64.3 Source: BPS, BAPPENAS, UNDP, “Indonesia: The National Human Development Report,” 2001.

4

4. Poverty is high in Lotim, particularly in the more isolated northern mountainous areas around Mt. Rinjani and the dry south, while the middle benefits from richer and better irrigated agricultural land. Income is more evenly distributed (low Gini ratio) in the Indonesia context and given the high poverty rates, this indicates that a high proportion of the population is close to the poverty line. Economic activity tends to be small-scale in agriculture, commerce and manufacturing and employment outside of the agriculture tends to be limited to the non-formal sector. Population densities that rival Java is associated with farmers having modest land assets and significant numbers of landless rural workers. 5. Human resources indicators reflect these economic circumstances. Despite comparatively good provision of and access to schools, Lotim has low adult educational attainment and rapid falloff of attendance at post primary school levels by children. A large proportion of adults over the age of 30-35 have been virtually by-passed by the educational system, with high illiteracy, limited capacity to function in Indonesia’s national language, and significantly worse outcomes for women than for men. Health outcomes have proven stubbornly poor despite good access to health services and facilities. Maternal, infant and child mortality remain high and life expectance low compared to elsewhere in Indonesia. Poor nutrition, housing conditions and health practices and norms at the household and community are a contributing factor to these outcomes. 6. These constraints of economic opportunity and social welfare have prompted many in Lotim to seek their futures elsewhere, and the region has experienced out-migration for many years, to destinations both in Indonesia and abroad. Young women are increasingly participating in this through employment as household workers in the Middle East, but migration is still predominated by men, resulting in a high level of female-headed households. 7. Facing these critical socio-economic problems, Lotim embarks on fiscal decentralization with tight resource constraints. Budget resource limitations are compounded by the additional civil servants who have also been decentralized. Significant increases in tax revenues are unlikely, nor desirable, at least in the short run because of the low tax base and likely incidence on the poor, and there is little capacity to borrow. To address its problems, Lotim will need to be ambitious in improving its service delivery, reduce leakage in public expenditure and mobilize more revenue charges. Full use needs to be taken of the opportunities that exist to: (i) improve planning, budgeting and financial management processes; (ii) improve service delivery by Dinas through greater involvement of user groups in expenditure planning and delegation to them of program implementation; and (iii) improved governance by engaging civil society in government and the DPRD. 8. From an electoral perspective, the willingness to carry out governance reform is affected by factors related to self-interest: • It will improve prospects of re-election (for DPRD members); • It may help prevent the potential rollback of decentralization legislation which is being called into question by senior members of the central government; • It can help improve the public image of the local government, especially if the focus is placed on running a “clean” government; • Reform-minded governments tend to attract a larger share of government and external donor resources; and,

5



Running a clean government reduces the likelihood of legal and social sanctions in the event that real judicial reform takes place at a national level.

9. The depth and breadth of the governance reforms that will actually take place remain largely in the hands of local government officials and politicians as well as with civic leaders who have the capability to bring public pressure to bear upon the government. 10. The report that follows is organized in three sections. The first presents the basic structure of Lotim’s budget – both revenues and expenditures – as well as the planning process which produces this and the financial management practices which underpin its implementation. In this first year of Kabupatens’ budget management under decentralization, shifts in budget priorities are already evident and these are identified. The second section examines in greater detail the potential for greater budget efficiency and effectiveness in a sample of Lotim’s sectoral service provision. Three sectors are reviewed, education, public works (roads and solid waste management) and agriculture. The third section analyzes how governance dynamics are evolving and may yet need to evolve to achieve a more balanced set of checks and balances between the citizenry, parliament and executive in the use of public budget resources towards the achievement of development priorities. Lotim’s Budget, and Planning and Budget Processes 11. A key objective of decentralization is to get better services at lower cost for the regions by moving decisions on service delivery closer to the people it affects. The budget is central to this objective: the planning and budget process allocates scarce budgetary funds over competing uses. For Lotim, budgetary funds are scarce indeed: total revenues in FY2001 were budgeted to be less than half the average for all local governments in Indonesia,1 and the lowest in the province of NTB. Options for raising revenues or issuing debt remain limited in the short run. Therefore, making the best possible use of existing funds through careful planning and budgeting to get the most out of available resources is therefore a must for Lotim.

Kabupatens (regencies) and Kota’s (cities) are referred to together as “local governments.” “Regions” is used for either Kabupatens and Kota’s or provinces, or both. Local governments and provinces together are referred to as “regions.” Note that for the fiscal data on a per capita basis referred to in the text and tables two kinds of averages can be calculated: (i) the average for the per capita values over all local governments; and (ii) the national average per capita for Indonesia’s local government level. Unless otherwise specified, the text refers to method (i), in order to compare Lotim with the “Typical” Local government in Indonesia. 1

6

The Budget for FY2001 12. Overview. Lotim’s budget Table 2: Lotim, Summary Budget structure is fairly representative for an 1998/99-2001 (Rp. Million). average kabupaten in Indonesia. A large chunk of its revenues is derived from the 1998/99 1999/2000 2000 2001 general grant or DAU, and the wage bill takes the largest cut form spending. Total 68,394 102,839 269,983 63,363 Expenditures Decentralization had a dramatic impact on Lotim’s budget. Total expenditures Wages 40,800 49,231 65,667 160,096 more than doubled compared to the (9 months) FY2000, and almost doubled on Development 17,468 13,125 29,349 93,678 Expenditures an annual basis (Table 2: Lotim Summary Budget). The wage bill rose Total Revenues 64,365 70,210 105,735 269,983 with similar magnitudes, and the revenue structure changed completely (Annex Central 57,192 62,591 89,658 236,699 Transfers Table 3: Lotim income structure). Lotim planned to spend nearly Rp. 280, 000 per SDO 41,558 50,375 66,031 person in FY2001. Compared to the average of Rp. 640,000 for Indonesia’s Shared 2,295 4,302 8,720 16,172 Revenues local governments this is low (Annex Table 2: Summary Indicators), but Own Revenues 4,041 2,586 4,415 13,145 compared to the past, this is a substantial (PAD) increase—even if the additional spending obligations on civil service and projects inherited from the center are Retributions 3,445 1,775 2,667 5,777 taken into account. The government’s own impression is that decentralization Source: ABPD Revisi. The numbers for 2001 include revisions made after disbursement of the contingency fund. has brought about no undue budgetary pressures for Lotim. And indeed, Lotim was able to allocate more than one third of its budget for development purposes, an increase from the past. At the same time, surviving the budgetary pressures of decentralization is hardly enough for a region that faces challenges ranging from insufficient infrastructure, low social indicators, and harsh climatic conditions. Addressing those challenges with the limited resources Lotim has will be a major task for the regional government in years to come.

Revenues 13. Like most of Indonesia’s local governments, Lotim relies heavily on transfers from the central government. The most important one is the general grant (DAU) followed by shared revenues from land and real estate taxes (PBB) and natural resource revenues (SDA). Own revenues are as of yet modest, even after Law 34/2000 opened up increased possibilities for raising taxes locally. Over time, this fiscal dependence on the center is far from healthy, as there are few things that make a local government more accountable to its electorate than a significant local revenue base. Even so, in the short run, this situation is likely to persist. 7

14. General Grant (DAU). The DAU is by far the most important source of revenues for Lotim. In the budget for FY2001 the Rp. 226 bn. amounted to 91 percent of revenues. The DAU is a general grant from central government which can be allocated as the region sees fit—as long as it Box 1 : Ongkos Administrasi performs the tasks and functions of The DAU is a general grant which is to be allocated through the Law 22. Nevertheless, some central regional budget. Central government has in principle no say in the “oversight” seems to be still at work use of the funds. But many regional treasuries think otherwise. Lotim’s finance bureau, for instance must each month go to the (Box 1: Ongkos Administrasi).2 regional treasury (KPKN) to fill out a request for its share of the DAU. Moreover, the KPKN requires proof of spending of last month’s allocation. And, of course, a little administrative cost is involved. The finance bureau is in a bit of a bind here, because it has no alternative but to “recover” the costs of this fee on their own disbursements. To eliminate this “onkos administrasi” the central government could consider direct disbursement of the DAU every month from the BUN directly into the regional governments accounts in the regional development banks, and without administrative costs.

15. The DAU for 2001 was allocated largely on the basis of past spending patterns, and only 20 percent on the basis of an equalizing formula. 3 The reason for this was that an allocation by formula alone would have risked a significant mismatch between spending obligations (on civil service and ongoing projects of the devolved bureaus) and revenues. Therefore, the allocation was largely determined by the “base amount” which consisted of 130 percent of past transfers for personnel (SDO) plus 110 percent of past development grants (INPRES). Yet, some mismatches still remained, and in mid-2001, Lotim received on another Rp. 9 billion from the central government contingency fund. 16. In the draft revised budget, the DAU including contingency fund constitutes 87 percent of revenues for Lotim. On a per capita basis, this puts Lotim with Rp.243,000 at less than half the average DAU for local governments in Indonesia. On the other hand, this is still almost double the amount of the region at the bottom end, which received Rp. 133,000 per capita. Own revenues (PAD) 17. East Timor’s own revenues accounted for less than five percent of total revenues (or just over 13 billion Rp.). On a per capita basis this is less than half the national average for local governments. The bulk of the own revenues is derived from retributions (levies), with charges for health services adding up to 1.8 billion, or almost 15 percent of total own revenues. The region received some 2 billion in revenues from government enterprises, and a further “other” revenues of Rp.5 billion. 4 The top item in revenues from government enterprises came from cattle fattening, an activity one would not normally consider as one in need of government involvement. 18. Local taxes. Lotim’s own taxes total less than ten percent of own revenues. The most important own tax was the street lighting tax which amounted to nearly 75 percent of total taxes. Ministerial decree 556/KMK.03/2000 dd. December 26 2001 describes the disbursement procedures for the DAU. These include monthly reporting on spending. 3 There was a third part to the allocation, which according to some regions visited, was strongly correlated with timely visits of regional governments to the Ministry of Finance. 4 The visiting team did not further investigate this category. 2

8

Second was the tax on grade C minerals (e.g., bricks). Law 34/2000 gives a region the right to issue regional taxes by means of regional regulations approved by the local parliament. Many regions have taken this opportunity and issued new taxes—and a significant part of those has been found to be conflicting the principles stated in Law 34/2000. 19. Revenue Effort. Overall, Lotim’s own revenues fall short of what could be expected on the basis of its regional GDP.5 If the region would achieve a nationwide-average revenue effort, the revenues for 2001 would have been Box 2: Cancelled? Rp.17.63 billion instead of the budgeted Rp.11.96 billion. But on a total budget of Regional regulations on taxes and levies need approval by the Rp.248.50 bn. this would hardly have regional parliament. The central government has thirty days from made a difference: the share of PAD the receipt of a regional regulation to issue a cancellation would increase from 5 percent of budget (dibatalkan). As of September 2001, and inter-agency working group out of the Ministry of Home affairs had reviewed and to 7 percent of the budget. Within PAD, classified a total of 1,041 regulations on tax (116) and levies regional taxes in Lotim are very low (933).. Of these, 103 (94 levy, 9 tax) were submitted for compared to the national average: only 10 cancellation—none of them from Lotim, but neighboring BIMA percent of the 11.5 billion budgeted PAD was less lucky with its tax on exports to other regions of consists of regional taxes. For other local industrial and agricultural products. The draft cancellations were forwarded to the Minister of Home Affairs for signature, but at governments this is on average about half. the end of 2001 he had still not signed. So now the question is Again, even if Lotim would bring its tax whether the regions can keep their illegal taxes… effort up to average, this would only add about 2 percent to overall revenues. Thus, expectations on PAD are best kept limited. For FY2001, Lotim’s government had budgeted for a study on expanding PAD, and the results of these study could inform policy makers what to expect in terms of increased revenues from this source. Box 3: Criteria for Regional Taxes 20. PAD Regulations. Lotim has issued some 13 regional regulations on local taxes and retributions since the start of FY2000, 7 of them in FY2001 up to August.[Box 2: Regional Regulations.] The authorities do not consider these regulations as “new” because these were already in place before they were banned by Law 18/1997.6 The authorities had a further 9 regulations on taxes and retributions in preparation by the time of the Bank team’s visit in end-August 2001. In the discussion on new taxes, the authorities indicated that they would consider taxes on traded goods such as

According to Law 34/2000, regional taxes should: o Have the character of a tax, not a levy o Have tax objects to be found in the region, with little mobility outside the region o Not conflict with the public interest o Not tax the same item as national or provincial taxes o Have adequate revenue potential Additional criteria that a region could consider are that a regional tax should: o Tax those that benefit most from government services financed by the tax (benefit principle) o Not tax the poorest sections of the population o Not overburden sectors the government sees as promising (e.g. agriculture) o Promote efficient use of resources (e.g. an idle land tax)

These estimates were used in the calculation of the DAU 2002. Based on annualized FY2000 outcomes, the average tax effort in Indonesia can be explained by the equation: PAD = 2.357 + 0.000957*(Non-Mining GDP). For 2000, one would expect Lotim to raise Rp. 8.67 billion. Actual receipts were Rp. 5.88 billion on an annual basis. 6 This phenomenon of “recycling” regulations was also observed by SMERU in a number of other regions, See Ilyas Saad, Indonesia’s decentralization Policy: The Budget Allocation and its Implications for the Business Environment, SMERU working paper [not numbered] September 2001. 5

9

agricultural products that are transported outside the region. In fact, PERDA 16/2001 imposes a tax on inter-island transport of goods. In the authorities’ view, such taxes would not burden to poor. They are probably wrong: besides that taxes on internal trade contravene Law 34/2000 (Box 3: Criteria for Regional Taxes), such taxes would hurt those that are producing the taxed goods (usually small farmers) because the price they receive net of tax is likely to be lowered by the tax. Other taxes, such as the levy on motor vehicles (PERDA 9/2001) is likely to conflict with the rule that regions cannot tax the same tax base that higher levels of government already tax. 21. NTB’s provincial government has issued several new taxes as well, including higher taxes on motor vehicles and fuel. The provincial government has realized the potential damaging effects of local taxes on the economy, and aims to establish a common approach on new taxes across the province’s districts. Such an initiative could build on the efforts of East Sumba. In that regency, a process to review proposed taxes has been put in place. 7 Beyond reviewing the proposed taxes against the criteria in Law 34 (see Box 3) Lotim could add criteria on its own—such as impact on specific sectors or impact on the poor. A review panel could then be organized with representatives from various stakeholders, which could make recommendations to regional government and the regional parliament. 22. Tax Administration. Before introducing new taxes, the government may want to review the performance of existing taxes. Until now, little information on compliance and tax yields is available. The collection of taxes is fairly informal, and is largely done by visiting potential tax payers such as restaurants and shops. Although several checks on the revenue receipts are done once they enter the Kas Daerah, little is known what should actually be collected—and which tax collector does a good job, and which one does not. Intensification of retributions collection could add to revenues as well. The amounts stated in the budget and government accounts are unlikely to be the full picture: Many service delivery units charge for their services in addition to the official charges. But these revenues never reach the region’s Kas Daerah.

7

See PAD workshop Sumba Timur PRODA-NTB/NTT, May 2001

10

23. Shared Revenues. Shared revenues for budget FY2001 were slightly less than Lotim’s own revenues. The bulk of shared (national) taxes for Lotim were composed of the land and building tax (5.3 billion or 73 percent) and the income tax (1.2 billion or 16 percent). The land transfer tax raises only minimal revenues in Lotim—less than 3 percent of the average of other Kabupatens. These taxes are collected by the central tax administration, but regional authorities have some influence on their collection. First, they have some flexibility in setting the exemption level for the real estate tax. Second, according to Law 22/99, the region will be in charge of land administration,8 and could potentially improve the information on the tax base for the land and real estate tax and the land transfer tax. And third, in some regions, the regional government has actively cooperated with the central tax administration to identify high income individuals. Lotim could perhaps as well pursue such program with the tax administration to enhance the performance of these shared taxes.

Table 3 : Revenue Shares according to Law 25/99, percent Central Government

Item

Provincial Government

Originating Local Government

Other Local Governments in the same province

All Local Governments in Indonesia (Equal Share)

Oil (non-tax, onshore) 85 3 6 6 -LNG (non-tax, onshore) 70 6 12 6 -Mining: Land-rent 20 16 64 --Mining: Royalty 20 16 32 32 -Forestry: Land-rent 20 16 64 --Forestry: Resource rent 20 16 32 32 -Fishery 20 -80 Property Tax 9a 16.2 64.8 -10 Land Transfer Fee 16 64 -20 Personal Income Tax 80 8 12 --a The Government’s share in the Property Tax is supposed to cover administrative costs. Source: Law 25/199, Government Regulations 104/2000 and Bambang Brojonegoro (2001) Fiscal Decentralization in Indonesia, Background paper to a ASEM-WBI Conference on Decentralization in East Asia, Bali, January 10-11 2002.

24. The share of the provincial motor vehicles tax accounted for less than ten percent of shared revenues (630 million or 9 percent). Additional transfers from the province included. The royalties category (which in addition to 32 percent of the actual district of activity is also distributed 32 across all kabupaten in a province) exceeds projected own revenues at 8 billion. This allocation was only determined at the end of March 2001. 25. DAK. Finally, the DAK this year (and anticipated next year) was an estimated Rp. 1 billion (less than 0.5 percent of the total budget) and exclusively directed at reforestation. In the future, special grants could become a revenue source for regions such as Lotim, but central regulations are still pending. Potential areas for a DAK that are currently discussed at the center include health and education, and given the region’s low income, and poor social indicators, it could be one of the first regions to benefit from special grants.

8

The decentralization of the land agency has been postponed by presidential decree.

11

Expenditures 26. Lotim’s first budget after decentralization is dominated by the wage bill. Routine spending accounts for some sixty percent of total, and over eighty percent of that is devoted to wages. Development spending accounts for about 38 percent of total. This distribution of the budget is much the same as it was before decentralization. Especially after 1997/98, when the wage bill for primary teachers was shifted from the province to the district level, did the wage bill dominate Lotim’s spending. New is that wages are no longer financed from the SDO grant. And the shift in additional civil service to the district, combined with expected and unexpected wage increases did cause significant budget problems for Lotim during the budget year. 27. Decentralizing the civil service and the wage bill. Lotim’s civil service now consists of some 10,500 civil servants, or some 1 percent of the population. This number is net of the contract workers and “honorar” personnel financed from the development budget. All civil servants working for the regional government are now paid from the district budget.9 Their pay takes up 54 percent of total spending, and almost 90 percent of recurrent spending. 28. Before decentralization, three types of civil servants existed in Lotim: regional civil servants, central civil servants seconded to the Lotim regional administration,10 and central civil servants working in the deconcentrated central agencies. At the end of 2000 the first two categories added up to some 8,200 people. 11 This number already included the personnel which was decentralized during FY2000: in that year, the Gus Dur government decided to abolish 6 departments altogether, or turn them into a state ministry without regional apparatus.12 Nationwide, some 130,000 civil servants working in the regional offices of these ministries (kanwils and kandeps) had to be placed in the regional governments. How many civil servants were decentralized to Lotim in this round is unclear, but they are included in the 8,200 number quoted earlier. 29. With decentralization, Lotim received an additional 2400 civil servants from various regional offices of departments on its payroll. 13 These departments through Law 22/99 have lost their regional deconcentrated apparatus, and had to reassign their civil servants.14 Some 2.1 million civil servants were re-assigned nationwide. The 28 percent increase in the number of civil As of 2001, there seem to be no more seconded central government staff working in the regions. The secondment was administratively managed through the Ministry of Home Affairs. The bulk of the seconded personnel was primary school teachers—nationwide some 1.1 million. 11 Data from the central civil service wage bill. These differ somewhat from the numbers underlying Lotim’s wage bill on the budget. From central data sources, the estimated wage bill would have been some Rp.125 billion—including an estimated 7 percent wage drift. Lotim’s first budget for 2001 estimated a wage bill of Rp.135 Tr. It is unclear what explains the difference, This could be a different assumption on wage increases underlying Lotim’s wage bill, a different number of civil servants, and/or inclusion of contract workers in Lotim’s budget amount. 12 The departments are: Department of Information, Department of Tourism, Social Welfare, Transmigration, Manpower, Cooperatives. 13 These numbers are taken from Central Government payroll data. Regional data as quoted in the joint provincial government/BPS publication “Nusa Tengara Barat 1999 are hard to reconcile with these numbers. For Lotim, the numbers quoted on pp.60-61 show 1188 regional civil servants and 1132 central civil servants seconded to Lotim. However, this excludes the teachers—5268 primary school teachers and 1337 secondary school teachers, and 646 high school teachers. But not all teachers are civil servants, and not all teachers are on the government’s payroll. The number of teachers quoted in central government sources is 7767 for 2000/2001 (http://www.pdk.go.id/, table 23.2) for all teachers (private and public) of which 6904 public. 14 Home Affairs, Health, Agriculture, Education, Forestry, Industry, Communications, Mines and Energy. 9

10

12

servants paid from the local budget was well below that of the national average increase of 54 percent.15 Lotim’s wage bill before the 2001 wage increases rose by some 27 percent compared to FY2000. 30. Therefore, the “hold harmless” clause in the DAU, which guaranteed a minimum of 130 percent of FY2000 SDO that financed the wage bill, was sufficient to pay civil servant wages before the pay increases of July 2001. According to the revised Lotim budget, those pay increases—retroactive for January 1 2001—increased the wage bill by a further 18 percent, or Rp. 25 bn. Only Rp. 9 bn. of that was compensated from the contingency fund, a Rp. 6.5 Trillion central government fund established to provide additional funds for regions that ran into trouble because of decentralization. 31. The increase in the number of regional civil servants (Pegawai Negri Sipil Daerah) was markedly larger than the increase in the number of civil servants paid from the regional budget. The reason is that many of the civil servants already on the region’s payroll were still registered as national civil servants before 2001. With the Big Bang decentralization this has now changed: with ceremonies throughout the regions in April and May 2001 were all national civil servants working for regional governments transferred, and “accepted” by the regions by means of a document signed by the bupati or walikota. 32. Total public sector pay and the number of government workers are larger than the wage numbers in the budget suggest. This wage bill misses out on two items: the non-civil servant workers, and the civil servant compensation paid from the development budget. Although numbers for Lotim are not available, the share of non-civil service workers adds 10 to 25 percent to the total number of workers in other regions. These contract workers range from contract teachers to daily workers in irrigation paid from fees, and they are financed from both the routine and the development budget, and sometimes from off-budget sources. In addition, civil servants receive project-related pay and fees ranging from management rewards to attendance fees for meetings. For a typical project in Lotim, the management overhead for a project is some [10] percent of the project costs.16 Thus, in reality, the wage bill is at least some 3 percentage point of total spending higher than the official numbers show. 33. Lotim seems to have little scope for reducing the wage bill in the short run. For one, the region’s wage bill is hardly excessive: with Rp.170,000 per capita the region’s wage bill is well below the nationwide local government average of Rp.283,000. Moreover, over two thirds of the wage bill goes to teachers (65 percent to primary teachers alone)—and solving the shortage of teachers (see section on education) would increase rather than decrease the wage bill. Finally, even if a reduction in the civil service ranks would be desirable, this can only be effected over the long run. Civil service rules imply that redundancy only saves part of the wage bill, as base wages still need to be paid to the retrenched civil servant. In addition, there is an implicit agreement between central and regional governments that none of the former central civil servants transferred to the regions can be made redundant in the first three years after decentralization. So a large part of the

Based on MOF payroll data. The 54 percent is for province and local government together, as no separate data for these levels are available. 16 See ABPD. 15

13

wage bill is to there to stay. Over time, the wage bill is even likely to rise, even without wage increases, because starting 2001 the region must take care of the pensions of new civil servants.17 34. Yet, because the wage bill is such a large part of spending it is worth extra scrutiny. The government could consider starting with an inventory of all civil servants in the regions—a “civil service census” to have a solid baseline of the number of civil servant, and their characteristics. Such a census could also detect possible ghost workers, and signal possible excessive absentia of some civil servants. The census would need to be followed up with an analysis of payroll management to ensure that the payroll stays “clean.” As part of the census, the qualifications and characteristics of the civil service could be registered as well, including background, training, and experience. Such information could be important for possible management decisions on the civil service in the future, including reallocation of civil servants, and projecting the changes in the wage bill. 35. Even without active retrenchment, the wage bill could decline through natural attrition. Nationwide, about 3 to 4 percent of the civil servants leave the service every year—whether through retirement, change of careers, or otherwise. Lotim could use this natural attrition to save on the wage bill by replacing those that leave with staff that is redundant elsewhere in the service rather than hiring new staff. Again, good information on the civil service will help in achieving this, together with training programs that would prepare a civil servant for a new assignment. 36. Finally, there is always the option of redundancy. Although a civil servant cannot be fired, redundant civil servants only receive 80 percent of their base wage18 and Government would save on functional or structural allowances. Reducing income in such a way would most likely be a strong incentive for the civil servant in question to leave the service. However, it would also bring hardship to the people affected, and if it applies to many, such redundancy policy could spark unrest among remaining civil servants. 37. Other routine expenditures. Lotim’s outlays on operations and maintenance (O&M) spending is a modest 6 percent of total spending. However, much O&M is hidden in the development budget. For instance, the regional government finances drugs for Lotim’s health centers. This started as part of a health project, but is now among the regular government spending financed from general revenues of the government rather than from project money from the center. Nonetheless, spending on drugs still figures on the development budget as a project. One of the reasons for this is the incentives in the civil service: projects come with additional income from management fees, per diems, and honoraria, which would not be allowed under routine spending. However, this practice distorts the true nature of spending, and consequently makes budget analysis and decision making by the government and the DPRD more difficult.

17 18

It is unclear whether for the transferred but not yet retired staff premiums for the state pension fund are to be paid. See Government Regulation No. 96/2000 and Government Regulation No. 32/1979 on dismissal of civil servants.

14

Figure 1 : Lotim - Shares of the Development Budget per Sector, FY2000 and 2001 Subsidy to Lower Regions Security and General Politics, Information, Communication, and Media Government Apparatus Law Technology Religion Housing and Welfare Health etc.

FY2001

Population and Welfare Education, Culture, etc FY2000

Living Environment Regional Development Tourism and Regional Communications Mining and Energy Transportation Trade Labor Water and Irrigation Agriculture and Forestry Industry 0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

Percent of Total Development Spending

Source: ABPD Lotim and Annex Table 4

38. Development spending. In the course of decentralization, Lotim was able to increase the development share of the budget. In contrast to many other regions, its development spending rose from 29 percent of the budget in FY2000 to 35 percent in FY2001.19 Total spending increased by 150 percent on an annual basis, from Rp. 29 bn, in (9months) FY2000 to Rp. 93 bn. In FY2001. Within this larger envelope, Lotim significantly shifted its priorities: education, heath, trade and the government apparatus itself saw large increases in shares of the development budget (Figure1). Among the losers regional development and settlement saw its share plunge. Transportation remains the largest sector on the development budget, followed by Government Apparatus. In this, Lotim is similar to other Kabupatens, but the region’s share of 17.6 percent on transport is far lower that the 28.7 percent average for all districts and cities. In contrast, Lotim now devotes a significantly larger share to health and education that the national average—about 24 percent of total compared to a local government average of 17 percent for Indonesia. Of course, on a per capita basis, social development spending still lags far behind the local government average for the nation: Rp.67,000 in Lotim versus 108,000 for the average local government in Indonesia. 39. The development budget includes about 190 individual projects in 21 categories. A major single development expenditure included a transfer for the two regionally owned banks. Missing in FY2001 were transfers to lower levels of government—sub-districts and villages. The government’s explanation for this was that regulations to transfer to lower regions were not yet ready at the time of budget and budget revision. This is likely to be problematic, as before decentralization villages received each year a development grant of Rp. 20 million per village. In 19

revised budget data

15

other districts in Indonesia, the regulations were apparently ready in time for the transfers to be effectuated. 40. Borrowing. In principle, Lotim has significant borrowing capacity. 20 Currently, the region does not borrow at all, and seems to have no outstanding debt. Law 25/99 and Government regulations 21 allow for borrowing from both external and internal sources in line with a region’s borrowing capacity. Broadly, this capacity depends on revenues and “hard to avoid” spending such as wage bill and ongoing project spending. Although it is as of yet unclear how central government interprets this, there seems to be significant scope for Lotim to borrow. If the rules from Government Regulation 107/2000 are applied, the region would have a total debt capacity of some Rp.106 billion. However, for 2001 an 2002, and perhaps beyond, access to borrowing is severely restricted. A Ministerial Decree has disallowed direct borrowing by the regions, and regions can only borrow from the central government. The reason for the ministerial decree is the central government fear that regional borrowing could disrupt macro-economic stability. There is some support for in the numbers for this position: total debt capacity of the regions amount to some Rp.47 trillion (35 trillion for local government, 12 for provincial government). If absorbed in just a few years, this would affect the center’s attempt to maintain macroeconomic stability. 41. Lotim’s debt capacity is only an indicative upper borrowing limit. Whether the region would actually be able to borrow this much depends on their creditworthiness as perceived by potential creditors—including banks, private individuals, and bilateral and multilateral financial institutions. Creditworthiness depends not only on the capacity to repay, but on willingness to repay, the quality of financial management and information of a region, and the overall legal environment for regional borrowing. As for the latter, Law 25 and PP 107 are probably too weak to allow for a significant amount of borrowing by regional government. The legal redress of creditors in case of default by a region is weak, and only the most creditworthy regions are likely to get access to credit from sources outside central government.

Note that in Lotim’s budget borrowing is still listed as revenues, and repayment is included in expenditures. This is the old system for the national budget, which meanwhile has been replaced. 21 Government regulation 107/2000 on regional borrowing limits total debt to 75 percent of revenues “minus necessary expenditures” and debt service to 35 percent of revenues “minus necessary expenditures.”

20

16

Planning and Budgeting Box 4: Vision and Mission of Lotim 42. The transition to a decentralized system has not been conducive to good VISION planning and budgeting. The To realize a religious society based on norms, rule of law and intergovernmental fiscal relations changed environmentally conscious within the unitary state of the Republic of Indonesia. dramatically, and expenditure responsibilities were devolved at the same MISSION time. Central decisions on resource Some missions have been determined to realize the future vision Lotim such as: allocation and on expenditure devolution of 1. the comprehension improvement on religion in their daily came late in the year preceding life. decentralization, whereas regulations on the 2. the implementation of Pancasila (five basic principles of RI) in the society life as a nation. planning and budgeting processes to be 3. consistently regional economy development based on a just people’s observed were released less than two economy and the acceleration of reg. economy to promote recovery. months before the start of the year.22 Some 4. realization of a real regional autonomy and responsible central government decisions affecting based region capability Lotim only became available during the 5. realization of a professional government apparatus, transparently and accountable as well as clean and free of year. These include the retro-active wage increase passed in July 2001 and the 6. KKN. developing the respect towards the legal system to uphold decision on Lotim’s share of resource the supremacy of law and human rights. revenues, which was only decided in June. 7. development of prosperous of society particularly in the basic needs properly. Central government allocations to projects 8. meeting realization of a democratic educational environment which in Lotim remained largely outside the provide the quality and skill full that could contribute to regional and national development. purview of the regional authorities—even 9. societal development in a sustainable management of natural if the projects were implemented in coresources. administration by the regional sector bureaus. In addition, much of the Source: Perda Kabupaten Timur Nomor 15, Tahun 2000 regulatory framework for budget management was issued too late (November 2001), and key pieces of regulation on financial management were still missing at the start of the budget year. Thus, the 2001 planning and budgeting exercise in Lotim was only an imperfect attempt to come to grips with decentralization.

Planning Cycle 43. There is an abundance of planning in Lotim. Building on the tradition of the fiveyear plans, decentralization has added a number of instruments for planning—including the POLDA, the RENSTRA, and the REPETADA (see Table 4: for an explanation of the terms). The POLDA is a broad statement of objectives for the Kabupaten, 23 and includes a statement of vision and mission for the Government (Box 4). The planning processes are aimed at achieving the objectives of the POLDA. Although participation of civil society and the village level has much increased24 the process is still much like

Government Regulation 108 was issued in November 2000. Peraturan Daerah Kabupaten Lotim Nomor 15 Tahun 2000, 23 December 2000. 24 Not least because the new regulations on planning prescribe this participation. 22

23

17

the Repelita process of before, and can be thought of consisting of a top-down process and a bottom up process. 44. Top Down. The MPR (Consultative Assembly, Indonesia’s highest constitutional body) provides broad development guidelines to Government (GBHN). Based on the GBHN, central government prepares the PROPENAS (national development program). Figure 2: Top-Down Planning GBHN

PROPENAS National

RENSTRA

National Consultation

National

PROPEDA Propinsi

RENSTRA

Regional Consultation

Regional

PROPEDA Kabupaten

RENSTRA

National

Consultation

Propinsi

Consultation

Kabupaten

REPETA

National

REPETADA

Regional

REPETADA

National

Consultation

Propinsi

Consultation

Kabupaten

Note: GBHN: Garis Besar Haluan Negara (National Guidelines); PROPENAS: Program Pembangunan Nasional (National 5 year plan); PROPEDA: Program Pembangunan Daerah (Regional 5 year development plan); REPETA: Rencana Pambangunan Tahan (National annual development plan); REPETADA: Rencana Pembangunan Tahuan Daerah (Regional annual development plan)

Coordination between central government and regions is done in a national coordination meeting (Rakorbangnas) in the Fall. At the provincial level, no such coordination meeting seems to be taking place. Regional governments (provinces, kabupaten, and kota) each prepare and develop Propedas (regional development plans) on their own. Propenas and Propeda have five year time horizons. As with the POLDA, they have no specific outputs or outcomes specified, nor does it have link to the budget, and the reports serve largely as a broad statement of objective. 45. Based on Propeda, each local government bureau (Dinas and Badan) prepares and develops a Rencana Strategis (Renstra) as strategic plan with a 3 years time horizon. The Renstra is aimed at establishing accountability for the heads of the region and the heads of the bureaus. The report includes some performance indicators, although they are not quantified. Based on the Renstra, the Repetada (annual plan) is developed. The Repetada is the only planning document that has a linkage with Lotim’s APBD—the budget. However, the Repetada does not include planned budget allocations, but is rather a listing of the programs that are to be funded in the budget.

18

46. Bottom Up. Top-down planning is mirrored by a bottom up process starting from the village level and below. Each sub-village (Dusun) develops a planning program in Table 4: Planning and Budgeting Documents Policy document GBHN, Broad guidelines Government guidelines

Legal form to Decision (TAP) of the MPR, higher status than a Law.

Function Every five years, the MPR prepares policy guidelines to direct the President, in the same meeting in which it elects the President

Propenas, (Program Pembangunan Law Nasional) 5 year national development plan

The Propenas defines cross-sectoral and priority policies that government needs to fulfill the GBHN. It forms the basis for advice to ministers and regions to develop and integrate their policy documents and programs.

Repeta (Rencana Pembangunan Law in 2000, not Tahuanan), annual development in 2001 plan.

The annual development plan defines priorities for the development budget which is part of the annual budget.

Renstra,(Rencana strategis) Ministerial decrees strategic plans of central government agencies

Strategic plans of central agencies, defining their missions and programs to implement the central government authorities defined in Law 22 of 1999 and government regulation 25 of 2000, including programs for supervising governance in regions and giving guidance.

Pola Dasar pembangunan daerah, Local Regulation (Basic policies of regional (PERDA) development)

Broadly defines development policies for the region

Propeda, program

regional

Regional Renstra

development Local Regulation (PERDA)

Development policies to suit local needs and capacities, in compliance with national law, including the Propenas.

Decrees of Head of Region, and heads of dinasses

Strategic plan of regional head and agencies, defining visions, missions and programs. Programs will cover control functions, service functions and development.

Repetada (Rencana Pembangunan Local Regulations Each region will develop an annual development plan based Tahunan Daerah)annual (PERDA) on the Renstra and review of evolving needs development plan Local Regulations APBD, regional budget Performance budget as per Government Regulation 108/2000 (PERDA) Source: Adapted from Report of the Indonesia Decentralization Assessment and Activity Identification mission, AUSAID, November 2001

Sub-Village Development meeting (Musbangdus). This process then feeds into the development of a Village (Desa) Development plan in Village Development Meeting (Musbangdes). Based on the village development plans, each Kecamatan (Sub-District) develops a Kecamatan planning program that is integrated into the Kabupaten Planning Program. In reality, it is hard to see whether the PROPEDA and REPETADA take into account the bottom up approach. 47. It is unclear how and when the top-down and bottom-up planning are integrated. The link between national plan and Lotim’s district plan may be there in name (the national plan is referred to in the regional plan) , but it is unclear who is responsible for ensuring the integration, or what would happen if a kabupaten decided to largely ignore the PROPENAS in its plan. Moreover, the PROPENAS has little or no fiscal consequence for the region. Thus, central government cannot “steer” the regional plans through financial incentives. At the provincial level, no integration or even coordination

19

of planning seems to take place between the regions. In a similar vein, the village planning, except for those projects that in the end make it in to the annual plan, are hardly coordinated with the district plan. Figure 3: Bottom-Up Planning

REPETADA RAPBD

RAKORBANG

DINAS

UDKP

BADAN

MusBangDes

MusBangDes

MusBangDes

MusBangDus

MusBangDus

MusBangDus

Note: REPETADA Rencana Pembangunan Tahunan Daerah (Regional Annual Development Plan); RAPBD Rencana Anggaran Pendapatan & Belanja Daerah (Regional Annual Budget Proposal); RAKORBANG: Rapat Koordinasi Pembangunan (Development Coordination Meeting); UDKP: Unit Daerah Kerja Pembangunan (Development Unit in District Level); DINAS/BADAN: Local Government Unit; MusBangDes - Musyawarah Pembangunan Desa (Village Development Meeting); MusBangDus: Musyawarah Pembngunan Dusun (Sub-Village Development Meeting).

48. A further weakness of the planning exercise is that it is largely void of considerations of affordability. Only the annual plan (PROPEDA) takes fiscal considerations into account. As in the REPELITA process before decentralization, this leads to much disappointment at all levels of government, once the fiscal headroom forces government to make sharp choices, and cuts in the “wish-lists” of villages, bureaus, and the district government as a whole. It is telling that Lotim’s budget makes reference to many Government Regulations, but not the POLDA, the PROPEDA, or even the REPETADA. 49. A third weakness of planning is its continued focus on the development budget and development projects, rather than on the budget as a whole. This should change over time, as regional governments are required to make “performance budgets” according to Government Regulation 105/2000. Such performance is hard to imagine 20

without a move to a programmatic planning and budgeting which integrates recurrent and development spending. But only a few regions—such as Semarang—have started preparation for moving in this direction. 50. Finally, There seems to be significant overlap in the role of some of these instruments. In part, this seems to be the result from transition from the old to the new. For instance, the Lotim’s POLDA of December, 2000, a planning document to set the basic policies for the rest of the regional government’s term (2000-2004) was just completed, when new central regulations required a new document—the PROPEDA. In Lotim, this document has much communality with the PROPEDA. Budgeting Cycle 51. Budgeting in Lotim remains largely an committee, all civil servants, are in charge of the process, and the Bupati and DPRD get in relatively late in the game (see Table 5). The budget process remains . Budget implementation involves a large number of checks on spending without much concern for results on the ground. Budget evaluation is hardly developed, and there is no independent audit of the regional government’s final account. 52. Budget Preparation. All spending units submit their budget proposals to the districts Komite Anggaran (budget committee) that consist of the Secretary of the Region (Sekwilda), the Finance Bureau, the Bappeda, the revenue office, and the legal bureau.

administrative exercise.

The budget

Table 5 : Lotim’s Planning and Budget Calendar

Approximate

Step

Date June/July

July-September September/October October

November

December During year

Rakorbang (development planning coordination meeting) Budget agencies prepare budget proposal National Rakorbang meeting Agency proposal to Komite Anggaran; budget discussions Submission of budget proposal (RAPBD) to DPRD DPRD Approval of ABPD Budget adjustment— usually one month before the end of the budget year Submission of unaudited PAD (Perhitungan Angaran Daerah)

53. The budget committee reviews revenue estimate prior to their review of the After budget year expenditures budget. Information on anticipated central government resources (i.e. DAU, DAK, Tax/Non-Tax sharing), which Source: Mission interviews. currently form the bulk of APBN regional revenues, are clearly important for the committee’s budget work. If there is some uncertainly about projected revenues, as was the case in FY 2001 with the late announcement of transfers from the center, the committee will mark certain expenditure items as conditional upon receipt of the central revenues.

21

54. The budget committee finalized the budget proposal document around October.25 The Bupati, as Kabupaten Chairman, then submit the Annual Budget Proposal (RAPBD) to the DPRD (Local Parliament). The DPRD then has at least two months to scrutinize the budget proposal. Early December 2000, Lotim’s DPRD approved the budget, whereas the regional regulation formalizing the budget was approved on March 1.26 55. After the DPRD has authorized the budget, the Finance Bureau and each spending units prepare the budget documents for the apportionment of the budget. Separate warrants are issued for routine budget (Daftar Isian Kegiatan or DIK) and for the development budget (Daftar Isian Proyek or DIP). The documentation underlying the warrants include the Petunjuk Operasional (operational guidelines-PO) and Lembaran Kerja (working paper-LK). The Finance Bureau identified the budget expenditures (line item starred) for which funding is conditional upon availability of revenues from central government funding within the budget year. 56. Budget Implementation. Once a budget document (DIP/DIK) has been approved, each spending unit submits quarterly spending plans, and a request for payment (SPP) to the finance bureau. The plans are accompanied by evidence of spending needs (wage bill, contracts signed) and past spending (invoices, etc.). The plans are checked against the DIPs and DIKs by the finance bureau—and sometimes by the internal auditors. Before a new SPP is approved for the next quarter, all spending is checked against the previous quarter SPP, and spending that is not in accordance with the SPP is deemed to be illegal, irrespective of whether such spending is included in the budget or in the warrant. Upon approval, the money is transferred to the bank account of the spending unit, from where actual payments are made. There is one bank account per dinas, and one per project. 57. Delays in issuing budget documents represents a significant problem to implementation. The main reasons for implementation delays are lack of information about budgets from center in FY 2001 and clarity of central regulation regarding decentralization fund. It creates more difficulty for the local government to achieve the development objective. 58. Budget Adjustment. The first year of decentralization, many regional governments in Indonesia were forced to engage in a major budget adjustment. The main cause of this adjustment was a centrally mandated increase in civil servants wages announced in July 2001, but retroactive until January 2001. This wage increase came on top of the increases due to the change in status (echelon) and pay of higher level regional civil servants effective at the time when the regional bureaus were merged with the central deconcentrated bureaus. For Lotim, the retroactive salary increase caused the wage bill to increase by some 18 percent. The impact of this was dampened by a

The date of finalization of the budget proposal makes it hard to see how the circular letter of the Minister of Home Affairs on budget preparation, dated November 17, 2000, could have influenced budget preparation. 26 Peraturan Daerah Kabupaten Lotim Nomor 3 tahun 2001. 25

22

allocation of Rp. 9 billion from the central government contingency fund.27 Some reductions in other routine expenditures dampened the increase in overall routine expenditures to 14.8 percent. An increase in overall receipts of 8.6 percent, including an allocation from the central contingency funds dampened the nominal decrease of in development spending to only 1.4 percent. 59. At the time of the team’s visit to Lotim, the budget adjustment was not yet passed by the regional parliament. Such budget adjustment traditionally occurs one month before the end of the budget year. The budget adjustment is more of a record of what was realized during the year rather than a decision on how to re-evaluate priorities in light of changed circumstances. 60. Procurement. Lotim’s procurement processes remain unchanged from before decentralization. For each procurement, a procurement committee appointed by the Bupati is set up. These tender committees usually consist exclusively of civil servants. Upcoming tenders have to be published in the local newspaper. The result of the tender committee has to be approved by the Bupati. The guiding regulations—presidential decree 18/2000—allows for preference for local bidders, and Lotim officials see procurement as a means to support local contractors rather than a way to get the best price for the goods procured. Lotim has not yet determined its own procurement processes—a move which would be allowed under Law 25/2000 (Art 62) and Government regulations 105 (Art.32). 61. World Bank operations in Indonesia suggest that improved tender processes can Box5 : Kontribusi generate savings of up to 20 percent. Assuming that 80 percent of the Lotim’s detailed budget on the revenue side shows a development budget is tenders this would post “Kontribusi,” or contributions. The finance office that it was common that contractors who won suggest that up to about 5 percent on total explained a contract with the regional government would expenditures could be saved on improving contribute to the region’s budget. The contribution was the tender process. One means to improve of course voluntarily, but in case a contractor would not procurement would be to allow for non- contribute, the government would try to persuade him to government representatives to participate in do so anyway. Halting payments on the contract was the procurement committees, a move which is considered to be one appropriate means of persuasion. allowed under Law 28/1999. Including outsiders in the committees would reduce the scope for collusion between committee and bidders, and thus limit the risk of corruption. A further move could be to publish the results of the tenders, and to open a channel of complaints on procurement processes. 62. Accounting and Reporting. Each budget user (Dinas/Badan) has the responsibility to maintain its own accounting system. The Bagian Keuangan (Finance Bureau) under the Sekretaris Daerah (Regional Secretary) has the responsibility to keep separate accounts for all spending units transaction based on Surat Permintaan The numbers are taken from: Rancangan Peraturan Daerah Tentang Perubahan Pertama APBD Kabupaten Lotim, August 2001. 27

23

Pembayaran (SPP = Payment Requests) and Surat Perintah Pembayaran (SPM = Payment Orders) issued by the Kasda. There is no accounting standard yet applied for the government entity. Basically Kabupaten follows cash basis accounting with modification on assets registered. 63. All units are required to submit a quarterly reports to the Bagian Keuangan. These are then consolidated before the submission of a report to the Bupati. The completion of the report includes the reconciliation of revenues between the Bagian Keuangan with the Dinas Pendapatan (Revenue Unit) and the Kas Daerah. However, it is not clear how Bagian Keuangan monitor and reconcile all unit revenues and expenditures. At the end of fiscal year, an unaudited financial report compiled by the Bagian Keuangan, is submitted by the Bupati to the DPRD for further review. 64. Financial Information System. Lotim has been developing an integrated financial management information system for some time now. However, there is no significant progress as of today. A high priority for central government is to issue Government accounting standard for central as well as local government. The development of regional accounting standards, permitted under current regulations presents a significant risk (e.g., in developing transparent bookkeeping from auditing purposes) and wastes efforts due to duplicated efforts. Furthermore, multiple audit and reporting standards make it difficulty for central government to produce a consolidated financial report for the whole government. 65. Central government now follows a modified cash basis to prepare financial reports. Local governments still follow MoHA regulations to prepare their financial reports. MoF has recently been preparing an accountancy standard for central as well as local government for their preparation of financial reports. However, socialization remains limited and training has not yet occurred for local staff. MoHA has also been updating standards, and the full compatibility of the MoHA and MoF standards for FY 2002ff is still not clear. 66. Budget Evaluation. Budget evaluation is hardly developed in Lotim The report that comes closest to evaluating budget implementation on results achieved is the accountability report of the head of region.28 Because the budget documentation does not specify targets, it is hard to evaluate outputs and outcomes, and the accountability report largely reports what the development budget was spent on. In some cases, outcome indicators are quoted, but from the report it is unclear whether these result from the policies pursued, or whether although the accountability report 67. Auditing Arrangements. Audit arrangements at the local level still remain confused. Law 25/99 determined that the audit of local government budget should be done by “prevailing regulations.” At that time, those prevailing regulations assigned the Government’s internal auditor (BPKP) the supreme audit authority (BPK), the Pertanggung Jawaban Kepala Daerah Kabupaten Lombok Timur, Tahun Anggaran 2000, 31 March 2001. This report is based on PP108/2000 on the accountability of the head of region and Regional Regulation No. 4/2001. 28

24

Inspectorate General of the Province and the Inspectorate General of Home Affairs a role in audit. Since then, Presidential Decree 74/01 has been issued, changing the audit arrangement. The decree assigns three internal auditors the right to audit local governments’ budgets: (i) Badan Pengawas Daerah (Bawasda) Kabupaten (ii) Badan Pengawas Daerah (Bawasda) Propinsi; (iii) Inspectors General of Line Ministries on technical aspect. The Inspector General of MOHA plays a role as coordinator of Bawasda assignment. 68. The Bawasda prepares an audit plan (Program Kerja Pemeriksaan Tahunan, PKPT) and submits it to Bawasda Propinsi for regional compilation plan and to MOHA IG for national compilation. The Bawasda Kabupaten may request the Bawasda Propinsi to conduct an audit assignment if it does not have adequate capacity and resources. Furthermore, the Bawasda Kabupaten will request the Bawasda Propinsi to audit units under Sekda. This is due to a structural weakness, which placed the Sekda at a higher echelon/rank than the Bawasda. Such a difference in seniority will often engender the lower rank to be reluctant to perform the audit. 69. Prior to FY2001, the district’s Bawasda audits were limited to the subdistrict and village level budgets. Based on Keppres 74/2001, the Bawasda has now expanded its audit scope to all Kabupaten budgets, including in theory the SekDa office under which it resorts. There are 142 audit objects with 2 weeks time for each assignment. The Bawasda still focused on compliance audits, and has yet to broach performance audits. There is still plenty room to increase the audit scope of Bawasda in terms of duration and depth of the audits. In addition, no reviews or evaluations on the result of budget spending are conducted in Lotim. The auditor has also yet to audit the whole Kabupaten’s financial report, or conduct a financial audit on BUMD (Local Government Owned Enterprises). 70. The limited capacity of Lotim’s Bawasda is already overloaded at the current scope of audit. The Bawasda has 20 auditor staff, including five structural staff. Only three staff had university degree. No staff had an accounting or auditing background. The Bawasda staff realized that they fundamentally lack the capacity to conduct their audit assignment. 71. As internal auditor, Bawasda submits the audit report to Bupati and copy to Auditee unit, Bawasda Propinsi and IG MOHA. Bawasda audit reports are not available to the public, nor to the DPRD. In the past, the issuance of an audit report was often preceded by a process to “correct” the mistakes found. This process, which sometimes resulted in negotiation between auditor and auditee, 72. Lotim’s DPRD has not yet exercised their budget controlling and monitoring authorities through the audit report, nor has it had public hearing on the audit reports. The DPRD chairman showed his interest to have audit reports as a tool to control and monitor the budget realization. As most DPRD meetings are open for public, hearings on the audit report could also increase public knowledge on the result of the audits.

25

73. Lotim’s DPRD could consider expanding the capacity of the internal auditor by allocating more funds. But perhaps more important is a position of the auditor independent from regional government. Central government could consider revamping Keppres 74/2001, and making the Bawasda an independent auditor reporting directly to the region’s parliament. But if that does not happen, there seems to be little that stops Parliament from appointing its own independent auditor—whether BPK or a commercial auditor. 74. Indonesia’s Supreme Audit Agency BPK, the only external auditor in the country, has according to Law 5/1973, the authority to audit all levels of government. The current draft law on the State Audit which has been submitted to Parliament in September 2000 puts this authority in doubt, however. Other countries have solved this by giving the Supreme Auditor the option to audit regional government expenditures, while at the same time prescribing that regional government chooses an external auditor—which could but need not be the Supreme Auditor. In France, the Court des Comptes accredits regional external auditors, and it plays a role in the appeals process against verdicts of Other countries, such as the Netherlands, do not prescribe an external auditor for local government, but regional assemblies usually chose one to report to an audit committee. A Budgetary Strategy for Lotim 75. Lotim’s budget has come through the Big Bang decentralization with significant, but manageable problems. The budget adjustment within the year remained limited compared to other regions, and development spending retained levels comparable to the past. The planning and budget process, although adjusted with several new elements, remains largely as before decentralization, although the DPRD and to some extend civil society has started to play a larger role in the budget process. Resources remain limited, and will likely be so for quite some time to come. At the same time, the region faces numerous socio-economic challenges for which government action—and government resources—are required. Looking forward, Lotim has to decide on an overall budgetary strategy to address its public policy issues. Broadly, this strategy can be divided into one to raise revenues, and one to use those revenues better through improved public expenditure management. 76. Raising revenues. It seems clear that the DAU will remain Lotim’s main source of revenues for the foreseeable future. But central government plans for more special allocations could benefit regions such as Lotim. And while revenue potential from own taxes and retributions is likely to remain limited, they should be carefully considered. § DAU. As a poor, resource-scarce region, Lotim is likely to benefit from a more equalizing DAU. To achieve a more equal distribution, the region could consider lobbying with DPR and DPOD representatives form the regions—although currently these representatives are predominantly drawn from the resource rich regions, and have shown little interest in a more equalizing DAU. Lotim could

26

also align itself with regions in a similar situation, in order to make its lobbying for a better DAU more effective. §

DAK. Lotim may be one of the first to benefit from special allocations from the center once they come on stream, but as of yet this is an uncertain prospect. However, the existing deconcentration and co-administration (medebewind) resources through the regional dinasses could already be quite big. The government could start taking these into consideration with its own budget allocation, and a first step in this would be taking stock of what resources actually flow to the region.

§

Own Revenues. While Law 34/2000 provides the regions with expanded opportunities for raising revenues, this is a double-edged sword. Raising the wrong taxes could hurt the local investment climate, vulnerable groups in society, or both. And even if successful, those efforts will have only a limited impact on the overall budget—given the low base that own revenues are starting from. Yet, the government could already start studying options for raising more own revenues. A special committee with representatives of private sector and civil society along with Government and DPRD members could be set up to consider options. Such a committee could consider possible taxes and retributions against criteria in the law, and criteria developed in by Lotim. Such a process could lead to a broader revenue base acceptable to the population, and least damaging for economic activity and the poor.

§

Revenue Administration. Before even raising new taxes or levies, Lotim could review the collection performance on current taxes and levies. In addition, it could consider closer cooperation with the central tax administration which collects land and real estate taxes and personal income taxes—both of which are shared with the region.

77. Improving public expenditure management. Lotim’s largest gains from autonomy are likely to come first from a better use of existing resources, rather than from an increase of those resources. Improvements in the budget management process are key for achieving those gains. To some extend, this process of improvement will be driven by central regulations, but in other areas Lotim could take the lead and improve its mechanisms by which it allocates and spends public money. 78. The prescription in the law on the topic of local budgeting is for the regions to have a “performance budget.” While this may be hard to achieve in the short run,29 the general direction of more performance orientation is a good one. Indonesia’s tradition of budgeting already contains a number of elements that could reinforce such a performance orientation, including strong strategic planning capabilities, and tight expenditure control. But equally, many elements are as of yet missing, and developing them in Lotim will, as elsewhere, not be easy, and will take a long time. The In fact, even in OECD countries, only one—New Zealand—has introduced a full performance budget based on output budgeting. 29

27

government could therefore consider as a first step to set up a special task force or committee to guide the reforms in budget management. A first task of the committee or task force would be to come up with a strategy that could guide the reforms. 79. Performance orientation requires a budget management process to do three things: (i) maintain fiscal discipline, or spend within your means; (ii) allocate money to the government’s priorities; (iii) provide incentives for efficient service delivery. In each of these areas, Lotim could take some immediate steps, and initiate medium-term reforms. In the short run, the government could: §

To improve fiscal discipline, Lotim could consider o Issue an overall spending limit at the beginning of the budget process, and agree on this with the DPRD; These spending limits should drive the planning process at the district and lower levels. o Include all resources—including central resources—in the budget documentation, even if the DPRD does not vote on it. o Make budget adjustment an exception rather than a rule, and set explicit regulations on virement (changes of spending from one DIP to the other).

§

To improve prioritization, the Government could: o Introduce a program classification for budget purposes rather than a split between development and routine budgets o Issue sectoral spending limits at the beginning of the budget process, and agree on these with the DPRD. These spending limits should drive planning by the dinasses for both recurrent and development purposes. o Require sector bureaus to cost out their Renstra for both recurrent and development spending. o Review all existing development spending and determine whether it is still in line with the Government’s priorities. o Start an experiment with one service delivery agency, or one dinas to prepare output indicators and link them to budget spending

§

To improve efficient use of resources, the government could: o Start a review of the civil service and other public sector workers, and consider strategies for reallocation and retrenchment; o Start an experiment with one dinas to allow more flexibility for in-year use of resources, and limited carry-over of funding from one year to the next. o Start an experiment with one dinas or one service delivery unit to solicit client feedback. o Review the DIP/DIK process to ensure timely availability of funding. o Streamline the ex-ante review and audit of spending to ensure timely implementation of the budget, while at the same time initiating external audit. o Widen the membership of procurement committees to include nongovernment members.

28

o Propose to the DPRD to hire an external auditor that would report to parliament. 80. In the long run, a whole host of measures would be required, including changes in planning processes and methodology, in budget classification, accounting, and reporting, monitoring of results, budget evaluation, financial information system, and even government organization. The next section further discusses service delivery options for Lotim. The final chapter goes deeper into the role of civil society in policy planning and budgeting. SECTORAL BUDGET MANAGEMENT 81. In this section the potential for greater budget efficiency and effectiveness in Lotim’s technical services is explored. Examples are drawn both from initiatives that are already underway in Lotim and that could benefit from further adaptation, and from potential opportunities which Lotim is yet to exploit. By efficiency is meant current levels or better achievement of activity’s goals at reduced unit costs – “stretching” the budget. Effectiveness is explored in terms of improvements in the demand responsiveness of delivery of a technical services by sector agencies. This section identifies areas of needed administrative capacity strengthening. After expanding briefly on these three notions – budget “stretching, demand responsive delivery of services and administrative capacity building – these are applied in more detail through a review of a sample of technical services in the education, public works, and agriculture sectors. 82. “Stretching” the Budget. Opportunities are readily apparent in Lotim for more efficient use of budget resources. These include through cost recovery for services, cost sharing for small-scale investments, and out-sourcing activities to non-civil servants. 83. New measures to introduce cost recovery for delivery of services have begun in Lotim just this past year. These consciously seek to correct initiatives in the past to establish sustainable financing mechanisms, which under recent conditions had degenerated into a system of informal charges by service providers with payments not usually entering into the local treasury. 84. These cost-recovery mechanisms are positive steps but a number of aspects merit further attention. It would be useful to make some quick estimates of actual unit costs of delivery of these types of services in Lotim. These would serve several purposes. First, having a benchmark of actual real costs would be important to know how far cost recovery would have to increase to completely cover these costs. Second, this could then be used to assess whether different types of service users could, under existing market and household conditions, afford these actual costs. Third the unit costs themselves should be scrutinized to see whether the process of service delivery can be improved so that unit costs can be reduced. In brief, once a strategy of phased introduction of cost recovery is embarked upon, it is important to have a quantitative

29

target and up front awareness of how this will impact the economics of the activity involved and different types of services users. 85. The broader issue this poses to local stakeholders, however, is how to introduce formal service fees when the background is a citizenry that is disillusioned with the recent history of non-transparent and often unequal application of informal charges. When a service administration carries this reputational baggage with it into the new environment of decentralization and democratization, it is inadequate to simply legislate a new set of formal service fees without direct actions to remove the informal fees. The key stakeholder in ensuring that this be adequately addressed is the DPRD, for whom reelection depends in some measure on bringing a more transparent and cleaner government into place. Simple measures in this direction could be determined in relation to the annual budget hearings for a Dinas, whereby new cost-recovery fees would only be approved by the DPRD in conjunction with a specific strategy and public campaign for clean governance of that service and elimination of informal fees. Citizen and NGO stakeholders could be mobilized to monitor implementation and scheduled to testify prior to the next budget cycle with progress by the Dinas being an important factor in consideration of their budget proposal for the subsequent year. In brief, the executive and legislative branches would be wise to avoid charging ahead too simply with formal fee increases without coupling this with clean government measures that are visible, credible and monitorable by the public. 86. Cost-sharing in the implementation of small-scale investments particularly involving infrastructure is a being pursued by a number of Dinas in Lotim. Cost-sharing is a useful way to extend the limited resources of the APBD by partnering with the capacity of beneficiary groups to contribute to projects that directly benefit them. Moving forward, the Dinas involved may find themselves coming under pressure to respond to two types of pressure from the partnering groups. One is to establish clear and transparent rules by which the negotiations over cost-sharing take place. It is possible to establish principles by which different partnering groups contribute different shares, but these need to be clearly established to avoid misunderstandings and conflicts. Second is to move rapidly to implementation where the APBD share of investment cost is transferred as budget to the partnering group rather than as goods delivered by the Dinas. 87. Where this has occurred elsewhere in Indonesia, the partnering group has generally not only been able to acquire the implementation and bookkeeping skills adequate to handle the financial and contractual responsibilities, but also able to deliver lower unit costs with acceptable quality compared to similar works delivered by contractors reporting to the Dinas rather than partnering group. A phased turnover of responsibility to the partnering group may be defined in consultation with representatives of functioning groups but more than a several year transition would be excessive. The DPRD would likely be interested in results of monitoring of unit costs and quality of such irrigation investments during such a transition since lowering costs means being able to reach more communities, faster, for the same budget allocation.

30

88. A third vehicle for making APBD resources stretch further is outsourcing and privatization of services and activities which in the future do not need to be carried out by civil servants. Lotim is already pursuing this in hiring contract teachers, and for livestock artificial insemination. There are other opportunities, such as in the physical operation of irrigation systems. 89. Demand Responsive Delivery of Services. The technical services of Kabupaten government will become more effective as their capacity increases to respond to demand from citizens for their services. This demand needs to be understood in various dimensions particular to each service and could include, for instance, timing, quality, cost and content. 90. An important part of capacity to respond to demand starts with the capacity to measure such demand in its important dimensions. As a general observation, the technical services of Lotim’s government are developing their tools for getting a better measure of relevant demands on services, but can do much more in this area. A number of tools for understanding demand are already being used. For instance, a University of Mataram survey on livestock farmers’ preferences for breed selection in artificial insemination services is guiding the Dinas in providing semen supply and insemination services. 91. A variety of simple mechanisms could be used more effectively. Surveys of clients should include not only seek to prioritize technical messages for extension programming but also do ex post assessments of quality and effectiveness as viewed by the target groups. Another source of useful information for programming and targeting of extension capacity is the revealed preferences that emerge from open menu revolving fund programs. Most of these programs involve a verification phase at the Kabupaten level, and much more effective use of the self-help group proposals reviewed could be made to program and target extension personnel with skills appropriate to the proposed activities. Bringing greater demand orientation into the operations of the Kabupaten’s Dinas’ will also be able to move forward more quickly with the strengthening of community organizations since these can most effectively articulate peoples’ priorities. An example of how this may develop in the future is presented in the agriculture section below for the irrigation sub-sector. Government Capacity Strengthening 92. The challenges identified in this report for improving public budget resource management, on both the revenue and expenditure side, highlight the need for strengthening of local government capacity in a number of areas as decentralization evolves. It is useful to think both in terms of new skills needed by civil servants and of new/strengthened processes in both the administration and legislature.

31

93. As the development budget moves in the direction of more implementation directly by community groups through direct resource transfers and cost sharing, Dinas will need to become more adept at facilitating these groups’ in chosing, planning and implementing priority activities. Further training in such facilitation skills will be an important complement to the existing technical skills of Dinas. 94. With streamlining of the public administration through outsourcing, Dinas will be less engaged in direct implementation and more engaged in process management, contracting, monitoring and quality assurance, and skills in these areas need to be strengthened. 95. Specific areas of strengthening were identified in Sekda and Bappeda. PAD enhancement and cost recovery strategies need to be tempered with ability to do incidence analysis so as to avoid new taxes and fees that overly burden poorer segments of society. Conducting such incidence analysis can be done as either a development planning analysis by Bappeda or in conjunction with each new perda by Dispenda, but in either case, the agencies involved will need to acquire new skills through training or recruitment. An important administrative capacity that may soon come under pressure is in the Sekda’s Sub-Bagian Kepegawaian which processes all of the district civil servant issues including teachers and non-teaching staff at school level which are the majority of the civil servants for Lotim to manage. If payroll management for these is transferred to the district, this will add enormously to this unit’s responsibilities and its capacity will need to be increased. 96. Building government capacity that puts demand-responsive service delivery at its core will require progress on defining new roles for the executive and legislature. Leadership in putting demand-responsiveness as a top priority would need to come from the DPRD, particularly in its annual review of Dinas budget proposals. Bappeda would need to be proactive in developing a proposed strategy, and implementing the approach approved by the DPRD, which while setting overall targets and timeframes, could also define a transitional period that begins with a pilot phase with several vanguard Dinas and services. Managing Cross-Kabupaten Issues 97. With decentralization being taken in hand by Kabupaten’s there is concern that public services and activities that require coordination or implementation across Kabupaten’s will be neglected to the detriment of public welfare. It is only natural that in the first year or two of Indonesia’s ambitious decentralization pace that Kabupaten governments will focus on assuring the basic functionality of their own domain before concentrating much effort on the more delicate tasks of cross-Kabupaten collaboration. But if this persists it could quite quickly lead to problems of deferred maintenance of infrastructure, degradation in the quality of services, unnecessarily high costs to the public of service provision at too disaggregated a level, and distorted incentives resulting from uncoordinated fees, taxes and regulations across Kabupaten borders. Although this report’s focus on only one Kabupaten has obvious limitations as a platform for exploring

32

such issues, results from interviews and fieldwork do provide some useful observations on the topic. 98. First, there appears to be a quite high awareness in Lotim of these concerns about cross-Kabupaten coordination with decentralization and willingness to overcome potential obstacles. Lombok may be a special case, since there the island has only three kabupaten and one kotamadya, there is a strong history of practical collaboration in various public sectors and the Lombok identity is quite strong. These characteristics may facilitate cross-Kabupaten collaboration although these favorable conditions do not apply for larger provinces with ten or more Kabupaten. 99. Second, in this first year of decentralization, there is ongoing collaboration that appears to continue with relatively little difficulty. A few examples are illustrative. In animal health, a Lombok-wide program to eradicate a communicable cattle disease that would facilitate inter-island trade and export requires cross-Kabupaten cooperation in monitoring, vaccination, and culling of infected animals. The responsible Dinas in Lotim reports that this collaboration is continuing satisfactorily for this priority program. In the fisheries sub-sector, exports of fish require laboratory sampling for health certification. There is only one qualified laboratory on Lombok in the capital, and the fisheries Dinas in Lotim considers that because of professional requirements, economies of scale and current volumes needing inspection, there is no merit in Lotim attempting to establish its own capacity for this. 100. There is tension between Kabupaten’s and the Provincial level regarding their respective roles, as reflected in a number of sectors looked at for this report. For instance in education, the Provincial Dinas for Education, Youth and Sport (Dikpora) has its role defined by a Provincial Decree including facilitation on various issues, but Kabupatens have so far not been responsive to attempts at coordination. In the longer run, if coordination modalities are not worked out, this will be to the detriment of the development of curriculum and quality standardization, teacher career development and promotion and educational management systems. 101. Safeguarding expenditure levels for O&M of public infrastructure classified as Provincial is also a real issue. Provincial budgets are incapable of maintaining previous allocation levels, and Kabupaten governments have so far proven unwilling to budget for these civil works, despite, at least in the case of Lotim, having substantially increased their O&M budgets for Kabupaten infrastructure. Coordination is also needed between neighboring Kabupaten’s to avoid boundary ruptures of maintenance quality, and for this coordinating meeings between respective Bappeda’s and Dinas’ will need to become a regular feature of future budget planning. The two major works involved are roads and irrigation networks. 102. Problems with fiscal incentives may also arise as Kabupaten’s issuance of Perda’s on local taxation and fees become more concrete, especially as differences in coverage and levels emerge. Comparisons of levels of tax and fee burdens may result in displacement of mobile economic activities, and dissatisfaction by service users and

33

taxpayers to the extent that they compare service quality and payments in their own and neighboring Kabupatens. Lotim officials are aware of this issue, as are Provincial level officials, but as yet little attempt has been made to deal concretely with this. Sectoral Perspectives Education 103. In 2000, the population of Kabupaten Lombok Timur (Lotim) was 968,755 people, of which 23 percent were in the 7 to 15 year old age group that is supposed to attend school. In many parts of Indonesia, parents often decide that children need to help with family economic activities, or cannot afford to attend school. But in Lotim there were 12% more children enrolled than the total school age population for primary school, more than 100% because those of junior secondary age were still at the primary level. Lotim has succeeded in mobilizing children to attend primary schools, but has teaching quality problems as evidenced by many primary school graduates having difficulties in reading, poor writing and arithmetic. 104. For grade 7 to 9 of junior secondary school, Lotim faces a dual problem: low enrollments and of poor quality. Of the total 71,074 population who are 13 – 15 years old, only 67% enrolled because of poverty, shortage of schools and parents preferred their daughters to get married at very young age or working. Even if they enrolled, many pupils drop out during the year. In 2000, 820 pupils (almost 2%) of primary and 1805 students (4%) of junior secondary school dropped out from the schools. 105. Private provision of education is important in Lotim as shown in Table 6. One fifth of the total students are enrolled in madrasahs (Moslim schools), almost all of which are private. Students in pesantren (from Quran school) reach almost the same number, such that the total proportion of community schools is 38 percent of all students. By school level, the community owns 99% of pre-schools, 5% of junior secondary, 15% of senior secondary, and 70% each of junior and senior secondary madrasahs, and 18% of primary madrasahs. In contrast, 82% of the primary schools, 24% of the junior secondary and 14% of the senior secondary schools are public or under the control of the dinas.

34

Table 6. Lombok Timur: Schools, Students and Enrolments, 2000 School

Pre schools

Number of schools and students

Total 97

Public

SD/SLTP/SMU (decentralized), % of Total

(Madrasah)** (centralized), % of Total Public 0

1

Private 96

Public 1

Private 99

Gross Enrollment, %

Private 0

Students

4,695

115

4,580

2

98

--

Primary schools

814

667

147

82

0

1

17

Students

146,338

125,798

20,540

85

0