Social Protection System Review of Indonesia

Chapter 4. Financing social protection

Abstract

Indonesia’s spending on social protection is low for a country at its income level but it has risen significantly in recent years and is emerging as a budgetary priority. This chapter locates social protection spending within broader Government of Indonesia expenditure and identifies how it fits into the intergovernmental budgetary system. It analyses spending on key social protection programmes as well as the spending dynamics of various programmes. It assesses the potential to scale up social protection, particularly the scope for higher levels of tax financing, and concludes by examining the fiscal incidence of taxes and transfers to understand their combined impact on poverty and inequality.

Overall public spending as a percentage of gross domestic product (GDP) is low for a country at Indonesia’s income level as a result of weak domestic resource mobilisation. Social protection currently accounts for a small proportion of public expenditure and is just one of several priorities confronting the Government of Indonesia (GoI). However, social protection spending has risen in real terms since 2000, has increased as a proportion of public expenditure in recent years and will be an important driver of overall public spending in 2019.

Given current weaknesses in the tax system, higher coverage by contributory social insurance arrangements will be an important mechanism for achieving a step-change in social protection spending and coverage over time. However, concerns over the sustainability of these arrangements exist.

Indonesia’s extensive decentralisation complicates the financing of social protection but also offers potential for higher spending if subnational revenues are increased. Social protection is not one of the core functions of the central government, which means that subnational administrations (and the inter-governmental budgetary system) have an influence on overall spending levels.

Changes to the structure of spending offer space for social protection to scale up

Total government spending (including expenditure by central government and transfers to local government) declined from 20.0% of GDP in the early 2000s to 15.0% of GDP in 2010 and stood at 14.6% of GDP in 2016 (Figure 4.1). This level of spending is very low for a country at Indonesia’s income level (IMF, 2017[1]) and limits the GoI’s potential to invest in the development of physical and human capital required to escape the middle-income trap. It also constrains the capacity for redistribution through the fiscal system, which is the most direct means of reducing income poverty and inequality.

Another notable feature of Indonesia’s fiscal system is the high degree of decentralisation. Under big bang decentralisation reforms of 2001 and 2005, Indonesia devolved substantial funds and authority to local governments, including responsibility for public service delivery and natural resource management. The intention was to improve these functions by empowering local administrations to reflect the country’s extremely diverse contexts. However, the reforms did not have the desired effect, resulting in the persistence of sizeable gaps in socio-economic outcomes between provinces and regions. Low capacity levels in sub-national government have been a critical constraint (OECD, 2016[2]).

Box 4.1. Deepening decentralisation in Indonesia’s intergovernmental system

The GoI structure comprises five levels of government: central government, provinces, districts and municipalities, sub-districts and villages. The end of the Suharto era was followed by a major push for decentralisation: Laws No. 22/1999 and No. 25/1999 significantly increased the political authority and resources of the 491 districts and municipalities. As a result, sub-national administrations have significant spending power and great discretion on how public revenues are spent, even though their revenue-raising is not extensive and there are concerns around the financial management capabilities of local government (Nasution, 2016[3]).

Central government spending accounts for the largest share of total expenditure but spending at this level declined from 15.8% of GDP in 2001 to 9.6% of GDP in 2017. Since 2007, the local government share of total expenditure has increased steadily, indicating deepening decentralisation of government functions. Between 2008 and 2009, a period during which central government spending declined by 3.8 percentage points (in GDP terms), regional expenditure only decreased by 0.9 percentage points. Between 2014 and 2016, district governments’ share of total expenditure jumped from 31.1% to 38.1% and was equal to 5.6% of GDP in 2017.

Central government is solely responsible for six absolute (core) functions: finance, foreign affairs, defence, security, religion, and state administration and justice. Central government shares responsibility for provinces and districts for other areas of spending known as concurrent functions. These include the provision of basic services – education, public health and social welfare, government administration, and infrastructure and public works – known as mandatory functions. Elective functions cover the economic sector and include transport, agriculture, industry and trade, capital investment, land, co-operatives, labour force and environment.

Social protection is thus classified as a mandatory, concurrent function of government. District governors and mayors are thus implicated in the implementation of national social protection programmes but are also empowered to develop their own programmes. This can create confusion and lead to duplication of efforts. It is thought to be a factor behind the inefficient implementation of national programmes, despite the existence of minimum standards in place for service provision (UNDP/UNCDF, 2013[4]).

Sub-national government receives large-scale transfers from the central government to implement concurrent functions. The largest of these is the Dana Alokasi Umum (DAU; General Allocation Fund), which covers local civil servant salaries. Although it still accounts for 50% of transfers, the DAU’s significance has been diminishing slightly of late. The second-largest intergovernmental transfer is the Dana Alokasi Khusus (DAK; Special Allocations Fund), which has grown rapidly in recent years. DAK usually targets remote and less developed areas to facilitate capital financing for selected local governments. The grant is channelled into basic education, preventive health care, basic infrastructure and road development, district markets and small-scale industry development, as well as development of regional art and culture.

Historically the second-largest (now the third-largest) source of transfers to sub-national government is the Dana Bagi Hasil (DBH) Revenue Sharing Grant. The DBH re-allocates revenues from general tax and the exploitation of natural resources, including mining, oil and gas.

Provinces raised 37% of their revenues through taxes in 2017, versus around 6% raised by local governments (districts and municipalities). The Organisation for Economic Co-operation and Development (OECD) Economic Survey 2018 finds that enhanced revenue-generation at a sub-national level, primarily through higher revenues from recurrent taxes on immovable property, would strengthen decentralisation, promoting local responsibility and accountability by better matching spending and tax (OECD, 2018[5]).

Figure 4.1. Government spending has declined as a proportion of GDP

Note: IDR = Indonesian rupiah.

Source: MoF (2016).

Figure 4.2 shows how the functional classification of central government spending has evolved since 2005. The GoI disaggregates spending into 12 functions: public services, defence, order and security, economy, environmental protection, housing and public facilities, health, tourism and culture, religion, education, social protection, and others (miscellaneous category). Growth in the social protection function group – from 0.2% of GDP in 2015 to 1.2% of GDP in 2017 – is partly due to changes in its composition in 2016.

Figure 4.2. The composition of public spending is changing

Source: MoF (2017).

However, the “social protection” function group covers a broader range of programmes than are recognised as social protection by this review. Its composition recently changed to include food/housing subsidies and social contributions, which were previously classified under the “general government administration” function.

The public services function group is the largest in spending terms, although its allocation shrunk considerably between 2017 and 2018. Allocations to economic services, the second-largest function group, increased from 0.8% of GDP in 2005 to 2.4% of GDP in 2017. As a share of total expenditure, economic spending jumped from 6.5% to 25.4% of total spending over the same period.

The GoI is constitutionally required to allocate at least 20% of its budget (excluding interest payments) to education, which it has achieved since 2009 (Jasmina, 2016[6]). Two-thirds of education spending occurs at the sub-national level and is thus not captured in Figure 4.2. In total, Indonesia allocated resources equivalent to 3.3% of GDP to education in 2016 (World Bank, 2017[7]). The expansion of basic education to include senior high school will apply further upward pressure on this allocation.

Public health spending was significantly lower than education spending at 1.4% of GDP in 2017. Total health spending was 3.0% of GDP in the same year, meaning private health expenditure was greater than public.

Figure 4.3 disaggregates central government expenditure by economic classification. Between 2001 and 2016, overall spending declined in line with declining central government spending, and the composition of expenditure changed significantly. In 2016, the largest areas of spending were salaries, goods and services, and capital expenditure; in 2001, the main categories had been interest payments, subsidies and development expenditure.

Figure 4.3. Fiscal space has opened up as interest payments and subsidies decline

Source: MoF (2016).

The decline in interest payments and subsidies has been an important source of fiscal space. State subsidies shrank from 4.7% of GDP in 2001 to 1.4% of GDP in 2016 (see Box 4.2). Interest payments declined from 30% to 15% of spending over the same period. However, both subsidy spending and interest payments increased in 2018, with subsidy spending expected to increase further in 2019. Central government spending is budgeted to increase by 12.4% in nominal terms in 2019 from 2018, driven by higher spending on energy subsidies (up 69% in nominal terms from 2018), higher PKH benefits and higher expenditure on PBI JKN, driven both by increased premiums and higher coverage (World Bank, 2018[8]).

Box 4.2. Spending on subsidies has declined significantly but rebounded in 2018

Indonesia’s main items of tax expenditure are energy and non-energy subsidies. The former are subsidies for fuel and electricity. The latter comprise nine categories: cooking oil, food, fertiliser, seeds, soybeans, public service obligations, credit programmes, tax subsidies and miscellaneous.

Spending on energy subsidies peaked at 4.1% of GDP in 2008 but declined dramatically after 2014 (Figure 4.4). This decline reflected a strategy to reallocate resources from energy subsidies to other areas, such as health, infrastructure and social assistance. Until 2014, fuel subsidies were the largest component of spending on energy subsidies; thereafter, they declined to a similar level as electricity subsidies until 2017.

Spending on subsidies jumped by an estimated 0.3% of GDP in 2018 due to an increase in the costs of the diesel subsidy (OECD, 2018[9]) and is set to increase further in 2019 but will remain below pre-2015 levels (World Bank, 2018[8]). In 2019, fuel subsidies and social assistance spending will be important drivers of real growth in public spending.

Figure 4.4. Energy subsidies declined dramatically after 2014

Source: Government of Indonesia Audit Board (2019).

Spending on key social assistance programmes is growing strongly

Social protection spending reflects the fragmented nature of its implementation – by level of government, source of financing and institutional responsibility. This makes it very difficult to calculate an aggregate figure for social protection spending, although programme-level data make it possible to identify trends in total expenditure.

According to the World Bank public expenditure review (2017[10]), total social protection spending in 2016 was IDR 177 trillion, equivalent to 1.4% of GDP and 15.4% of total government spending. Of this amount, IDR 78.3 trillion was spent on social assistance and IDR 99.6 trillion on social insurance, or 44% and 56% of total social protection spending respectively.

Three important trends stand out. First, that social protection spending is increasingly a budgetary priority: in 2012, social protection spending accounted for 10.7% of total public spending and equated to 1.2% of GDP. Secondly, the disparity between social insurance and social assistance is diminishing: in 2012, social assistance accounted for 36% of social protection spending. Lastly, Indonesia spends significantly less on social assistance than other countries at its income level, on average (1.5% of GDP).

Since 2005, Indonesia has invested significantly in social assistance programmes, helped by the phasing-out of fuel subsidies. In 2010, poverty reduction became integral to the administration of President Yudhoyono, which sought to redesign social assistance programmes to achieve broad-based economic growth and fiscal sustainability. Spending on social assistance has since climbed, reaching 0.7% of GDP in 2015. Although this is lower than social assistance spending in a number of peer economies, the current upward trend is expected to continue, thanks to the reform and expansion of key programmes.

Figure 4.5. Social assistance spending increased as fuel subsidies declined

Note: HH SA = household social assistance. BLSM = Bantuan Langsung Sementara Masyarakat

Source: MoF (2017).

The administration that took office in 2015 identified social assistance, in particular the PKH cash transfer, as a mechanism for reducing inequalities of income and opportunity. In the same year, social protection surpassed spending on fuel subsidies for the first time thanks to a major reform in that area (Figure 4.5), although subsidy spending rebounded strongly in 2018. This section examines the allocation of social assistance spending to the programmes identified in Chapters 2 and 3.

Penerima Bantuan Iuran (PBI)

PBI, known as Jamkesmas until 2014, are non-contributory members of the Jaminan Kesehatan Nasional (JKN, national health insurance). The number of beneficiaries increased from 76 million in 2013 to 92 million in 2018. This increase in coverage has been accompanied by strong growth in expenditure (Figure 4.6).

Figure 4.6. Premium subsidies for PBI beneficiaries is rising sharply as coverage grows

Source: MoF (2019).

Total programme expenditure in 2017 was IDR 25.4 trillion, equating to 2.1% of government expenditure or 0.19% of GDP. A recent study shows that programme costs are expected to increase to up to 4.5% of total government expenditure by 2030 as the GoI advances towards universal health coverage (UHC) (Dartanto, 2017[11]).

Enrolment of the poor and near poor, combined with improving health services, represents a significant fiscal challenge. BPJS Health aims to overcome this challenge by incentivising non-poor informal sector workers to join the contributory scheme. To ensure the system’s sustainability, this strategy must be accompanied by improved cost control, higher insurance collection rates and activities to promote public health.

Rastra and Bantuan Pangan Non Tuna (BPNT)

Rastra is the second-largest social assistance programme in terms of coverage. In 2016, Rastra expenditure amounted to IDR 22.1 trillion (Figure 4.7), equal to 1.9% of total government expenditure and 0.18% of GDP. A significant increase in expenditure from 2014 to 2015 has been partly attributed to inclusion errors; Rastra’s inefficient financial structures are well-known (Shin et al., 2017[12]).

Figure 4.7. Rastra spending is on the rise but beneficiary numbers are steady

Source: Compiled from MoF (2016) and World Bank (2012) data.

In 2017, the GoI introduced the BPNT (non-cash food assistance) transfer programme for poor households, which is based on e-vouchers and intends to reduce the leakage associated with Rastra. The Ministry of Social Affairs (MoSA) administers the programme, which is intended to gradually replace Rastra. In 2017, approximately 1.4 million households in 44 cities benefitted from BPNT, which had a budget of IDR 1.6 trillion, equal to 0.13% of total government expenditure or 0.01% of GDP.

Program Indonesia Pintar (PIP)

Improving education outcomes is an important objective of social assistance in Indonesia. Multiple programmes are implemented, costing roughly IDR 14.4 trillion and account for 1.2% of total government spending or 0.1% of GDP. Programme Indonesia Pintar (PIP), administered by the Ministry of Education and Culture, is the largest, covering roughly 20 million students. Spending and coverage have grown rapidly since 2000. (Figure 4.8).

Figure 4.8. PIP spending and coverage are on the rise

Source: MoF (2019).

Program Keluarga Harapan (PKH)

PKH, a conditional cash transfer programme and is emerging as Indonesia’s flagship social protection programme. As modelling in Chapter 3 demonstrates, it is the best targeted and most effective social assistance programme. Piloted in 2007, PKH expenditure has increased significantly since 2012 in line with growth in coverage. Following a recent national scale-up, PKH covered some 6 million households in 2017 at a cost of IDR 12.8 trillion, equal to 1.1% of total government expenditure or 0.10% of GDP (Figure 4.9).

Figure 4.9. PKH spending has increased rapidly

Source: MoF (2019).

Expenditure on PKH increased further in 2018, when the GoI expanded programme coverage to 10 million households. In 2019, a doubling of benefit levels means it will be one of the main drivers of growth in overall public spending. The World Bank has provided a loan to support the development of information systems required to sustain this scale-up (Box 4.3).

Box 4.3. PKH loan recognises the investment case for social protection

In 2017, the World Bank approved a USD 200 million [United States dollar] loan to help develop the architecture for PKH. The loan programme will last until 2021 and intends to support the increase in PKH coverage through strengthening information systems and delivery mechanisms and improving co-ordination with other social protection programmes (World Bank, 2017[13]).

Although this amount represents a small proportion of the GoI’s intended spending on the programme over the coming years, it is nonetheless notable that Indonesia is borrowing on commercial terms to finance social assistance. This reflects an understanding that social protection constitutes an investment in human capital and thus the country’s long-term economic prospects. The loan also demonstrates a mechanism by which donors can support social protection in middle-income countries that no longer access concessional financing.

Bantuan Langsung Sementara Masyarakat (BLSM)

BLSM is a time-bound unconditional cash transfer with one objective: to provide cash assistance to poor households to offset anticipated price inflation related to a specific policy. In effect, its purpose is to reduce the exposure of poor households to economic risks arising from declines in fuel subsidies.

Figure 4.10. BLSM is a time-bound but significant intervention

Source: MoF (2016).

The GoI has intervened through BLSM on several occasions. In 2005-06, for instance, subsidy cuts resulted in household fuel price inflation of over 125%. A similar price-shock occurred in 2008-09. On both occasions, the government disbursed BLSM to prevent increases in poverty. In 2015, an estimated 15.8 million households received IDR 600 000 over two phases, at a cost of IDR 9.5 trillion, equal to 0.9% of total government expenditure or 0.1% of GDP (Figure 4.10).

JKN’s success in reaching the missing middle is threatening its sustainability

In 2015, some 70 million non-contributory and contributory JKN members claimed a total of IDR 67.75 trillion, equal to 0.55% of GDP. Almost half of this expenditure (IDR 30.42 trillion) is financed by contributory members, such as formal sector employees, self-registered members (informal workers) and employers. The GoI pays the balance.

There are several concerns with the financial stability of the JKN system. First, the number of beneficiaries exceeds the number of monthly contributors, resulting in a claims ratio of above 100%. This partly relates to selection issues: beneficiaries only need to have been registered for 30 days or more to receive treatment, creating the potential for individuals to register only when they or their households need medical assistance.

Informal workers also make irregular contributions due to the nature of their income. Their average contribution level is also low for the same reason, and most informal workers opt for the lowest monthly premiums.

At the same time, health care availability and quality affect the decision to enrol. In many areas, services are considered suboptimal, causing people to opt out or resist paying higher premiums. Increased use of facilities by newly covered individuals can also strain quality of services.

The GoI aspires to provide UHC by 2019. Even without UHC, BPJS Health’s accumulated debt is expected to reach IDR 173 trillion in 2019, unless the existing payment scheme is changed (Table 4.1).

Table 4.1. JKN’s deficit is widening

	2014	2015	2016
Beneficiaries (million)	152	162	186
Deficit (IDR trillion)	1.54	4.00	9.79

Source: Dartanto, T. (2017[11]), Universal Health Coverage in Indonesia: Informality, Fiscal Risks and Fiscal Space for Financing UHC, International Monetary Fund (IMF)/Japan International Cooperation Agency (JICA), Tokyo.

Expenditure by labour-related social insurance programmes is lower than spending on JKN and coverage is increasing more slowly. However, the design of the system established in 2015 and overseen by BPJS Labour will have significant implications for the financing and sustainability of social insurance over the long term as the population starts to age.

At present, the majority of labour-related social insurance spending relates to Jaminan Hari Tua (JHT, old age security). Old-Age Security is a defined-contribution provident fund, with employer and employee contributions. Male and female beneficiaries receive a lump-sum pension payment upon reaching the retirement age of 55. JHT expenditure was IDR 16.75 trillion in 2015, equal to 1.42% of total government spending or 0.15% of GDP.

Other insurance programmes include the survivor pension, JKK and a few insurance schemes targeted at specific employment sectors, for example construction workers (both formal and informal), members of the armed forces or civil servants. These individual programmes combined accounted for IDR 1.46 trillion in 2016, equalling 0.13% of total government expenditure and 0.01% of GDP.

Since 2015, the programme is complemented by a new pension scheme for formal workers, which is run on a defined-benefit basis. This scheme is currently in the accumulation phase, and as such, its costs are relatively low. The GoI is keeping contribution rates and the retirement age relatively low. However, to protect its long-term sustainability, these parameters will both need to increase within a relatively short period of time.

Low revenues are a major constraint on Indonesia’s development

Indonesia’s low level of domestic resource mobilisation, for a country at its income level, is a critical constraint on its ability to continue scaling up social assistance. It also limits the potential for redistribution through the fiscal framework, which is an important mechanism for tackling income poverty and inequality. The 2018 OECD Economic Survey for Indonesia devotes a chapter to raising public revenue through mechanisms that promote inclusive growth. It provides much greater detail on domestic resource mobilisation than is available here, as well as policy recommendations for enhancing revenues that are relevant for this analysis (OECD, 2018[9]).

Efforts to increase social insurance coverage can ease the direct burden on public finances. However, they must work in tandem with policies to increase tax revenues through broader formalisation policies. If these function coherently, a virtuous circle is achievable whereby a larger proportion of the population is formally employed and contributing to social security arrangements at the same time as paying more in taxes, with the consequence that workers’ livelihoods are protected and the government can afford to spend more on tax-financed social protection for individuals who are unable to work. Of course, taxes and social security contributions can also militate against formalisation by increasing the cost of employment; careful policy design and implementation are critical.

According to Ministry of Finance, tax revenue contributes a large majority of the total state budget revenue: 11.1% of GDP in 2016 versus 2.3% from non-tax revenues (Figure 4.11). The gap between tax and non-tax revenue has steadily increased since 2009, reflecting a gradual decline in the GoI’s reliance on income from the oil and gas sector.

Figure 4.11. Government revenues have declined as a proportion of GDP

Source: MoF (2019).

According to Revenue Statistics in Asian and Pacific Economies 2018, which allows for direct comparison of the level and structure of countries’ tax revenues by harmonising tax data, Indonesia’s tax-to-GDP ratio is substantially lower than that of its neighbours (OECD, 2018[14]). According to this report, Indonesia’s tax-to-GDP ratio was equivalent to 11.6% of GDP in 2016, versus 14.3% in Malaysia, 17.0% in the Philippines and 18.1% in Thailand (OECD, 2018[14]). Indonesia’s tax revenues as a percentage of GDP steadily declined since peaking in 2008 (OECD, 2018[14]), although they recovered in 2018 and are projected to increase further in 2019 (World Bank, 2018[8]).

Figure 4.12. Tax revenues are very low by regional standards

Source: OECD (2018[14]), Revenue Statistics in Asian and Pacific Economies 2018, https://doi.org/10.1787/9789264308091-en.

Figure 4.13 depicts the composition of tax revenues between 2001 and 2017. The share of income tax in total tax revenue collection increased by 1.5 percentage points to 34.1% in 2016, exceeding the pre-2012 levels, after lower collections between 2012 and 2015. Taxes on income from oil and gas declined, from 5.2% of tax revenues in 2014 to 1.8% in 2016.

Figure 4.13. Direct taxes account for over half of tax revenue

Source: MoF (2016).

The potential exists to enhance revenues from VAT, which is a relatively efficient source of revenue from an economic growth perspective. Indonesia can broaden the VAT base by removing a number of exemptions and lowering the VAT registration threshold, which is high relative to OECD countries. Such reforms would raise more revenue without increasing the rate, although it would nonetheless be important to calculate the impact on low-income households of removing the exemptions. Excise taxes on tobacco are also an important source of income (see Box 4.4).

Weak tax compliance contributes to Indonesia’s low tax-to-GDP ratio. Between 17.8% and 35.7% of Indonesians aged 15 or over were registered for personal income tax (PIT), which is low by emerging economy standards (OECD, 2018[9]). The GoI has taken steps to increase tax compliance and reduce tax evasion over the last decade, including multiple bilateral tax agreements and three tax amnesties in 2008, 2015 and 2016.

Box 4.4. Increasing tobacco excise taxes for revenues and health

As discussed in Chapter 1, smoking is emerging as a major threat to the Indonesian population’s long-term health, and there is a strong case for increasing tobacco taxation. Indonesia has one of the highest rates of smoking and tobacco use in the world – three-quarters of men smoke – and is one of the few countries where smoking rates have increased in the past decade. Taxation has proven a cost-effective means of reducing smoking internationally. Indonesia's tax rate remains below the World Health Organization’s recommendation of 70% of the retail price.

In its Economic Survey of 2018, the OECD recommends that Indonesia increase and harmonise excises across tobacco products (OECD, 2018[9]). At the same time, it is important that higher taxation be accompanied by public information campaigns to increase awareness of the dangers of smoking, as well as tighter regulation of tobacco advertising. In late 2018, the GoI cancelled plans to increase the excise tax on tobacco in 2019, making it the first year in which this has not happened since 2014 (Reuters, 2018[15]). Smoking is not the only major lifestyle-related health threat affecting Indonesia. With obesity rates also rising fast, there is a case for the GoI to use taxation to influence behaviour, and it is thought to be one of a number of countries in South-East Asia to be considering a tax on sugar-sweetened drinks (OECD, 2018[9]).

The oil and gas sector generates the majority of non-tax revenue but its contribution has declined as a percentage of GDP since 2001 (Figure 4.14). Other sources include natural resources, profits from SOEs, public service institutions and other non-tax revenues. Non-tax revenues peaked in 2006 at 8.2% of GDP but plunged sharply in 2009 to 4.6% of GDP. Since then, non-tax revenues have remained under 5% of GDP. In 2016, they amounted to 2.3% of GDP. External grants are a small and volatile component, accounting for 0.6% of total government revenue in 2016.

Figure 4.14. Declining oil revenues are bringing down non-tax revenue

Source: MoF (2018).

Tax buoyancy measures the efficiency and responsiveness of revenue mobilisation in response to growth in national income. Indonesia’s tax buoyancy between 2001 and 2016 was 0.84, implying that tax revenues rose at a slower rate than the economy grew over this period. This has been a continuous concern in Indonesia, largely due to the burgeoning informal sector, weak government monitoring and poor law enforcement that make tax evasion possible.

Indonesia’s fiscal deficit is narrowing and debt levels have declined

Indonesia is required by law to keep its combined fiscal deficit (central and regional government) below 3% of GDP. The fiscal deficit has gradually approached this threshold in recent years, reaching 2.5% of GDP in 2017 (Figure 4.15), reflecting an expansionary fiscal policy. However, the central government deficit declined to 1.7% of GDP in 2018, giving weight to estimates that the overall deficit will have narrowed to between 1.9% and 2.2% of GDP in 2018 (Reuters, 2019[16]). The deficit is expected to decline further in 2019 (OECD, 2018[9]; World Bank, 2018[8]).

Figure 4.15. The fiscal deficit widened after 2010

Note: RHS = right-hand side. LHS = left-hand side.

Source: MoF (2019).

Indonesia’s debt-to-GDP ratio has declined significantly over the past two decades, from 88.8% in 2000 to 29.8% in 2018 (Figure 4.16). Fiscal discipline, sustained and robust GDP growth, effective debt management and successful negotiations for lower interest rates have contributed to the decline. The GoI also sold several state-owned enterprises to pay off foreign debt in 2003-04 and restructured its foreign debt three times (2003, 2006 and 2008).

Figure 4.16. The debt-to-GDP ratio has declined significantly

Source: Directorate General of Budget Financing and Risk Management (2018).

While the debt-to-GDP ratio declined between 2000 and 2012, the GoI increased borrowing in 2014 to finance planned infrastructure investments. Nonetheless, the current debt-to-GDP ratio remains well below the 60% threshold established by Law No. 17 of 2003 and is lower than that of Malaysia, the Philippines and Thailand.

Recent upgrades to Indonesia’s credit rating reflect its strengthening macrofiscal environment. Standard and Poor’s Rating Agency, Fitch Ratings and Moody’s Investors Service all raised the country’s sovereign rating in 2017/18, with the result that it is now rated as investment grade by all three. These upgrades make it easier for Indonesia to attract foreign and domestic investment.

The combined impact of taxes and direct transfers is to reduce inequality

Taxes and social transfers offer governments powerful and flexible tools for achieving core development objectives while promoting social equity. These instruments serve multiple cross-cutting purposes, among them to 1) finance vital government activities; 2) create incentives to align private behaviour with social interests; 3) correct market failures; 4) tackle poverty and vulnerability; and 5) promote a more equitable distribution of income, wealth and other resources.

Fiscal incidence analysis can provide complex evaluations to assess the objective of equity, as often the real burden of taxation falls on individuals who might or might not be directly liable for the payments. For example, corporations pay corporate income tax (CIT) but the incidence ultimately falls on the shareholders, who receive lower after-tax dividends, employees who receive lower wages and consumers who pay higher prices.

Similarly, the outcomes of government transfers (and social protection programmes more broadly) often involve pathways complicated by targeting errors, opportunity costs and stochastic long-term payoffs. Both supply- and demand-side barriers create challenges for lower-income households to benefit from government expenditure aimed at reducing poverty and inequality.

This review carries out fiscal incidence analysis based on 2016 Survei Sosial Ekonomi Nasional (SUSENAS; National Socio-Economic Survey) data (Statistics Indonesia, 2016[17]). It simulates the five leading sources of tax revenues from direct taxes (PIT, CIT, property tax) and indirect taxes (VAT, excise tax), which together accounted for over 95% of government tax revenues from 2001 to 2017. Modelling the incidence of CIT, which accounts for about 32% of the tax burden, requires strong assumptions; in this case, the report assumes that the burden is distributed equally between owners of capital and consumers.

To analyse the impact of transfers, the model includes three major government transfer programmes: Rastra, PIP/Bantuan Siswa Miskin and PBI. PKH cannot be modelled because it is not specifically identified in the 2016 SUSENAS.

The total microsimulated tax revenues from the five taxes accounts for 96.6% of total reported tax revenues for 2016. Figure 4.17A demonstrates how the tax burden is distributed across consumption deciles, illustrating the burden as a percentage of household expenditure. The total tax burden is highest among households in the first decile: the effective tax rates reach 38.5% for the first decile, compared with 44.8% for the top decile.

Figure 4.17. The tax system is mildly progressive

Source: Authors’ calculations based on Statistics Indonesia (2016[17]), SUSENAS https://microdata.bps.go.id/mikrodata/index.php/catalog/769.

Individuals at the higher end of the income distribution pay a much larger share of total taxes than lower deciles (Figure 4.17B). The top decile pays a total of USD 31.3 billion, more than the bottom six deciles combined. The highest decile also accounts for 33.2% of the total tax burden, compared with 2.6% for the lowest decile. Given that the effective tax rates do not vary by such a wide margin, this disparity in tax payments reflects a concentration of consumption at the top of the distribution.

Direct and indirect taxes both weigh on low-income individuals (Figure 4.18). The impact of PIT increases with income, showing the progressivity of this instrument, although it is nonetheless surprising to see households in the lowest deciles paying any PIT at all. At the same time, the effective rate of VAT also increases slightly with income, indicating that consumers at higher-income levels consume more goods that are not exempt from income tax. The effective rate of excise taxes, imposed on cigarettes and alcohol, initially increases with income then declines.

Figure 4.18. Direct and indirect taxes weigh on consumption

Source: Authors’ calculations based on Statistics Indonesia (2016[17]), SUSENAS https://microdata.bps.go.id/mikrodata/index.php/catalog/769.

By modelling counterfactuals that eliminate taxes and social protection benefits, this report analyses the distributional impact of Indonesia’s taxes and transfers. The simulation assumes the baseline consumption is the final consumption that reflects taxes paid in line with the model above and social protection benefits received.

Counterfactual 1 reports household consumption in the absence of the social protection benefits. Subtracting the household per capita transfer amount from the household per capita consumption in the baseline scenario calculates the adjusted household consumption. The difference is the equity impact of the social protection system.

Counterfactual 2 reports household consumption in the absence of the tax system (but with the existing social protection benefits included). The model calculates this by adding the household per capita tax burden to the baseline household per capita consumption, which acts as a rebate of 100% of the taxes that households have paid and that households entirely consume. The difference between the baseline and this scenario represents the equity impact of the tax system.

Counterfactual 3 combines the other counterfactuals by removing the per capita tax burden and the per capita transfer amount for each household. In this scenario, the income aggregate for each household is their market income. Under the simplifying assumption that the marginal propensity to consume is one, the change in per capita income corresponds to a one-to-one change in per capita consumption.

Figure 4.19 shows the impact of taxes and transfers on inequality, as measured by the Gini coefficient. The Gini is lowest for the baseline; it increases from the baseline to counterfactual 1 and increases by a further 2.3 percentage points in counterfactual 2, indicating that the combined effect of taxes and transfers is to reduce inequality (as confirmed by counterfactual 3).

The reduction in inequality is consistent with the results of fiscal incidence analysis using the Commitment to Equity methodology (CEQ, 2017[18]) and based on SUSENAS data for 2012, which found that the combined impact of taxes and transfers was to reduce inequality by one percentage point (Jellema, Wai-Poi and Afkar, 2017[19]). This report confirms the finding in Chapter 3 that PKH (which is not included in the fiscal incidence analysis here) has the largest impact on poverty and inequality of any transfer, despite receiving a very low spending allocation. This finding leads the World Bank to conclude that “Indonesia has historically spent most on those programmes and policies that least reduce inequality in the short term, and little on those that have the greatest impact” (World Bank, 2016[20]).

Figure 4.19. Transfers do not outweigh the impact of taxes on poverty

Note: Counterfactual 1 removes social protection benefits from household consumption; Counterfactual 2 adds households’ average tax payments to their consumption and Counterfactual 3 combines both counterfactuals.

Source: Authors’ calculations based on Statistics Indonesia (2016[17]), SUSENAS https://microdata.bps.go.id/mikrodata/index.php/catalog/769.

Figure 4.19 shows the impact of taxes and transfers on poverty, as measured by the national poverty line. This exercise is important, since taxes can be both progressive and impoverishing at the same time.

Comparing baseline and counterfactual 1 illustrates the poverty-reducing impact of the transfers. The difference in poverty headcount ratio between the two scenarios is 2.4 percentage points, meaning that 6.2 million individuals would fall below the poverty line without the government transfers simulated here.

In counterfactual 2, when the model transfers the collected tax revenues back to households, the poverty rate decreases from the baseline, from 10.9% to 4.1%, reflecting increased household disposable income and per capita consumption. The result suggests that microsimulated taxes push about 6.8% of the population (17.5 million individuals) below the poverty line. The poverty-increasing impact of taxes is thus larger than the poverty-reducing impact of transfers.

The results of the analysis presented here should be treated with caution, principally because it only looks at social assistance transfers and does not include PKH. These account for a very small proportion of the broader social spending included in the fiscal analysis, which includes in-kind health and education benefits, as well as subsidies. Once these are taken into account, (Jellema, Wai-Poi and Afkar, 2017[19]) find that the fiscal system succeeds in reducing poverty as well as inequality.

References

[24] Chen, D. et al. (2018), Indonesia Economic Quarterly June 2018: Learning more, growing faster, World Bank, Jakarta, http://documents.worldbank.org/curated/en/305361528210283009/pdf/126891-WP-PUBLIC-on-6-5-18.pdf (accessed on 22 June 2018).

[11] Dartanto, T. (2017), Universal Health Coverage in Indonesia: Informality, Fiscal Risks and Fiscal Space for Financing UHC, International Monetary Fund (IMF)/Japan International Cooperation Agency (JICA), Tokyo.

[23] Gehrke, E. and R. Hartwig (2015), “How can public works programmes create sustainable employment?”, Discussion Paper, No. 11/2015, Deutsches Institut für Entwicklungspolitik [German Development Institute], Bonn, https://www.die-gdi.de/uploads/media/DP_11.2015.pdf (accessed on 22 June 2018).

[1] IMF (2017), IMF Fiscal Monitor: Tackling Inequality, October 2017, International Monetary Fund, Washington, DC, http://www.imf.org/en/Publications/FM/Issues/2017/10/05/fiscal-monitor-october-2017 (accessed on 22 June 2018).

[6] Jasmina, T. (2016), “Public Spending and Learning Outcomes of Basic Education at the District Level in Indonesia”, Economics and Finance in Indonesia, Vol. 62, pp. 180-190, https://ideas.repec.org/a/lpe/efijnl/201615.html (accessed on 5 December 2018).

[19] Jellema, J., M. Wai-Poi and R. Afkar (2017), “The Distributional Impact of Fiscal Policy in Indonesia”, in The Distributional Impact of Taxes and Transfers: Evidence From Eight Developing Countries, The World Bank, http://dx.doi.org/10.1596/978-1-4648-1091-6_ch5.

[3] Nasution, A. (2016), “Government Decentralization Program in Indonesia”, ADB Working Papers, No. 601, Asian Development Bank, Manila, https://www.adb.org/publications/government-decentralization-program-indonesia (accessed on 16 January 2019).

[18] Nora Lustig (ed.) (2017), Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty, Brookings Institution Press and Commitment to Equity (CEQ) Institute, New Orleans.

[5] OECD (2018), OECD Economic Surveys: Indonesia 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-idn-2018-en.

[9] OECD (2018), OECD Economic Surveys: Indonesia 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-idn-2018-en.

[14] OECD (2018), Revenue Statistics in Asian and Pacific Economies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264308091-en.

[2] OECD (2016), OECD Economic Surveys: Indonesia 2016, OECD Publishing, Paris, http://dx.doi.org/10.1787.

[22] Rahman, A. (2017), Tax compliance in Indonesia : the role of public officials as taxprayers, University of Twente, Enschede, The Netherlands, http://dx.doi.org/10.3990/1.9789036542425.

[16] Reuters (2019), “Indonesia posts narrowest budget deficit in 6 years in 2018”, Nikkei Asian Review, https://asia.nikkei.com/Economy/Indonesia-posts-narrowest-budget-deficit-in-6-years-in-2018 (accessed on 11 March 2019).

[15] Reuters (2018), “Indonesia to keep tobacco excise steady in 2019”, Reuters, https://www.reuters.com/article/indonesia-tobacco-tax/indonesia-to-keep-tobacco-excise-steady-in-2019-minister-idUSL4N1XG3M7 (accessed on 11 March 2019).

[12] Shin, J. et al. (2017), “Indonesia : Selected Issues”, Country Report, No. 17/48, International Monetary Fund, Washington, DC, https://www.imf.org/en/Publications/CR/Issues/2017/02/11/Indonesia-Selected-Issues-44654 (accessed on 22 June 2018).

[17] Statistics Indonesia (2016), Survei Sosial Ekonomi Nasional 2016 Maret (KOR), https://microdata.bps.go.id/mikrodata/index.php/catalog/769 (accessed on 22 June 2018).

[4] UNDP/UNCDF (2013), “Strengthening the Governance of Social Protection: The Role of Local Government Regional Analysis”, United Nations Development Programme (UNDP), UN Capital Development Fund (UNCDF), http://www.asia-pacific.undp.org/content/dam/rbap/docs/Research%20&%20Publications/democratic_governance/RBAP-DG-2014-Strengthening-Governance-of-Social-Protection.pdf (accessed on 16 January 2019).

[8] World Bank (2018), Indonesia Economic Quarterly December 2018 - Strengthening competitiveness, World Bank, http://documents.worldbank.org/curated/en/276171544466081210/pdf/132825-WP-P165633-PUBLIC-dec12-6pm-dc-add-series-IEQ-Dec-2018-ENG-for-web.pdf (accessed on 9 March 2019).

[7] World Bank (2017), Improving quality of spending in Indonesia: 2017 Budget and beyond, http://www.anggaran.depkeu.go.id/content/Publikasi/bidang%20pendidikan/improving%20quality%20of%20spending.pdf (accessed on 22 June 2018).

[10] World Bank (2017), Indonesia Public Expenditure and Financial Accountability, World Bank, Jakarta, https://openknowledge.worldbank.org/handle/10986/29820 (accessed on 31 January 2019).

[13] World Bank (2017), Program Appraisal Document on a Proposed Loan in the amount of US$200 million to the Republic of Indonesia for a Social Assistance Reform Program, World Bank, Washington DC, http://documents.worldbank.org/curated/en/353221496152466944/pdf/Program-Appraisal-Document-PAD-disclosable-version-P160665-2017-04-15-04202017.pdf (accessed on 16 January 2019).

[20] World Bank (2016), “Indonesia’s rising divide”, World Bank, Washington, DC, http://documents.worldbank.org/curated/en/267671467991932516/Indonesias-rising-divide (accessed on 7 December 2018).

[21] World Bank (2014), International Development Association Project Appraisal Document, for Productive Safety Net Project 4, Report No: PAD1022, World Bank, Washington, DC, http://documents.worldbank.org/curated/en/477011468037548038/pdf/PAD10220PAD0P1010Box385319B00OUO090.pdf (accessed on 29 June 2018).