GOVERNANCE E Q U I TA B L E G R O W T H , F I N A N C E & I N S T I T U T I O N S N OT E S Personal Income Tax Piggybacking Muhammad Khudadad Chattha Jürgen René Blum Roy Kelly © 2023 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover photo: iStock cnythzl >>> Acknowledgements This policy note was written with the guidance of Alma Kanani (Practice Manager, Governance Global Practice, World Bank) and Habib Rab (Lead Economist, World Bank, Indonesia). The note has been enriched through the peer review comments from Anna Custers (Economist, World Bank), Roy Bahl (Regents Professor Emeritus, Georgia State University), Bhimantara Widyajala (Director, Ministry of Finance, Indonesia), Fadliya (Deputy Director, Ministry of Finance, Indonesia), Rika Hijriyanti (Head of Section, Ministry of Finance, Indonesia), and other colleagues from the Ministry of Finance, Indonesia. Dr. Dubravka Jurlina Alibegović (Former Minister of Public Administration, Croatia), Anita Grebenar (Ministry of Finance, Croatia), and Bernardica Stipić (Ministry of Finance, Croatia) provided details about personal income tax piggybacking in Croatia. Nola Safitri (Program Assistant, World Bank) and Alfi Naufida (Consultant, World Bank) helped organize a peer learning session on Croatia’s personal income tax piggybacking for the Ministry of Finance, Indonesia which helped with the Croatia case study in this note. Carissa Hanjani (Consultant, World Bank) provided excellent data support. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 3 >>> Abbreviations CIT Corporate income tax DAK Dana Alokasi Khusus (Special Allocation Fund) DAU Dana Alokasi Umum (General Allocation Fund) DBH Dana Bagi Hasil (Revenue Sharing Fund) DGFB Directorate General Fiscal Balance DG Tax Directorate General Tax DPRD Dewan Perwakilan Rakyat Daerah (Local parliament) NIK Nomor Induk Kependudukan (ID number) NPWP Nomor Pokok Wajib Pajak (Tax ID number) OSR Own-source revenue PAD Pendapatan Asli Daerah (Locally generated revenue) PIT Personal income tax PPh Pajak Penghasilan (Income tax) Rukun Tetangga / Rukun Warga (Neighborhood Associa- RT/RW tion / Citizens Association) SNG Subnational government VAT Value-added tax EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 4 >>> Contents Key Messages 7 1. Revenue Mobilization in Indonesia 10 2. Classification of Subnational Government Revenue and PIT 12 Piggybacking 3. Autonomy of Subnational Governments and Fiscal Contract 16 4. Impact of PIT Piggybacking on Subnational Government 19 Revenues and Equity 5. Correct Identification of Taxpayer Residence 22 6. Reform Trajectory 24 7. Conclusion: Implications for Indonesia 29 Annex 1: Methodology for PIT Piggybacking Revenue Simulation 32 Annex 2: PIT Piggybacking Revenue Simulation Results (Total 33 Revenue) Annex 3: PIT Piggybacking Revenue Simulation Results (Per Capita) 34 >>> Figures Figure 1: Low tax collections compared to structural peers 10 Figure 2: Widening tax gap in Indonesia 10 Figure 3: Decline in Tax Effort Over Time 10 Figure 4: CIT and VAT are the biggest sources of revenue 10 Figure 5: Districts primarily rely on central government transfers for their 11 spending needs Figure 6: Typology of SNG revenue sources by level of autonomy 12 Figure 7: 2019 GDP per capita of countries with PIT piggybacking compared 13 with Indonesia Figure 8: Simulation of potential district revenue from the introduction of PIT 21 piggybacking as a share of PAD Figure 9: The structure of the taxpayer identification number (NPWP) 23 Figure 10: PIT revenues in Indonesia compared with those of other countries 27 >>> Tables Table 1: How PIT piggybacking can improve the “fiscal contract” and revenue 8 Table 2: Structural differences between shared revenues (DBH) and own-source 15 revenues (PAD) Table 3: PIT piggybacking with or without rate setting autonomy 19 >>> Boxes Box 1: PIT Piggybacking in Denmark 17 Box 2: PIT piggybacking in Croatia 25 >>> Key Messages Personal income tax (PIT) piggybacking is a local surcharge levied by SNGs on top of the taxable personal income or on the personal income tax liability already being levied by the central government. In contrast to tax sharing arrangements, piggybacking provides more SNG autonomy since the SNG is granted the power to set and levy the piggybacking rate, typically within certain bounds established by the central government, thereby strengthening the accountability between SNGs and their residents. Different versions of PIT piggybacking have been implemented largely in high-income countries, including Denmark, Norway, US, Canada, Spain, and Portugal. Croatia, a middle-income country, has adopted a PIT piggybacking system. While PIT piggybacking is an important source of SNG revenue in these countries, if Indonesia were to adopt a PIT piggyback, it would be one of the first few major middle-income countries to do so. A. For Indonesia, piggybacking on the PIT could serve two main objectives: i. Strengthening SNG accountability to their taxpayers (fiscal contract) PIT piggybacking could potentially strengthen SNG accountability to their taxpayers (fiscal contract) if the following conditions are met: a. SNGs are given the power to set the PIT piggybacking rate within the possible limits set by the central government. b. The PIT surcharge revenues levied by the SNG are allocated to the residence of taxpayers to strengthen the accountability link between the PIT surtaxes paid and the public services received. Linking the taxes imposed and the SNG services received by the taxpayer could encourage greater accountability between SNG-elected leaders and their constituency (taxpayers) and greater fiscal efficiency through linking the local-level revenues and expenditures. c. Taxpayers/voters are informed that the PIT piggyback is a local tax (Pendapatan Daerah, in the case of Indonesia). The salience of the PIT piggybacking will be improved by modifying the PIT returns to show the local surcharge separately, by implementing information campaigns to explain the change to citizens, and by ensuring that PIT piggybacking rate setting procedures are detailed to require approval from the local parliament (DPRD, in the case of Indonesia). EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 7 ii. Increasing public revenues Option 2: The Indonesian government could introduce an option to replace the tax sharing portion of the PIT PIT piggybacking could potentially improve domestic revenue DBH system by providing SNGs with the power to levy mobilization in Indonesia. Thus, if the Government of a range of PIT piggybacking rates with a floor rate of Indonesia wanted to shift from relying on a PIT tax sharing 25% and a ceiling above this. This option would also system (DBH) to introducing a piggybacking PIT-PAD system, provide SNGs with the flexibility to increase their PIT it could consider two options as summarized in Table 1. revenues.2 Providing this range would allow SNGs the options to levy a slightly higher tax rate than the current Option 1: The Indonesian government could PIT structure which would enable the policy change to introduce an option to allow SNGs to levy a separate be either revenue neutral or revenue positive. Having PIT surcharge on the central government tax base or a floor rate for the PIT opsen would reduce the risk PIT tax liability. Our simulation results show that such that the shift from a DBH structure to a PAD opsen an arrangement could potentially increase district structure would be revenue negative. OSRs by an average of 8.31%.1 Under this option, the central government could keep the existing PIT Enabling additional tax revenues to be collected through DBH system and continue sharing those PIT DBH the introduction of PIT piggybacking could help provide the revenues under the existing allocation formula and, government with needed fiscal flexibility to address economic at the same time, introduce a separate PIT surcharge crises in the future, like the COVID-19 pandemic. This is because option for SNGs. This option would be simple and PIT, being a progressive tax, can help raise more revenues from transparent to the various stakeholders, enabling a high-earning individuals without burdening lower- and middle- strong fiscal contract accountability link between the income households. This in turn could provide the government SNG and its residents. with additional fiscal space for countercyclical spending without significantly hurting private consumption. >>> Table 1: How PIT piggybacking can improve the “fiscal contract” and revenue Provide SNGs with bounded autonomy in setting the Provide SNGs with bounded autonomy in setting the PIT PIT piggybacking rate. piggybacking rate • If piggybacking were designed as a separate • If piggybacking were designed as a separate SNG PIT SNG PIT surcharge, then the fiscal system could surcharge, then the system could be revenue positive if Fiscal be revenue neutral if an SNG decides not to levy the SNG decided to levy a separate PIT surcharge. Contract Improve a PIT surcharge. • If piggybacking were designed to replace the current DBH • If piggybacking were designed to replace the system, it would be important to establish the floor rate at current DBH system, it would be important to 25% to keep the revenue neutrality while setting a ceiling establish a floor rate at 25% to keep the system rate available to SNGs higher than 25%. Revenues would revenue neutral. only be positive if the SNG levied a rate higher than the central-government-mandated 25% rate floor. If there was no SNG discretion in setting the PIT PIT revenue yield would be positive if the mandated fixed PIT piggybacking rate, there would be no improvement in piggybacking rate were to be set above 25% but the fiscal strengthening the fiscal contract. When replacing the contract would not be strengthened if there would be no rate No PIT DBH system, this would imply the need for the setting autonomy given to SNGs. Mandating such a central- Change central government to impose an SNG fixed rate of government-imposed minimum PIT piggybacking rate would 25% of the PIT liability with no rate setting autonomy go against the principles of regional autonomy. given to SNGs which would go against the principles of regional autonomy. Revenue Neutral Revenue Positive Revenue Implications Note: DBH = Dana Bagi Hasil (Revenue Sharing Fund), PIT = Personal Income Tax, SNG = Subnational Government. 1. The simulation is based on district-level PIT tax sharing (DBH) data shared by the Ministry of Finance. The model assumes that the PIT piggyback/opsen tax rate is set such that it raises additional district-level revenue equivalent to PIT DBH. This analysis only focuses on the district-level PIT piggyback and excludes districts in DKI Jakarta. Further details about the simulation are given in Annex 1. 2. For administrative simplicity, flexibility should be given to SNGs for them to choose only a selected number of rates between the ceiling and the floor. A minimum rate of 25% would ensure that the total PIT revenues would not be reduced because of the switch from a DBH tax sharing structure to a PAD opsen structure. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 8 B. The effectiveness of a of a PIT piggybacking option in the future. In addition, this time could be used to prepare for the required administrative PIT piggyback depends on changes to implement the PIT piggybacking in a transparent, complementary and requisite accountable, and cost-effective manner. For example, changes will be needed in the various tax forms, business reforms: processes, and IT systems. A well-designed implementation and support strategy is also required to provide awareness, i. Correct identification of taxpayer residence socialization, and education to taxpayers, tax-withholding Accurately allocating the PIT piggyback revenues to the agents, tax department personnel, and the SNGs. place of taxpayer residence is critical in strengthening the “fiscal contract” between the SNGs and the taxpayers. The remainder of this note is structured as follows. Section From a theoretical perspective, SNGs should levy the 1 provides an overview of the revenue mobilization context PIT piggybacking on their residents, creating a feedback in Indonesia and the vertical fiscal gap for SNGs. Section 2 accountability loop between SNGs and residents.3 explains the typology of SNG revenue sources, defines PIT Unfortunately the taxpayer residence data currently piggybacking, and explains how countries across the world available within DG Tax as part of the tax registration have structured PIT piggybacking. Section 3 discusses the role number (NPWP) is not systematically updated, and of SNG autonomy in PIT piggybacking, brings in international thus, does not accurately reflect the actual residence of experience, and discusses the possible changes envisaged taxpayers. Therefore, granting SNG autonomy to set a by the Director General of Fiscal Balance (DGFB) on the SNG piggybacking rate is less likely to achieve the intended autonomy dimension. Section 4 discusses the revenue and “fiscal contract” accountability and efficiency objectives. equity implications of PIT piggybacking. Section 5 discusses This makes it important to efficiently generate a more the challenge of inaccurate residential addresses in the PIT accurate residence data in the PIT database before database for Indonesia and ways that it can be improved prior introducing PIT piggybacking. to the introduction of PIT piggybacking. Section 6 explains ii. Visibility/salience of the surcharge to taxpayers the reform trajectory of various countries that have introduced To strengthen the “fiscal contract” between the SNG and PIT piggybacking. Finally, Section 7 discusses implications for its residents, it is also important that local residents are Indonesia and the possible way forward. fully aware of the newly introduced PIT piggybacking revenue structure. C. The current lack of accurate taxpayer residence data makes it the wrong time to introduce PIT piggybacking. As discussed above, the availability of accurate residence information of taxpayers is an important pre-condition for implementing PIT piggybacking. Hence, it would be important for DG Tax to use the time now to generate a more accurate and updated information on taxpayer residence in order to deliver an accurate revenue allocation of the current PIT DBH system and to provide the foundation for the possible introduction 3. Residents consume local services, such as education, garbage disposal, and health facilities. Hence, in ideal circumstances, the PIT piggyback should be levied on residents living within the jurisdiction of the SNG. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 9 1. >>> Revenue Mobilization in Indonesia Indonesia collects lower tax revenues as a share of GDP compared to structural peers (Figure 1). The tax gap between actual and potential revenues was estimated at 6 percentage points of GDP in 2018 (Figure 2).4 The widening tax gap in Indonesia shows declining tax effort over time (Figure 3). Corporate income tax (CIT) and value-added tax (VAT) are the largest contributors to tax revenue, together constituting 68% of total tax revenue (Figure 4). Personal income tax (PIT) contributes only 8% of total tax revenues. >>> Figure 1: Low tax collections compared to structural peers Source: IMF Revenue Longitudinal Data, 2018. >>> Figure 3: Decline in Tax Effort Over Time >>> Figure 2: Widening tax gap in Indonesia Source: World Bank staff estimates. >>> Figure 4: CIT and VAT are the biggest sources of revenue CIT 18% Total VAT 50% 15% PIT 5% Excise Property 5% Other Tax 6% 1% Source: World Bank staff estimates. Source: IMF Revenue Longitudinal Data, 2018. Note: CIT = corporate income tax, PIT = personal income tax, VAT = value-added tax. 4. For details, see Box A.4 in Indonesia Economic Prospects. 2021. A Green Horizon: Toward a High Growth and Low Carbon Economy. Washington, DC: World Bank Group. https://openknowledge.worldbank.org/bitstream/handle/10986/36732/166956.pdf. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 10 Although district reliance on the central government’s own-source revenue (OSR/PAD) as a share of total district intergovernmental transfers has declined over the last two revenues. PAD constituted 6.6% of total district revenues in decades, these transfers remain the primary source of district 2001 but increased to 14.8% in 2018, showing a significant revenue (Figure 5). This decrease in central government decline in the vertical fiscal gap. reliance has been driven in part by an increase in district >>> Figure 5: Districts primarily rely on central government transfers for their spending needs 600 100 PAD 500 DAU 80 DAK % total district revenue 400 DBH 60 Rp trillion 300 Other 40 200 20 100 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Source: Indonesia Database for Policy and Economic Research. Note: DAK = Dana Alokasi Khusus (Special Allocation Fund), DAU = Dana Alokasi Umum (General Allocation Fund), DBH = Dana Bagi Hasil (Revenue Sharing Fund), PAD = Pendapatan Asli Daerah (Locally generated revenue), RHS = central government transfers to districts as a share of total district revenue, Rp = Indonesian rupiah. DAU and DAK are different types of central government transfers to the local governments. DBH is a tax sharing arrangement for central government taxes. PAD are local governments’ own-source revenues. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 11 2. >>> Classification of Subnational Government Revenue and PIT Piggybacking The level of influence, or autonomy, that SNGs have over different revenue sources varies largely, depending on the revenue instrument, such as intergovernmental transfers, tax revenues, and non-tax revenues. This is illustrated in Figure 6 which shows different revenue instruments on a continuum of SNG autonomy. When it comes to revenue instruments, tax sharing arrangements like the current structure of the personal income tax and the property tax on forestry, mining, and estates between the central government and SNGs in Indonesia offer little autonomy to SNGs and are classified as a form of revenue transfer. On the other end of the spectrum, taxes and non-tax revenues that have been delegated to SNGs, such as the urban and rural property taxation in Indonesia’s case, are forms of OSRs. In these OSR cases, SNGs have the power and responsibility to set and levy rates and administer/collect taxes and non-tax revenues, within the bounds set by the central government. As Figure 6 indicates, piggybacking taxes are also a form of OSRs as the SNG has the power to levy the piggybacking surcharge tax rate, although the tax base and the tax administration are under the responsibility of the higher level of government. >>> Figure 6: Typology of SNG revenue sources by level of autonomy Revenue Sharing +/- room for maneuver Earmarked General/ Shared Taxes +/- room for maneuver Tax Sharing Piggybacking Own Source Tax Revenues + room for maneuver (Partial-Full power over rates and base) Non-Tax Revenues + room for maneuver (Partial-Full power over rates and base) Unconditional No Discretion Expenditure Discretion Revenue & Expenditure Current and • Block Grants Shares of Local surtaxes Property Taxes capital grants • Equalization national taxes on national Property Transfer Tax User Charges earmarked Grants (CIT, PIT, VAT, taxes (CIT, PIT, Professional Tax Service Fees for education, Excise Tax) VAT, Excise Local PIT Regulatory Fees health, roads, Tax) Vehicle Tax Exactions housing, Entertainment Tax Impact Fees water, etc. Hotel Tax Parking Fees Environmental Tax Transfers: Received by LGs Own Source Revenues: Levied by LGs Source: OECD (Organisation for Economic Co-operation and Development) and United Cities and Local Governments (UCLG). 2019. 2019 Report of the World Observatory on Subnational Government Finance and Investment – Key Findings. Paris: OECD; UCLG: Barcelona. Note: CIT = corporate income tax, LG = local government, PIT = personal income tax, SNG = subnational government, VAT = value- added tax. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 12 Piggyback taxes are local surtaxes levied by the SNG on a On the other hand, states in the US set their own tax rates, common national tax base wherein the SNGs have some determine their own income tax base (although many autonomy to set the tax rate. PIT piggybacking is a local tax determine their PIT tax base adjusted from the federally levied on either the taxable personal income or the personal determined tax base), and administer/collect the state-level income tax liability which is in addition to the tax levied by the income tax themselves with no federal level involvement. It is central government. important to emphasize that the state-level income tax in the US operates as a pure SNG tax (i.e., the state sets its own tax Multiple high-income countries have implemented PIT base and tax rates, and administers/collects its constituents’ piggybacking. If Indonesia were to implement PIT piggybacking, PIT), although it is sometimes referred to as a piggyback it would be one of the first middle-income countries to do so. system since the PIT tax base used by the states is closely Indonesia’s outlier status is worth noting, as shown in Figure linked to the federal income tax base. 7, because high-income countries typically face significantly fewer challenges to PIT administration and lower constraints Local governments (municipalities, cities, and towns) in 17 of to PIT enforcement. the US states also levy a personal income tax (Waznak 2019). Local governments in 12 of those states levy their own rates, The piggyback arrangements in these high-income countries determine their own base, and administer/collect their own have varying levels of autonomy given to SNGs. For instance, income taxes. In these US cases, the PIT is structured as an in Denmark, local governments5 have the power to set the independent tax and cannot be considered a piggyback PIT surcharge rate on the central-government-determined tax system. However, local governments in five of the 17 states base while the central government remains responsible choose their local PIT tax rate but piggyback this income tax for all aspects of tax administration. This is similar to the on the state-level-determined tax base and leave aspects of provincial-level PIT system in Canada (with the exception of administration/collection to the state tax agencies. Quebec Province) where the provinces choose and levy their tax rate on the common federal tax base, while the federal- level Canada Revenue Agency collects both the federal and provincial level PIT.6 >>> Figure 7: 2019 GDP per capita of countries with PIT piggybacking compared with Indonesia 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 - Croati a Sweden Iceland Denmark Belgium Korea, Rep. Portugal Indonesia Montenegro Spain Finland Switzerland United States Norway Canada Italy Source: World Bank data. 5. Local government refers to district-level government. 6. In Canada, provinces and territories that have entered into tax collection agreements with the federal government for collection of personal income taxes (“agreeing provinces,” i.e., all provinces and territories except Quebec) must use the federal definition of “taxable income” as the basis for their taxation. This means that they are not allowed to provide or ignore federal deductions in calculating the income on which provincial tax is based. Quebec is an exception; it defines its own PIT tax base and rates and collect its own constituents’ personal income taxes. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 13 It is important to emphasize the difference between a shared tax (called DBH in Indonesia) and an OSR (called PAD in • Scenario 1 is the current PIT shared tax (DBH) system Indonesia). Shared taxes are essentially intergovernmental wherein the four defining characteristics are defined by transfers where the central government is responsible and the central government, but with SNGs receiving a portion controls the tax base, tax rate, and tax administration while of the PIT shared revenues (currently 20% of the total PIT sharing a portion of the collected revenue with the SNGs. The collected). These shared revenues are allocated between revenue sharing formula is also determined by the central the province and the “income tax producing” district, with government, as in the case of Indonesia (e.g., splitting the a portion shared across all districts within the province. PIT between provinces and the PIT “producing” districts, with a portion also being distributed across all districts within a • Scenario 2 is an option wherein the four defining province). Under the shared tax approach, SNGs only receive characteristics are still defined by the central government, a portion of the PIT revenue as defined by law. SNGs are but with the SNG receiving the portion of the PIT revenues neither responsible nor accountable on how much and how which are collected based on the central-government- the money is collected but are only responsible on how the determined fixed opsen rates. The major difference is that received PIT money is spent through their budget process. the PIT DBH 20% share is replaced by a PIT opsen rate of 25%, set by the central government, for which SNGs In contrast, an SNG PAD is a tax (or non-tax revenue) are required to levy. The other minor difference is that the instrument which is under the control of the SNG, levied PIT revenues will now accrue only to the province and to within any limits prescribed by the central government under the producing districts, with no amount shared across all legislation. These PAD are therefore seen as directly under the districts with the province. This option is essentially the responsibility and accountability of SNGs. While the revenue same as a shared tax because there is no rate discretion base may be defined by the central government, with some given to the SNG; the main difference is a reclassification autonomy and flexibility given to the SNG to grant exemptions from a DBH revenue to a PAD revenue. and deductions, among others, the policy decision on choosing the tax rate itself is given to the SNG, often within a range set • Scenario 3 is a “piggyback” tax, as commonly known, by the central government. wherein the SNG is given the autonomy and responsibility to set the PIT surtax rate within the limits established The PAD/OSR administration is typically under the SNG, by the central government. The PIT surtax could either although the revenue administration may be co-administered be levied on the common national PIT tax base (e.g., with the central government, with other SNGs, and/or with Canadian provinces, except for Quebec). It could also be certain administrative functions outsourced to the private levied on different tax bases (e.g., local governments in sector (e.g., postal system for bill delivery, banking system for Denmark which allow for special deductions on interest receiving tax payments). The key is that the revenues from and commuting expenses) or as a flat percentage of the PAD accrue to the SNG, and the tax rates (levies) are the PIT tax liability (e.g., Croatia). Under a piggybacking determined and levied by the SNG. Under the PAD approach, arrangement, the tax base is typically fixed by the central the SNG does not only receive the revenues from the PAD government, and the tax administration/collection is revenue instrument but also chooses and levies the rate, thus entirely the responsibility of the central government. the SNG becomes accountable for how much is collected and Such a “piggyback” PIT is considered as an OSR since how that money is spent. the SNG is given the responsibility to choose the tax rate and has the option to “levy” that tax rate; thus, the SNG Table 2 illustrates the progression between a DBH to a PAD, becomes accountable and responsive to its citizens in broken down by the power to determine the tax base, the terms of determining the amount of tax revenue collected tax rate, the administration responsibility, and the revenues and spending that tax revenue appropriately through the received. Applying this to the changes being considered in SNG budgeting process. Indonesia, we can see four different scenarios. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 14 • Scenario 4 is a pure OSR wherein the SNG chooses income taxes and some of the local government income the tax rate within the limits established by the central taxes in the US. SNGs under a pure OSR structure government. The SNG also chooses to adjust the central- are responsible for setting the tax rate but, in contrast government-determined tax base within the limits allowed to the “piggybacking” structure, may be able to adjust by the central government, administers and collects the tax base within limits and would be responsible for the taxes (and fees), and receives the tax revenues. administering/collecting the PIT. PIT examples under Column 4 would be the state-level >>> Table 2: Structural differences between shared revenues (DBH) and own-source revenues (PAD) Scenario 1 Scenario 2 Scenario 3 Scenario 4 PIT opsen Piggybacking Tax Key Aspects / Shared Tax (Opsen A: without rate (Opsen B: with rate Own-Source Revenue (PAD) Functions (DBH) discretion) discretion) Central government Tax Base Central Central government Central government With possible SNG discretion on exemptions, Definition / Choice government tax incentives for local development within the limits set by the central government. SNG SNG Tax Rate Central Central government Usually within the Definition / Choice government Usually within the limits set by the central limits set by the central government government SNG Central Tax Administration Central government Central government With possible co-administration with the government central government, other SNGs, and third- party agencies SNG SNG Shared with SNG SNG Based on revenue based on the Based on surtax rate Revenues collected from the percentage set chosen by SNG within the Based on the tax rate chosen by SNG within central-government- by law limits set by the central the limits set by the central government mandated fixed surtax government amount This is a true SNG OSR as it provides SNGs with This is also a true SNG OSR because the discretion/autonomy to SNG has the autonomy to choose the tax set the surtax rate within rate and levy the tax. It helps to strengthen This is essentially the the limits set by the the “fiscal contract” between SNG and same structure as a central government. residents and can improve the efficiency shared tax (DBH); the and accountability to link revenues and only difference is that It helps to strengthen the expenditures through the planning/budgeting Comments the law defines the “fiscal contract” between process. Opsen Type A as a SNG and residents and PAD, which confuses can improve the efficiency The main difference between scenario 3 and the distinction. and accountability 4 is which level of government is responsible to link revenues and for collecting/administering the tax—in expenditures through scenario 3, it is the central government while the planning/budgeting in scenario 4, it is the SNG. system. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 15 3. >>> Autonomy of Subnational Governments and Fiscal Contract Greater autonomy of SNGs in determining their local OSRs and rates can help strengthen the “fiscal contract” which is the fiscal–governance accountability relationship between citizens and SNGs. Bird (2011) argues that residents are more likely to hold local governments accountable if public services are financed by local own-source revenue rather than by transfers. In Brazil, increases in the share of SNG tax revenues was found to lead to a larger increase in local public health and education services than correspondingly large increases in transfers (Gadenne 2016). Similarly, in the Democratic Republic of Congo (DRC), Weigel (2020) finds that active property tax collection led to improvements in citizen’s political participation. Similar examples can also be found in Thailand, Mozambique, and Burundi (Wainer 2019). PIT piggybacking could in theory strengthen the “fiscal contract” under three conditions. First, SNGs need to have the power to set the tax rate, even if this is within the bounds defined by the central government. Second, the piggyback tax revenue levied by the SNG and collected by the central government should be allocated to the taxpayer’s place of residence. Third, citizens need to be informed that the PIT piggyback is an SNG surcharge (a PAD) rather than a central government tax (a DBH) so that they can hold the SNG accountable for the level of the PIT piggybacking surcharge and how those funds are linked to local-level expenditures. Countries that have implemented PIT piggybacking provide SNGs with the power to set the surcharge rate. For instance, in Denmark, Sweden, and Croatia, municipalities have the authority to set the piggyback rate while the central government is responsible for determining the tax base and for all aspects of tax administration. Norway, on the other hand, has experimented with a ceiling rate to address concerns of over-taxation by municipalities, but the result was simply that all municipalities charged the ceiling rate (Lotz 2012). Providing SNGs the power to set the PIT surcharge rates (even if within the bounds set by the central government) essentially makes the PIT surcharge a local own-source revenue. Allowing SNGs to set the PIT surcharge rate increases their accountability and strengthens the “fiscal contract” between the SNG and its residents. Typically, the central government will provide a rate setting framework for the EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 16 SNG to ensure that the combined PIT tax rate of the central any involvement from the federal government.7 In contrast, government and the SNG is set at a reasonable level to meet in Canada, the provincial level PIT tax rates are set by the the broader national revenue goals as well as efficiency and provincial government, but the tax administration is by the equity objectives. federal government, except for Quebec. In some countries, such as Japan and Korea, local Personal income taxes are not levied by local municipal/city governments piggyback their local income tax on the central governments in Canada but are levied in 17 of the 50 states in government tax base, applying their own PIT tax surcharge the United States. Of those 17 states, five use a piggybacking rate but taking on the responsibility for managing the billing structure wherein local governments levy their PIT surcharge and collection-related tax administration aspects. In the on the state tax base and the PIT is administered by the state United States, state-level governments typically levy their tax agencies. In the other 12 states, local governments levy PIT on some variation of the federal PIT income tax base, their own PIT rates and administer the income tax themselves but are fully responsible for all tax administration, without (Walczak 2019). BOX 1: PIT PIGGYBACKING IN DENMARK The Danish income tax has a long and unique history. The income tax was initially established as a local income tax in 1803, based on a local assessment of personal capacity to pay tax. By 1861, the local income tax assessment shifted to rule-based assessments, limited to 3 percent of estimated income. The modern central government income tax was introduced in 1903, 100 years after the local income tax had been operational. When the central government income tax was enacted, there was a decision that the local income taxes should be based on the income as defined by the central government income tax law. Until 1969, local governments had the autonomy to set their own rates and rate structures, including the prerogative to impose progressive tax rates. After 1969, the proportional (flat rate) tax was introduced, replacing the local-level progressive rates. The local personal income tax in Denmark is collected as a “piggyback” on the central government’s PIT system. The local government sets a flat tax rate which is applied to the PIT base assessed for the central government. As Figure B1.1. illustrates, the local flat rate is added to the national-level progressive income tax rates. The marginal administration costs for the local PIT administration are negligible; the main extra requirement is the need to match the PIT collected from the taxpayer residence. >>> Figure B1.1: Income Tax in Denmark, 2007 Source: Lotz, Jorgen, and Jens Blom-Hansen og Søren Hartmann Hede. 2013. The Changing Role of Local Income Taxation in Denmark. 7. The District of Columbia and the 41 states levy their broad-based taxes on individual income. New Hampshire and Tennessee tax only individual income from dividends and interest. Seven states do not tax individual income of any kind. https://www.taxpolicycenter.org/briefing-book/how-do-state-and-local-individual-income-taxes-work. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 17 Under the Danish system, the municipal and central government rates are added and applied to the PIT tax base. To prevent excessive tax increases, Denmark has adopted a national maximum tax ceiling within which the total local government and central government tax rates must fall below. Denmark has a formalized system of negotiations to keep the rates below the maximum rate ceiling (Lotz 2012). In addition to this unique local income tax structure, Denmark also has a complex and effective fiscal equalization transfer system which takes into account the local-level expenditure needs and fiscal capacity. The fiscal capacity measure includes the potential PIT revenue capacity, thus the structure of the PIT rate discretion and the intergovernmental transfer system are closely linked. The Danish PIT piggybacking trajectory can be seen as an additive approach from independent income taxes. That is, the Danish PIT system started first with a local-level PIT in 1803 followed by the central government PIT in 1903. Merging these two separate PIT structures and administration into a single, piggybacking tax enabled Denmark to combine the local government rate setting autonomy with the efficiency of a single tax administration under the central government. Having a formalized system to coordinate the tax rate levels to fit within the maximum country-wide tax rate ceiling, along with a highly effective intergovernmental fiscal equalization transfer system, has contributed to making the Danish system work. Ensuring citizens (taxpayers) understand that PIT piggybacking residential relocation decisions due to the possibility of is a local surcharge will be important in strengthening the “fiscal regional variation in the piggybacking rates set by SNGs. The contract.” This can be done by modifying the PIT tax return so extent of the locational impact depends on the difference in the that the local piggybacking surcharge is shown as a clearly PIT piggybacking rates across SNGs and whether the tax rate separate surtax being levied by the SNG,8 by implementing difference is significantly large enough to influence taxpayer information campaigns that explain the change to citizens, and residential relocation. The possible regional differential will be by outlining a PIT piggybacking rate setting procedure that constrained by the central government policy regarding the involves taking an approval from the local parliament (DPRD), allowable piggybacking surtax amounts. among others. While information campaigns may be costly, small behavior interventions such as changing the tax return format can be effective in bringing up the salient feature of PIT piggybacking as a local surcharge at low cost. Such policy and administration interventions can help citizens understand that the PIT piggyback is a PAD levied under the authority of the SNG and is separate from the central-government-imposed income tax. These interventions would also amplify the tax rate level chosen which would further strengthen the accountability link between the SNG and their residents. In addition to the general distortionary effects of a central government income tax on the labor market, PIT piggybacking can, in theory, have an impact on inter-SNG taxpayer 8. The change in the tax declaration forms and the withholding forms will be necessary to enable the DG Tax to properly account for and allocate the SNG portion of the PIT. This is especially true if SNGs are allowed to have some limited flexibility in setting differential opsen tax rates. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 18 4. >>> Impact of PIT Piggybacking on Subnational Government Revenues and Equity As mentioned earlier, a key distinction between PIT tax sharing (DBH) and piggybacking is that the latter is associated with more SNG revenue autonomy and stronger SNG fiscal accountability.9 This autonomy in rate setting can strengthen the fiscal contract as well as encourage enhanced revenue mobilization and equity (Table 3). >>> Table 3: PIT piggybacking with or without rate setting autonomy Key Questions What are the implications of PIT piggybacking to: the fiscal contract? With Rate Setting Autonomy (within bounds) Strong fiscal contract, if local residents are aware of the implications of shifting to a PAD piggybacking system. This requires piggybacking surcharge revenues to be allocated to taxpayer residence. With the added ability (at the margin) for SNGs to be able to mobilize additional revenues to respond to residents’ demand and willingness to pay for improved services Without Rate Setting Autonomy (fixed by law) Unlikely to strengthen fiscal contract since this arrangement would be very similar to the PIT DBH (Piggybacking surcharge revenues should be allocated to taxpayer residence) Possible increase if (i) PIT piggybacking is Possible increase if (i) PIT introduced as an addition to PIT DBH; or (ii) piggybacking is introduced as if PIT piggybacking replaces PIT DBH such revenue mobilization? addition to PIT DBH; or (ii) PIT that SNGs are allowed to set rates above piggybacking rate is set above 25%, but are required to have a floor rate 25% of PIT liability set at 25% of PIT liability PIT producing districts will PIT producing districts will get more get more piggyback revenues interregional equity? piggyback revenues (details are given in (details are given in Figure 8, Figure 8, Annex 2, and Annex 3) Annex 2, and Annex 3) 9. It should be noted that in many middle-income and lower-income countries, the PIT tax base is quite narrow, often levied primarily only on formal sector or public sector employment income. In addition, the PIT tax base may experience substantial tax base and tax revenue leakages because of tax administration challenges. This narrow tax base can weaken the potential regional autonomy and accountability objectives as well as the achievement of the broader revenue, equity, and efficiency objectives. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 19 We simulated the revenue and equity impact of a potential introduction of an additional district-level PIT piggyback arrangement (Figure 8, Annex 2, and Annex 3). Our results show that a PIT piggyback added to the exiting PIT DBH, with a rate that raises an equivalent amount of revenue as the current district-level PIT DBH, could potentially increase district PAD/OSRs by an average of 8.31%. Hence, an additional PIT piggybacking has the potential to be a significant source of SNG revenues which bolsters the case for its introduction in the future. The simulation shows that PIT piggybacking would accrue to the PIT producing districts that are mostly urban centers.10 However, it is important to mention that even though larger urban districts receive more total revenue from PIT piggybacking, it might not necessarily mean that they will receive more than other district governments on a per capita basis (Annex 3). Any revenue disparities across districts emerging from the potential introduction of PIT piggybacking could be addressed via intergovernmental transfers such as the DAU. Taken together, the simulation maps show that at an absolute level, urban districts and the island of Java-Bali will receive the most revenues from the introduction of a PIT piggybacking. However, there would be a more geographically varied increase in OSR (Annex 2) as well as revenue increase on a per capita basis (Annex 3). Implementing an SNG PIT surcharge system will potentially and relatively benefit SNGs that have a larger PIT tax base. Thus, SNGs that have a larger formal sector and/or public sector employment are better able to take advantage of the PIT surcharge system in substantially generating more SNG revenues. As the potential revenue impact from such a PIT surcharge system varies across SNGs, the central government will need to consider these differential SNG fiscal capacity measures when designing their intergovernmental fiscal transfer systems. 10. Annex 2 provides the simulation results in absolute figures rather than as a share of PAD to illustrate this point. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 20 >>> Figure 8: Simulation of potential district revenue from the introduction of PIT piggybacking as a share of PAD Note: This simulation is based on district-level PIT tax sharing (DBH) data shared by the Ministry of Finance. The model assumes that the PIT piggyback/opsen tax rate is set such that it raises additional district-level revenue equivalent to PIT DBH. This analysis only focuses on the district-level PIT piggyback and excludes districts in DKI Jakarta, because of data challenges. An important caveat to the simulation results is that the district-level PIT DBH aggregates include 8.4% of the PIT generated in the district and an equity component that is added to the DBH. This equity component is 3.6% of total PIT revenues in the country. Since our data does not disaggregate the PIT generation and equity components, the simulation results may have upward bias in districts that receive more of the equity component. Further details about the simulation are given in Annex 1. Simulation results for absolute PIT opsen revenue and on a per capita basis are given in Annex 2 and 3 respectively. 5. >>> Correct Identification of Taxpayer Residence From a theoretical standpoint, PIT tax sharing or piggybacking revenues should be assigned to SNGs based on the residence of taxpayers. This is because citizens consume local services (e.g., education, health, garbage disposal) mostly in the place of their residence, and taxpayers can vote in the local elections where they reside in order to hold their elected leaders accountable for their taxation and expenditure decisions. In the case of PIT DBH, inaccurate identification of taxpayers’ current residence may lead to mismatches in the allocation of resources across SNGs. In the case of PIT piggybacking, incorrect identification of taxpayers’ residence acts as a barrier to strengthening the “fiscal contract.” Hence, the ability of DG Tax to identify the residence of taxpayers more accurately is a crucial pre-condition to introducing PIT piggybacking in Indonesia. PIT in Indonesia allocated as PIT DBH consists of PPh21 and PPh25/29. PPh21 is collected as a withholding tax by the employer which is remitted to the DG Tax regional KPP tax office linked to the place of employment. PPh25/29 is the final tax filed by the taxpayer that is paid to the regional KPP tax office department where the taxpayer is registered. Taxpayers are required to register with DG Tax to obtain a 15-digit tax identification number (NPWP). Figure 9 provides details of what different parts of the NPWP represent. As illustrated, the NPWP has a three-digit code (tax office code) identifying the KPP office where the taxpayer originally registered. This NPWP does not change depending on where the taxpayer lives (residence). It is always based on where the taxpayer originally registered to receive their NPWP. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 22 >>> Figure 9: The structure of the taxpayer identification number (NPWP) Currently, the residence information available with DG Tax and b. Strengthening and formalizing the existing head of linked to the NPWP is often not up-to-date. This is because village reporting mechanism. Informally, citizens who DG Tax does not proactively update the information on where move to a new location should report to the village head the taxpayer maintains his or her residence. Technically the (RT/RW) and are required to submit a copy of their national taxpayer can report a change of address to the tax department. ID to the village head. Submitting a copy of their national ID However, taxpayers often do not update their place of does not necessarily mean that their national ID address residence with DG Tax. Thus, the residential address in the tax is updated to reflect their new residency. There may be an files may not reflect the current address of taxpayer residence. opportunity to formalize this reporting mechanism through The employers when paying the PPh21 do submit information regulation which could provide more accurate taxpayer that identifies the taxpayer NPWP, the amount of wages, and residency information. Given that the PIT DBH revenues the amount of PPh21 withheld. Apparently, employers also do and any future PIT piggybacking revenues will benefit the provide the taxpayer address in their files in the annexes that constituency where the taxpayer’s residential address on are submitted to the tax department. file belongs, local governments should have an incentive to encourage the updating of residence information. To better ensure that taxpayer residential addresses are correct and to enable the introduction of PIT piggybacking DG Tax is in the process of developing the core tax in Indonesia, the possible options that the DG Tax can take administration system (Sistim Inti Administrasi Perpajakan/ include the following: SIAP) which provides an excellent opportunity to put in place regulations and related business processes that can improve a. Integrating DG Tax’s systems with national identity the accuracy of the actual place of taxpayer residence. card database. The residential addresses available as part of the national resident identity card (KTP/NIK) may have more updated information on the actual taxpayer residence. Integrating the two databases may help improve the identification of taxpayer residence in the DG Tax information system. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 23 6. >>> Reform Trajectory There are three possible reform trajectory models to phase in a piggyback PIT system: Option 1: Additive approach from independent income taxes (Denmark and Canada) This is the trajectory that Denmark and Canada followed in its current PIT piggybacking structure. These countries started with independent personal income taxes at the local and central level (Denmark),11 or at the provincial and federal level (Canada).12 SNGs in both countries had and continue to have the power to set their own PIT rates. In both countries, the central government is responsible for the PIT revenue administration. While there are similarities between the two countries, there are some major differences. In Denmark, local governments can set their own PIT flat rate, but there is an agreement that there is a maximum PIT tax rate ceiling (Danish: skattelot) within which the local government and central government income tax rates must fit (Lotz 2012). This has been set at 52.05% since 2019.13 In Canada, in contrast, provincial governments can set their own PIT rates. The trajectory sequence for Denmark and Canada started from a system of independent personal income taxes at both the SNG and central government levels. To improve efficiency and coordination, these two countries centralized the tax administration while maintaining rate setting flexibility with the SNGs. These systems then evolved into their current PIT piggybacking system. Comment: This additive approach is not an appropriate model for Indonesia as the starting points are very different. Indonesia is starting from a position with a shared income tax, with no independent local income taxes. 11. In Denmark, local governments have had local income taxes since 1803, while the modern central government income tax wasn’t introduced until 1903. At that time in 1903, there was a decision that local income taxes should be based on the income as defined by the central government income tax law. Municipal income taxes could be structured with progressive rates until 1969 when the proportional, flat rate structure was imposed on local-level income taxes (Lotz and Hede 2013). 12. In Canada, there is an agreement that the Canada Revenue Agency (CRA) would collect provincial income taxes, but in exchange, the provinces must use the federal tax base. The CRA collects all provincial personal income taxes, except for the Province of Quebec which administers its own PIT system. 13. See the Danish Ministry of Taxation website at https://www.skm.dk/11466.aspx. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 24 Option 2: Additive approach with maintaining PIT its existing PIT tax sharing system and to give its local DBH and adding an SNG PIT piggyback surtax governments the power to levy an additional surtax on the (Croatia) national-level PIT (see Box 2 for more details). This was the approach taken by Croatia in its 1994 decentralization reforms. Croatia’s strategy was to keep BOX 2: PIT PIGGYBACKING IN CROATIA In 1994, Croatia embarked on an ambitious decentralization reform, restructuring the expenditure and revenue responsibilities between the central and subnational governments. At this time, the PIT and the local government surtax/piggyback on PIT liability were introduced (Gnip and Tomić 2010).a The main reason for the introduction of PIT piggybacking/surtax in Croatia was to provide local governments with more fiscal space and autonomy.b Initially, cities with more than 40,000 inhabitants were given the power to set a surtax on the PIT liability levied by the central government at a maximum rate of 30% of their share of the personal income tax (i.e., up to 7.5% of the total income tax, while Zagreb, as the capital city, could levy a PIT surtax on their share of the personal income tax up to 60%, i.e., 27.5% of the total income tax) (Alibegovic 2021). In 2001, the PIT surtax system was expanded to all local governments—municipalities were given the power to levy a maximum rate of 10%; cities with less than 30,000 population, a maximum of 12%; cities with more than 30,000 population, a maximum of 15%; and Zagreb, a maximum of 18%. The surtax is administered by the central government tax department and local governments pay 1 percentage point of the total revenues collected to the tax department for administration (Bronić 2007; Government of Croatia 2021c, Article 5). In 2012, it was reported that 213 municipalities (50% of the total number of municipalities in Croatia) and 75 towns (50% of the total number of towns in Croatia) had introduced the PIT surtax. Of those, only 41 municipalities and 9 towns introduced the maximum legally allowed rate of surtax. Only the City of Dubrovnik with the status of a large municipality introduced the maximum surtax rate.c Thus, only 9 percent of all local governments introduced the maximum legally allowed rate (Alibegovic and Slijepcevic 2012). As of September 2021, 302 local governments are levying the PIT surtax—34 local governments levy from a rate of less than 4%; 158 local governments levy a rate of 5 to 9 percent; 107 local governments levy a rate of 10 to 14 percent; two local governments levy a rate of 15%; and one local government (Zagreb) levies a rate of 18% (Stipić and Grebenar 2021). While local governments rely on the central government for most of their financing needs via tax sharing arrangements and intergovernmental transfers, the PIT surtax serves as the most important source of local government OSR in Croatia. The central government is responsible for levying and collecting its own PIT and collecting the PIT surtax on behalf of the local governments. Taxpayers thus pay the standard PIT rates to the central government and the surtax flat rate which can be levied by the local governments, both through the central government tax administration. It should be noted that the Croatia intergovernmental transfer system includes a system of tax sharing of the central government PIT and real estate transfer tax, and a system for sharing of non-tax revenues (annual concession fees from mineral and spring water and water for public water supply, as well as from the state budget and domestic and international institutions) (Alibegovic 2021). Of note is that the grant transfer formula was changed recently to take into account the local government fiscal capacity, which is measured through the maximum PIT surtax rate, not based on actual local government revenues collected (Alibegovic 2021). The Croatia PIT piggybacking trajectory could be considered an “additive” approach where the local government was given the authority to levy a PIT surtax on top of the PIT being collected by the central government. This additive approach EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 25 increased the total PIT revenue yield and promoted regional autonomy while reducing the risks to the central government’s collected revenues. Croatia allocates the surtax revenue collected on behalf of the local government based on the residence of the taxpayer.14 In Croatia, every citizen is obliged to report change of residence to the Ministry of the Interior. The Law of Residence sets a deadline of 15 days. There is a penalty (500 to 5,000 Croatian kunas for person). The Law of Residence also recognizes temporary residence (if it lasts longer than three months for several reasons, such as seasonal employment, education, or long-term health treatment). Every employee is obliged to report the change of his/her residence address to his/her employer based on the new residence address he/she has reported to the Ministry of the Interior. Employers have the duty to ask their employees for changes in important data, especially those concerning income (salary), even before the issuance of the form on the annual income, paid contributions, income tax, and surtax on income tax. Despite this system, it has been reported that there may be cases of tax evasion wherein taxpayers register their residence in a municipality in order to be taxed at a lower PIT surcharge rate. Therefore, to improve their taxpayer information and tax administration system, Croatia in 2009 initiated reforms to link the Ministry of Public Administration, Ministry of Interior, Tax Administration, and other third-party information using a common PIN (personal identification number) (Stipić and Grebenar 2021). Note: a. For legal documents, see Government of Croatia (2021a; 2021b; and 2021c). b. Prior to 1994, local communities (mjesne zajednice) independently introduced local self-contribution (mjesni samodoprinos) as an additional fiscal instrument via referendum to finance local needs. The central government was responsible for the administration of the local self-contribution, making it similar to a piggyback arrangement. After the introduction of the PIT surtax in 1994, the local self-contribution became a legacy fiscal instrument. c. Zagreb also now levies the maximum rate of 18%. Under this approach, Indonesia could maintain the existing SNGs then receive the incremental revenues resulting from level of the PIT and continue to share the PIT 80:20 with the their piggyback surtax levy which would be collected by the SNGs based on the current arrangements. In this way, the central government tax administration along with the central PIT being paid by taxpayers would remain constant, with the government PIT. amount of PIT revenue being collected remaining constant, and the amounts being shared and allocated across SNGs Under this approach, the PIT taxpayers pay the PIT amount would remain the same. There would be no revenue risk, they used to pay through the current PIT system but are holding everything else constant. required to pay the extra incremental PIT amount levied as a piggyback PIT by the SNG. The total tax administration remains To increase PIT revenues at the margin, while increasing at the central government level to reduce administration and SNG autonomy and accountability, as in Croatia, it is possible compliance costs. to introduce a PIT piggybacking arrangement which would allow SNGs to levy a separate flat PIT surtax on the central This approach safeguards the current levels of the PIT government’s PIT tax liability. This could be set up as a PIT revenues, thereby reducing fiscal risks. This approach could piggyback for the SNGs where the tax base and the tax also simultaneously allow for a small incremental increase in administration remain under the central government authority, PIT tax collections at the national level and could increase but where the SNGs have the power to choose and levy a the “fiscal contract” accountability between the taxpayers and tax rate within the limits set by the central government. The their SNGs. For instance, assuming SNGs were given the 14. There is one exception which is income from rentals; leases; renting of flats, rooms, and beds to travelers and tourists; and property rights. For these income sources, the PIT surcharge is allocated to the location of the real estate or accommodation unit (Stipić and Grebenar 2021). EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 26 authority (option) to levy a PIT surtax between 1 and 4 percent of development (Figure 10). The additive approach (i.e., on the tax liability established under the central government introduction of PIT piggybacking in addition to the PIT DBH) PIT, and if a taxpayer was paying a 20% income tax rate, then would be equivalent to a PIT rate increase for individuals that taxpayer would be required to pay a tax rate of 20.8%, if who are in the PIT net. Since improving PIT collections is an SNG levied a flat 4% surtax.15 an important tax reform for Indonesia, a natural question is whether a rate increase would discourage PIT base Comment: This additive approach would be the best option broadening. While this could be true, the tax rate appears not as Indonesia is also starting from a shared PIT system. to be the main binding constraint to PIT reform in Indonesia. Having a separate SNG PIT surtax option would improve the The main constraints instead are administrative, thus, the transparency and the fiscal accountability linkages between Government of Indonesia could introduce the additional PIT the SNGs and their taxpayers. piggyback on top of the current PIT system while investing in strengthening the overall PIT tax administrative capacity. It is important to note that Indonesia collects significantly lower PIT when compared to countries at a similar level >>> Figure 10: PIT revenues in Indonesia compared with those of other countries Source: IMF Revenue Longitudinal Data and Economic Outlook Data. 15. A taxpayer making Rp 100,000 a year at a PIT tax rate of 20% would pay Rp 20,000 a year. If an SNG levied a maximum rate of 4% on the tax liability of Rp 20,000, the taxpayer would pay an extra Rp 800 to the SNG. This means that the taxpayer would pay Rp 20,800 in total taxes on an income of Rp 100,000, for an average tax rate of 20.8%. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 27 Option 3: Substitution approach (Substituting PIT to have some PIT rate discretion within the limits set by the DBH of 20% with an SNG surtax (with no local rate central government. option) Comment: The proposed approach reclassifying the PIT tax This was an option under review in Indonesia where the sharing (DBH) to a PIT opsen (PAD) in an overall revenue central government simply replaces the current PIT tax sharing neutral manner will neither improve fiscal space nor achieve of 20% with a fixed, uniform 25% SNG surtax on the central the objectives of strengthening regional autonomy and government PIT tax base. The central government continues encouraging SNGs to be more responsive and accountable to collect its 80% of the PIT, while the SNG is given the to their residents (the fiscal contract). Increasing real SNG authority to levy a fixed 25% rate on the PIT which would be revenue autonomy requires that SNGs be given some control collected by the central government. Thus, the total amount of over the rate (even if within the constraints set by the central PIT revenue collected remains the same, since all SNGs are government). The proposed approach does limit fiscal risk required to levy the same 25% rate. through preserving the overall fiscal space from the PIT tax instrument, but the additive approach (option 2) also will The main difference under this approach is that the PIT preserve the overall fiscal space while strengthening regional revenues collected by the central government on behalf of government autonomy and accountability. Thus, option 2 the SNG are reclassified as OSRs (PAD). In addition, there appears to be better than option 3. is a distributional impact because the SNG PIT piggybacking revenues only accrue either to the income-producing If there is a need to phase the SNG piggybacking PIT, then provinces or to the income-producing districts. Under this it will be important to ensure that the reclassification of PIT proposal, unlike under the PIT DBH system, the surrounding revenue from a DBH to a PAD is followed. This can be done districts in the province would not receive any of the PIT by introducing the option for SNGs to have discretion in setting revenues collected. the PIT SNG surtax rate to achieve the intended objectives in the long run. Unfortunately, as shown in Table 2, this substitution approach does not provide any SNG rate discretion, and thus essentially operates in the same way as a shared tax (DBH) rather than as an own-source revenue (PAD). The main difference is that the PIT tax revenue is legally reclassified from a shared tax revenue (DBH) to an own-source revenue (PAD). As proposed, this would not be classified as a piggyback tax in the conventional sense and would not achieve the important policy objective of granting more autonomy and accountability to the SNGs. If it has been determined that it will not be possible to immediately introduce a genuine piggyback PIT structure, then a second-best solution may be to break the reform into two phases. Phase 1 could be an interim stage wherein the PIT revenue received through PIT tax sharing (DBH) is reclassified as a PIT opsen (PAD). Once the reclassification has been legal and accepted, then the government could move toward introducing a genuine PIT piggyback which would provide SNGs with the autonomy and accountability EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 28 7. >>> Conclusion: Implications for Indonesia Our analysis shows that the Government of Indonesia should consider introducing PIT piggybacking as an additional SNG revenue instrument in the medium to long term. Such an introduction could: a. provide SNGs with additional OSRs and help bridge the vertical fiscal gap (our simulation results show that PIT piggybacking could potentially increase district OSRs by an average of 8.31%); b. strengthen SNG accountability to their taxpayers (fiscal contract); and c. provide the Indonesian government with increased flexibility to address economic downturns in the future, like the COVID-19 pandemic, since a progressive PIT system could help strengthen automatic stabilizers and countercyclicality of fiscal policy. However, PIT piggybacking should only be introduced after the following pre-conditions have been achieved: a. The PIT piggybacking arrangement should be introduced such that SNGs would have some flexibility in setting their PIT piggybacking tax rates. Lack of any SNG discretionary rate setting power would practically mean that the PIT piggyback is not an OSR (PAD) but instead just a variation of the current PIT DBH. Giving SNGs some discretionary power to set the rate is an important pre-condition to strengthening autonomy and accountability (fiscal contract). b. DG Tax should have the ability to identify taxpayer residence more accurately. The lack of ability to accurately identify taxpayer residence weakens accountability and efficiency relationship between citizens and SNGs. c. There should be increased citizen awareness on PIT surcharge as an SNG revenue instrument (PAD) rather than as a central government tax (DBH). EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 29 Based on this, we recommend that the Government of Indonesia invest in these administrative pre-reforms before considering possible introduction of a PIT piggybacking. Specifically, the government could do the following over the next few years to prepare for the possible introduction of PIT piggybacking: a. Improve the accuracy of taxpayer residential information within the DG Tax’s PIT tax administration systems by integrating it with the national identity card database and/ or by improving the current regulations and business processes and their implementation so that the taxpayer residence identification can be updated more accurately. b. Prepare the administrative systems, forms, and procedures for filing the PIT to make it clear that there is an additional incremental PIT surcharge being levied by the SNG. This will help taxpayers understand that this PIT piggybacking amount is an SNG-levied tax and that the SNG is accountable for this incremental amount. This change in form is also needed to ensure that the DG Tax will be able to allocate the SNG-levied PIT surtax to the right SNG. c. Along with the administrative changes in the tax forms, systems, and procedures, prepare a communication outreach strategy to raise awareness among citizens that the PIT piggyback is an SNG tax, to improve accountability between taxpayers and their SNGs. The Government of Indonesia has the option to introduce PIT piggybacking either as an additional separate SNG revenue instrument (similar to Croatia) or to replace the PIT DBH with piggybacking. While both options are feasible, the additive approach used by Croatia (i.e., introduction of the SNG PIT surtax in addition to tax sharing) would be a better option because overall, it is revenue positive, simpler, and more transparent.16 16. Replacing the PIT DBH with piggybacking can also be revenue positive if the tax rates set by SNGs are higher than the current PIT DBH. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 30 >>> Bibliograpy Alibegovi, Dubravka Jurlina, and Sunčana Slijepčević. 2012. “Decentralization in Croatia: Problems and Possible Solutions.” EUGOV Working Paper 33: 35. Alibegovi, Dubravka Jurlina. 2021. “Overview of Surtax on Personal Income Tax in Croatia.” Paper presented at the Government of Indonesia – World Bank Webinar on Surtax on Personal Income Tax in Croatia, October 18. Bird, Richard M. 2011. “Subnational Taxation in Developing Countries: A Review of the Literature.” Journal of International Commerce, Economics and Policy 02 (01): 139–61. Bronić, Mihaela. 2007. “Personal Income Tax and Surtax Sharing in Croatia.” Newsletter: An occasional publication of the Institute of Public Finance 9 (27): 1–10. Danish Ministry of Finance. Skråt skatteloft - en historisk oversigt (accessed February 28, 2021), https://www.skm.dk/11466.aspx. Gadenne, Lucie. 2016. “Tax Me, But Spend Wisely? Sources of Public Finance and Government Accountability.” Economic Research Papers. University of Warwick, Department of Economics. https://ideas.repec.org/p/ags/uwarer/269315.html. Government of Croatia. Income Tax Act (accessed February 26, 2021a), https://www.porezna-uprava.hr/en_propisi/_ layouts/15/in2.vuk2019.sp.propisi.intranet/propisi.aspx#id=pro135. ———. Local Tax Act (accessed February 26, 2021b), https://www.porezna-uprava.hr/hr_propisi/_layouts/15/in2.vuk2019. sp.propisi.intranet/propisi.aspx#id=pro1621. ———. Act on Financing of the Units of Local and Regional Self-Government (accessed February 26, 2021c), https://www. porezna-uprava.hr/hr_propisi/_layouts/15/in2.vuk2019.sp.propisi.intranet/propisi.aspx#id=pro1974. Grdovic Gnip, Ana, and Iva Tomić. 2010. “How Hard Does The Tax Bite Hurt? Croatian vs. European Worker.” Financial Theory and Practice 34: 109–142. https://www.researchgate.net/publication/227448251_How_hard_does_the_tax_bite_hurt_ Croatian_vs_European_worker. Lotz, Jorgen, and Jens Blom-Hansen og Søren Hartmann Hede. 2013. The Changing Role of Local Income Taxation in Denmark. https://english.sim.dk/media/13658/jorgen-lotz-jens-blom-hansen-and-soeren-hartmann-hede.pdf. Lotz, Jorgen. 2012. “You Get What You Pay For: How Nordic Cities Are Financed.” IMFG Papers. University of Toronto, Institute on Municipal Finance and Governance. https://ideas.repec.org/p/mfg/wpaper/07.html. Stipić, Bernardica, and Anita Grebenar. 2021. “Surtax on Personal Income Tax in Croatia.” Paper presented at the Government of Indonesia – World Bank Webinar on Surtax on Personal Income Tax in Croatia, October 18. Wainer, Andrew. 2019. Taxation with Representation: Citizens as Drivers of Accountable Tax Policy. Save the Children. https:// www.savethechildren.org/content/dam/usa/reports/advocacy/taxation-with-representation-2019.pdf. Walczak, Jared. 2019. Local Income Taxes in 2019. Tax Foundation, Fiscal Fact No. 667. https://taxfoundation.org/local- income-taxes-2019/. Weigel, Jonathan. 2020. “The Participation Dividend of Taxation: How Citizens in Congo Engage More with the State When It Tries to Tax Them.” The Quarterly Journal of Economics 135 (4): 1849–1903. https://academic.oup.com/qje/article/135/4/1849/5851770. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 31 >>> Annex 1: Methodology for PIT Piggybacking Revenue Simulation This annex provides further details about the methodology for estimating the district-level revenue from the introduction of the PIT piggybacking in Indonesia. In conducting the analysis, the simulation used aggregates of the PIT DBH allocation by district for 2019, provided by the Ministry of Finance. Further details about the simulation are given below. Assumption The PIT piggyback rates are set by districts such that it raises the same level of revenue as the district-level PIT DBH. Implication of assumption for simulation This implies that every district picks a PIT piggyback rate of 15% on the district-level PIT liability and the behavioral response for the tax rate increase is such that the total PIT piggyback revenues by district are equal to 0.8*0.15*(district- level PIT aggregate), where 0.8 represents the behavioral response. This behavioral assumption is important because it represents how much additional PIT revenues the Indonesian government can generate from the introduction of the piggyback. It does this by scaling down the revenue collections from a rate increase to incorporate changes in individual behavior while responding to a rate increase. Limitations • This analysis only focuses on the district-level PIT piggyback and excludes districts in DKI Jakarta. This is because the data does not allow us to disaggregate the district and provincial component of the PIT DBH for districts within DKI Jakarta. • The district-level PIT DBH aggregates represent the actual PIT transferred by the central government to the districts. These aggregates are an imperfect proxy for the total PIT revenues generated by district. This is because the PIT is shared such that 80% of collections goes to the central government, with the remaining 20% going to the SNG level—specifically 8% to the revenue-generating province, 8.4% to the revenue-generating district, and 3.6% given to remaining districts within the province based on equal shares. Since the data only provides the total district-level PIT DBH, this does not allow us to disaggregate the revenue generating and equity component of the PIT DBH at the district level. Hence, we use the district-level PIT DBH aggregates as a proxy for PIT generation by districts, which has the abovementioned limitation. • The actual revenue collected from the PIT piggyback will depend on the final rate chosen by the districts and the administrative capacity of DG Tax for those districts. The simulation assumes that every district picks the same opsen rate of 15% to be levied on the PIT liability and that the behavioral response of 0.8 applies to all districts equally. It does not account for variation in administrative capacity or behavioral responses by district. EQUITABLE GROWTH, FINANCE & INSTITUTIONS NOTE <<< 32 >>> Annex 2: PIT Piggybacking Revenue Simulation Results (Total Revenue) Note: This simulation is based on district-level PIT tax sharing (DBH) data shared by the Ministry of Finance. The model assumes that the PIT piggyback/opsen tax rate is set such that it raises additional district-level revenue equivalent to PIT DBH. This analysis only focuses on the district-level PIT piggyback and excludes districts in DKI Jakarta. An important caveat to the simulation results is that the district-level PIT DBH aggregates include 8.4% of the PIT generated in the district and an equity component that is added to the DBH. This equity component is 3.6% of total PIT revenues in the country. Since our data does not disaggregate the PIT generation and equity components, the simulation results may have upward bias that increases in districts that receive more of the equity component. Further details about the simulation methodology are given in Annex 1. >>> Annex 3: PIT Piggybacking Revenue Simulation Results (Per Capita) Note: This simulation is based on district-level PIT tax sharing (DBH) data shared by the Ministry of Finance. The model assumes that the PIT piggyback/opsen tax rate is set such that it raises additional district-level revenue equivalent to PIT DBH. This analysis only focuses on the district-level PIT piggyback and excludes districts in DKI Jakarta because of data challenges. An important caveat to the simulation results is that the district-level PIT DBH aggregates include 8.4% of the PIT generated in the district and an equity component that is added to the DBH. This equity component is 3.6% of the total PIT revenues in the country. Since our data does not disaggregate the PIT generation and equity components, the simulation results may have upward bias that increases in districts that receive more of the equity component. Further details about the simulation methodology are given in Annex 1.