May 2016 · Number 13 Closing Indonesia’s infrastructure gap: the key role of fiscal policy reforms Jamie Carter, Ndiame Diop, Arvind Nair, Alex Sienaert1 Indonesia’s fiscal policy performance examined Introduction Low deficits, low debt, fiscal prudence Fiscal policy in Indonesia has delivered a dramatic reduction in the government debt Indonesia has a hard-won reputation for burden, after it exploded as a result of the sound fiscal management, based on its 1997/98 crisis. This success has been adherence to strict fiscal rules, and falling supported by strict rules that cap the fiscal government indebtedness over the 2000s. deficit and debt ratios at 3 percent and 60 The government debt burden fell sharply percent of GDP, respectively. following the immediate recovery period Policymakers now face the added challenge from the 1997/8 crisis, from over 87 of funding a sustained lift-off in public percent of GDP in 2000 to a low of 23 infrastructure spending, a key objective of percent of GDP in 2012. This rapid the current administration. The task is a reduction was driven by consistent formidable one as the government seeks to surpluses of revenue, notably natural decisively close Indonesia’s large resource-related revenues, over non- infrastructure gap while maintaining its interest spending (there were primary hard-won reputation for fiscal prudence, Budget surpluses every year from 2001- even as revenues have been weakened by 2011, averaging 1.5 percent of GDP), the global commodities downturn and coupled with rapid nominal GDP growth slower GDP growth. Meeting this (the value of total output rose more than challenge will require consolidating and five-fold over this period in current US extending recent energy subsidy reforms, dollar terms). Consequently, Indonesia’s significantly improving fiscal revenue government debt burden fell not only to a collection and improving budget moderate level in absolute terms, but also management, including by enhancing the to well below peer economy levels (Figure quality of key Budget assumptions. 1), culminating in Indonesia regaining an investment grade sovereign credit rating in 1 This Practice Note has been cleared by Shubham Chaudhuri, Practice Manager, GMF06. 2012.2 debt sustainability is not a concern for Figure 1: The government debt burden fell sharply Indonesia at present.3 over the 2000s, bottoming out in 2012. (General government gross debt, percent of GDP) Significant under-spending in 100 infrastructure Indonesia All developing economies 90 For much of the period 2005-2014, the 80 central government’s spending on 70 infrastructure was significantly crowded 60 out by large energy subsidies. Energy 50 subsidies rose in the mid-2000s and 40 reached unsustainable levels in recent years 30 despite several ad hoc, episodic retail price 20 increases. In 2014, spending on energy 10 subsidies accounted for more than one-fifth 0 of the central government’s budget. This 2000 2002 2004 2006 2008 2010 2012 2014 was more than three times the allocation for Sources: IMF. infrastructure such as roads, water, electricity and irrigation networks, and three times government-wide spending on From 2012 onwards, however, health. Total investment in infrastructure— macroeconomic conditions have become that is, investment by the central more challenging; global commodity prices government, sub-national governments, have slumped and overall external demand state-owned enterprises and the private growth has been sluggish, weakening sector—has remained at only 3 to 4 percent Indonesia’s external balances, and of GDP over the past decade. This is far requiring significant monetary policy and below the rates of above 7 percent of GDP exchange rate adjustments. Nominal GDP before the 1997 Asian financial crisis and growth has fallen sharply, with the value of the 10 percent and 7.5 percent spent by annual output contracting by over 3 percent China and India, respectively (Ihsan 2012, in current US dollar terms from 2011 to 2013). 2014. During this period, Indonesia’s fiscal stance became more expansionary (with the High volatility of spending and revenues primary balance swinging into modest but sustained deficits, averaging 0.9 percent of The fiscal picture in the last decade also GDP from 2012-2014), and the government shows significant in-year and year-to-year debt burden inched up (by 2 percent of volatility in major budget components, GDP, to 25 percent, in 2014). Yet, overall, especially energy subsidy and capital Indonesia’s fiscal management has expenditures, and natural resource-related remained prudent; fiscal deficits have revenues. Overall, the fiscal balance shifted continued to be capped at low levels; and, towards moderate deficits in recent years aided also by the robust risk profile of its (Figure 2). During 2010-2012, higher debt stock, government debt distress risks energy subsidy spending relative to remain low. Consequently, government nominal GDP was the major drag on the 2 Defined as having investment grade sovereign credit ratings 3 In particular, IMF staff assess that “public debt is sustainable from two out of the three large international credit ratings and robust to macroeconomic shocks”, IMF Article IV Staff agencies. As of April 2016, only Standard & Poor’s rated Report, March 2015. Indonesian sovereign credit below investment grade (by one notch, at BB+). May 2016 · Number 13 · 2 fiscal balance, followed since 2013 by comprised an average of 30 percent of declining revenues relative to GDP, revenues from 2004-14. This reliance, especially from natural resources. In sharpened by the decade-long commodity addition, there were pronounced swings in boom in commodity prices and production, capital expenditures from year to year. has come with risks, as commodity prices are volatile, both within-years and across The key sources of year-to-year years, and prices of significant commodity fluctuations in the Budget have been exports for Indonesia have fallen similar when evaluated over a longer period significantly since their peak in 2011 (Table 1, for 2004-2014). The major year- (Figure 3), in turn driving down the value to-year fluctuations on the spending side of exports and state revenues. The baseline were subsidies and capital expenditures. In expectation of World Bank staff is for addition, these same expenditure global commodity prices to remain components have been very volatile within subdued in coming years (Figure 3). years, with outturns typically deviating Consequently, in the absence of significant significantly from initially budgeted increases in non-resource revenues, the amounts for the year (Table 1). downward trend in the natural resource sector is likely to put significant pressure On the revenue side, the major source of on revenue collection and fiscal deficits in year-to-year and within-year fluctuations in the medium term. the last decade has come from natural resource non-tax revenues (Table 1, for 2004-14). Indonesia remains reliant on natural resource revenues, which Figure 2: A mix of expenditure and revenue Table 1: Natural resource revenues, and capital and factors have pressured the fiscal balance in subsidies spending, have been volatile both across recent years and within years (Contributions of revenue and expenditure (Standard deviations of major Budget items, changes to change in fiscal balance from the percent, 2004-2014) year before, percent of GDP) 3 Variation Variation across within 2 years years Total revenues 7.0 13.7 1 Tax 6.0 9.3 Non-tax from natural 0 17.8 43.5 resources All other revenues 11.8 18.6 -1 -2 Total expenditures 6.5 11.2 Revenues Interest expenditure Personnel 9.5 2.4 -3 Energy subsidies Material 11.9 8.6 Capital expenditure -4 Capital 24.2 10.7 Other expenditures Fiscal balance Interest 6.6 10.3 -5 Subsidies 23.3 105.3 2008 2009 2010 2011 2012 2013 2014 Transfers to regions 6.0 6.7 Notes: All items refer to changes compared with the Notes: “Variation across years” is the standard deviation from the previous year; a positive number indicates that the change linear trend of annual Budget outturns, “Variation within years” is in the item increased the fiscal balance. the standard deviation of the difference between outturns and Sources: Calculations based on Budget outturns. original Budgets (APBN). Sample period: 2004-2014. Sources: Calculations based on MOF Budget data. May 2016 · Number 13 · 3 Figure 3: Indonesia’s natural resource revenues have been tightly linked to volatile global commodity prices, which are expected to remain subdued in coming years (Annual change, percent) 80 Oil & gas-related income Commodity price tax forecast period 60 Non-tax natural resource revenues 40 20 Global commodity prices 0 -20 -40 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Notes: Percent change in revenues refers to revenues in nominal Rupiah; “Global commodity prices” are the simple average of World Bank energy and non-energy commodity USD price indices and forecasts. Sources: Calculations based on MoF data; World Bank Commodity Outlook. Over-optimistic budget assumptions affected by political negotiations, being formulated by the Ministry of Finance The macroeconomic projections in the (MOF) and then reviewed and revised by initial budgets presented to Parliament— Parliament as part of the wider Budget notably for GDP growth (Figure 4) and also approval process. This politicization of crude oil production—have tended, in assumptions can be expected prima facie to many years, to be more positive than generate less accurate Budget assumptions outcomes (Table 2). In contrast, private than if these were purely technically-driven sector analyst forecasts have tended to be so as to minimize forecast errors. In more accurate, at least for GDP growth particular, a positive bias is to be expected (Figure 4), a key macroeconomic given normal political pressures to spend assumption, and the variable for which more in the short-term. samples of analyst projections are readily available. The tendency for Budget assumptions to be too optimistic, relative to contemporaneous independent projections and to outcomes, is widespread across countries, and applies in both advanced and developing economies (Frankel 2011). The reasons for this optimism bias are multiple and differ across countries, and are difficult to establish conclusively. In Indonesia, two plausible, and related, reasons stand out. First, some key assumptions, such as GDP growth and oil production, may have a perceived “targeting” function, signaling that the government is confident about the economic outlook and has ambitious objectives. Second, assumptions are May 2016 · Number 13 · 4 Figure 4: Budgeted GDP growth assumptions have Table 2: Key Budget assumptions have differed tended to exceed actual outcomes considerably from outturns, especially since the (Real GDP growth, percent) post-2011 commodity down-turn (Average difference between outturn and Budgeted assumption; percent discrepancy in parentheses) 7 2004- 2012- 2004- 6 2011 2014 2014 Nominal GDP 393,498 -112,614 255,468 5 (IDR billion) (8.6%) (-1.2%) (4.4%) 4 Real GDP (%) -0.2 -0.8 -0.4 Actual-budget gap (-4.2%) (-13.8%) (-6.8%) 3 Budget Inflation (%) 1.6 1.8 1.7 Actual (20.9%) (25.4%) (22.0%) 2 Currency 112 1,038 364 Consensus (Rp/US$) (1.2%) (9.8%) (3.8%) 1 Interest rate (SBI 0.1 -1.0 -0.2 0 avg., %) (1.9%) (-22.2%) (-2.3%) Crude oil price 15.8 6.9 13.3 -1 (US$pb) (21.8%) (6.5%) (16.4%) -2 Oil production -72.7 -80.0 -74.7 (1000 bpd) (-7.6%) (-9.7%) (-8.1%) Notes: “Budget” is assumption in original Budget (APBN); Sources: MOF. “consensus” is the average of commercial analysts’ forecasts as of December of the prior year. Sources: MOF; Consensus Economics. Strict adherence to fiscal rules However, the rules have also had unintended consequences, as can be seen Indonesia’s fiscal rule, set out in by considering the changes adopted as part Regulation 23/2003, (1) limits the of annual revised Budgets (APBNP), cumulative deficit of the general promulgated around the middle of each government Budget to a maximum of 3 year. The pattern has been for capital percent of GDP for the year, and (2) expenditures to be revised up in revised constrains the cumulative general Budgets, only for the final outturn to be government debt to a maximum of 60 below the original Budget allocation (as has percent of GDP for the year. Implicitly, the been the case in every recent year, except rule has been interpreted to cap central 2009; Figure 5). However, when Budget government fiscal deficits at 2.5 percent of pressures are pronounced, as in 2008 or GDP, in order to allow a buffer for regional 2014, there have been substantial government cash deficits of up to 0.5 downward revisions. More generally, percent of GDP. These rules have conferred revised Budgets have tended to have three major advantages: they are simple and easy characteristics: (i) a proposed deficit that is to understand, and have become closely close to the 2.5 percent of GDP ceiling for associated with its track record of fiscal central government expenditure, (ii) prudence and rapid debt burden reduction understating (often significantly) actual over the 2000s. In particular, they have subsidies expenditure, and (iii) overstating provided a valuable, strong signal to actual capital expenditure. lenders of a commitment to prudent borrowing including even, for example, One hypothesis that would explain this during the 2008/9 global financial crisis pattern is that ministries have purposely period, when the non-cyclically adjusted underspent their capital Budgets in order to budget deficit was kept at a small 1.5 ensure that the fiscal rule is not breached. It percent of GDP despite slower growth. may also be that ministries have been May 2016 · Number 13 · 5 allocated additional money too late in the to have hampered the quality of year (following the revised Budget) for it to infrastructure spending. Delivering large be feasible to develop and execute amounts of infrastructure at the lowest investment spending. This would mean that price requires certainty. Ideally, certainty the need to cap the fiscal deficit to less than would be provided on a medium-term basis the 3 percent of GDP limit has been a so that ministries can plan for future years contributing factor to the volatility of but, at a minimum, there is a need to ensure capital expenditures, and especially to that spending agencies have confidence reduced capital spending at times of higher that the Budget they receive at the start of than expected energy subsidy spending the year is fixed (both in terms of not being (2010-2012; Figure 6) or lower than changed explicitly, but also that they will expected revenues (2012-present).4 not be discouraged from spending it in full). The gains from increased consistency and A further indication that adhering to the certainty of capital Budgeting could be fiscal rules may have affected Budget material; for example, Infrastructure UK’s management in recent years is the presence Infrastructure Cost Review indicates that of significant arrears, since pushing cash extending funding certainty in OECD payments to subsequent years is a way of countries has been associated with unit cost suppressing the fiscal deficit for the current savings of 10–20 percent, particularly for period. The APBN-P 2014 estimated that at routine maintenance and renewals work. the end of 2014 government arrears Raising the efficiency of existing spending amounted to IDR 45 trillion (approximately by 10-20 percent, while also facilitating 0.4 percent of GDP). higher execution rates by making it easier to plan in advance, would constitute a Regardless of the exact causes, the significant boost to infrastructure spending historical pattern of setting a low initial outcomes. capital spending Budget, significantly altering it during the year, and then underspending against plans, is very likely Figure 5: Revised Budgets have tended to increase Figure 6: Capital expenditures have consistently capital allocations, but outturns have almost without been below revised Budget levels, significantly exception remained lower than originally Budgeted offsetting a tendency for subsidy costs to exceed (Capex deviations from original Budget, percent) revised Budget levels Subsidy spending (outturn vs. revised budget) 20% Revised budget Actual outturn 50% Capital execution (outturn vs. revised budget) 5 15% Fiscal deficit under revised budget (RHS) 40% 4 10% 5% 30% 3 0% 20% 2 -5% 10% 1 -10% -15% 0% 0 -20% -10% -1 -25% -20% -2 -30% -35% -30% -3 2008 2009 2010 2011 2012 2013 2014 2008 2009 2010 2011 2012 2013 2014 Sources: Calculations based on MOF Budget data. Sources: Calculations based on MOF Budget data. 4 While this paper focuses on infrastructure spending challenges frequent under-execution of the capital budget, and volatility of arising from the fiscal rule and fiscal policy, major public capital spending, as outlined for example by Ahern et al. (2012). investment management challenges have also contributed to the May 2016 · Number 13 · 6 In sum, the interaction of the fiscal rule weaker international commodity prices, with high and volatile energy spending, notably oil. Consequently, against a target volatile resource-related revenues, and of increasing total nominal revenues by inaccurate Budget assumptions, has tended 13.6 percent in 2015, actual revenues historically to make infrastructure spending collected fell by 11.4 percent in 2015 a “residual” spending item. (World Bank 2015a). This poor performance highlights the need to broaden Securing the fiscal space for an the tax revenue base from currently low infrastructure lift-off: the levels (10.9 percent of GDP in 2014), which will require fundamental tax challenge administration reforms to increase tax compliance, such as improving taxpayer Indonesia’s current administration under data gathering and analytics – a major, President Joko Widodo took office in medium-term challenge. October 2014. A stated priority of this administration was to plug Indonesia’s Within the context of weak medium-term infrastructure gap, by redirecting spending revenue collection, sustaining the current, decisively away from energy subsidies, positive infrastructure spending impulse, lifting revenues, and ramping up total let alone increasing it, will be challenging. (public and private) infrastructure spending Even the rise seen so far in infrastructure to well above the 3-4 percent of GDP spending contributed to an increase in the average achieved in the previous decade fiscal deficit to 2.7 percent of GDP in 2015. (World Bank 2013). This is close to the de facto fiscal deficit limit under Indonesia’s fiscal rule. Now, in the second year of the current administration’s term, and after a slow Budget credibility relies on recognizing start, public infrastructure spending has explicitly, and responding effectively, to all indeed gathered significant pace; nominal these factors. If more and better-planned spending by the central government on infrastructure spending is to be central to infrastructure increased by 42 percent in the budget allocation, as appears to be the 2015 even if the IDR 209 trillion spent fell strong commitment of the current short of the government target by a wide government, then comprehensive fiscal margin (IDR 67 trillion). In real terms, management improvements will be needed. public spending in infrastructure increased by 14.5 percent, significantly lifting real gross fixed investment growth, to 5.1 Policy considerations for percent. The public spending momentum is sustaining higher infrastructure maintained in the 2016 Budget with IDR spending 313 trillion allocated for capital spending. To close the infrastructure gap in the years Consolidate recent energy pricing to come, achieving further increases in both reforms the quantity, and the quality, of infrastructure spending, will be needed. As discussed above, subsidy costs, the most significant of which have been for energy, Policymakers confront a tension between particularly fuels, have been a major the government’s infrastructure spending contributor to fiscal pressures and risks in commitments, revenue collection Indonesia in recent years. By 2014, fuel constraints, and adherence to the fiscal rule. subsidies had swollen to about a fifth of Revenues face headwinds from slower central government spending, or 2.4 nominal GDP growth since 2012, and percent of GDP. Not only did fuel subsidy May 2016 · Number 13 · 7 spending trend higher over the years due to Optimize resource revenue collection, rapidly increasing volumetric demand as a while reducing reliance on resource result of solid economic growth, but revenues fluctuating global oil prices and exchange rates also made subsidy costs particularly The recent downturn, as well as continued difficult to project accurately and to plan volatility, in resource revenues, is for. Over 2010-2014, fuel subsidy spending unavoidable as it is linked to global averaged 2.2 percent of GDP per year and resource prices which are volatile, on a exceeded initially Budgeted costs by an downward trend, and outside of Indonesia’s average of 0.5 percent of GDP, driving control as the country is a price taker for much of the overall volatility in most commodities. expenditures and accounting for about half (0.75 percent points) of the 1.5 percent of However, this does not mean that Indonesia GDP deterioration in the fiscal deficit from cannot benefit from improvements in 2010 to 2014. Consequently, fuel subsidies resource revenues collections. While the have been responsible for much of the case for increasing tax and royalty rates is adjustments required for discretionary weakened by lower prices, there are still items, especially capital expenditures. significant gains to be made from improving natural resource revenue Recognizing the problem, and the strong administration. For instance, a joint World case more broadly for eliminating energy Bank and Ministry of Finance study found subsidies in Indonesia as elsewhere (Diop that close to 50 percent of mining royalties 2015), the current government has already that could have been collected, given made major reforms (World Bank 2015b). existing production, prices and royalty As of the start of 2015, gasoline subsidies rates, were not collected from 2011-2014, were eliminated and the diesel subsidy was accounting for close to 1 percent of GDP capped at a maximum of IDR 1,000 per (World Bank 2014). Mining royalty, and liter. Consequently, Budgeted fuel subsidy likely also other resource revenue, costs have fallen from 2.4 percent of GDP administration could be significantly in 2014 to 0.6 percent of GDP in 2015. improved to being collection closer to its Ongoing reforms have also begun to reduce potential. electricity subsidy costs (Budgeted at 0.6 percent of GDP in 2015). At the same time as optimizing resource revenue collection, the government could However, while the government has also look to reduce the dependence of the remained publicly committed to lower fuel budget on resource revenues over the subsidy spending, implementation of the longer-term. Fiscal analysis has shown that new fuels pricing system has thus far been revenue buoyancy in the last decade has uneven. In particular, the monthly remained stagnant, implying that non- adjustment of retail gasoline and diesel resource revenues have not kept pace with prices envisaged in the implementing expanding economic output (World Bank regulations has not taken place. This risks 2014). Reversing this trend is critical and undermining the credibility of the reform, involves actively broadening the tax base, and in particular raises concerns that fuel improving revenue administration as well subsidies could reemerge over time. as improving compliance, among other Avoiding this outcome will be a crucial measures. element of raising and improving infrastructure spending in Indonesia. May 2016 · Number 13 · 8 Improve the quality of macroeconomic separate and independent government forecasts to reduce the optimism bias in bodies that calculate the economic the Budget-setting process assumptions, such as the Netherlands Central Planning Bureau. The economic Indonesia’s recent fiscal history suggests assumptions may also be calculated by non- that Budget assumptions which over- government organizations. Canada, for estimate outcomes can have a direct example, bases its assumptions on an bearing on public capital expenditures, average of leading private sector forecasts; since these are a major discretionary item Chile employs an independent non- that can be cut or delayed during the course governmental panel of experts as part of a of the year in order to ensure that the fiscal comprehensive approach to budget deficit rule is not breached. Such in-year formulation which has been seen as a adjustments impede not only the level, but strong model for resource-rich countries also the efficiency, of infrastructure (Frankel 2011; Calitz, Siebrits and Stuart spending. More generally, transparent, 2013). Regardless of the exact method plausible and mutually consistent used, the development of credible and underlying macroeconomic assumptions capable institutional frameworks, such as are crucial to the realism and credibility of separate government bodies to formulate a government’s Budget (IMF 2007). As the assumptions, takes time and effort. noted in the OECD’s 2009 review of Budgeting in Indonesia, the greatest risk is In Indonesia, the OECD’s 2009 review of for the assumptions to be “too optimistic”, Budgeting in Indonesia highlighted that the thus making it seem that more resources are role of the House of Representatives available than is really the case (Blöndal, (Dewan Perwakilan Rakyat, DPR) stands Hawkesworth and Choi 2009). out when compared to that of most OECD member country parliaments, going beyond Best practice is to ensure the independence the usual democratic Budget oversight role of those responsible for the calculations, to played by the legislature. This includes the insulate them from political pressure. high degree to which the DPR is involved However, putting in place the institutions, in the details of the Budget and in the and building the capacity, to ensure that number of occasions through the Budget transparent and credible economic process. In addition “the Indonesian assumptions form the basis for annual, and practice of politically negotiating the multi-year, Budgets continues to be a economic assumptions, albeit within challenge faced by many, if not most, calculated ranges, between the government countries, which have responded with a and Parliament is not applied in OECD wide variety of institutional frameworks. In countries” (Blöndal, Hawkesworth and some countries, such as the Nordic Choi 2009, p.16). countries and Mexico, for example, assumptions are the responsibility of the The introduction of an independent economics departments within finance institution which sets the underlying macro ministries. Similarly, in Australia, the assumptions for the Budget is a longer-term macro assumptions used in the Budget are institutional change which could be used to made independently by the Treasury help to depoliticize the assumption-setting department. This set-up is similar to that in process. In the short-term, adding greater Indonesia but the difference is that in transparency and verification on how the Indonesia the assumptions are reviewed by Ministry of Finance formulates the Parliament rather than an independent government’s assumptions could also body, which inevitably politicizes the strengthen the assumptions. For example, discussion. In other countries, there are the MOF could highlight which variables May 2016 · Number 13 · 9 can be most clearly viewed as exogenous, spending decisions. One example would be and not as possible targets which can be the decisions around public sector pay influenced by government policy, and rises. It would be possible to set aside discuss with the DPR the possibility of money for the annual public sector pay setting such assumptions in a technical increase in the original Budget (APBN), manner, limiting political discussion over but to actually set the level in the revised their level. The calculation methodology Budget (APBN-P) later in the year, with the could be described in detail in background annual increase thus being awarded from notes or briefings to the DPR. Providing June or July. This way, if there was a need information on how other countries set to lower expenditure in the revised Budget similar variables could reassure the DPR on then this could be done by setting the pay the appropriateness and robustness of such rise at a more affordable level rather than technical approaches. by all savings coming from front-line public services. Currently, despite being Increase flexibility in less economically over a fifth of central government important Budget expenditure to absorb expenditure, personnel expenditure does in-year pressures not bear any of the risks associated with fluctuations in Budget resources. There is always a possibility that in-year expenditure adjustments will need to be Progressively move towards a more made, even with appropriate forecasts, due medium-term fiscal framework to unforeseen developments in the presence of a fiscal deficit limit. The steps above set The above discussion highlights out how to minimize this risk, but it is also improvements in fiscal management that important to have effective modalities in could be made within the existing Budget place to respond as efficiently as possible policy framework. However, as set out should the risk transpire. A goal of this above, the combination of annual Budget should be to insulate high value economic cycles and the fiscal rule has also been expenditure such as infrastructure associated with significant capital investment from in-year changes in the expenditure fluctuations and squeezes, and overall resource envelope in the Budget. periods of arrears building up. Put another way, the current annual fiscal rules have This could be achieved in part by having a many positive elements: they are easy to buffer, for instance only setting the Budget understand, easy to manage and ensure to target a 2 percent of GDP fiscal deficit, conservative borrowing levels. However, so that small adverse changes do not they are also a blunt instrument, having require remedial action. Arguably, recent assured fiscal discipline but at a high cost fiscal management has included an element at times to the amount, and quality, of of this, although it has not been of sufficient infrastructure spending. magnitude to fully offset the positive skew of the final Budget assumptions. The There is rich international experience with downside of this approach, however, is that fiscal rules, showing mixed results, in part it leads to lower overall expenditure, which because they have been implemented with is sub-optimal given the potentially high many variations (Schaechter et al 2012; returns to public investment and Bova et al 2015), including in the time- Indonesia’s ambitious development goals. period over which they measure fiscal performance, the scope of the rule, the An alternative would be to have a types of expenditure that they cover and mechanism to absorb shocks by holding whether they are measured on a cash or back until the revised Budged some major accruals basis. Perhaps the key lesson from May 2016 · Number 13 · 10 the international experience is that while stakeholders to monitor; and (iv) can be fiscal rules can play a useful role in implemented given existing institutions, or reinforcing fiscal discipline, they should the government is prepared to set up the not be seen as a panacea, as they can have necessary institutions. unintended consequences, and their success depends critically on wider fiscal For example, this could entail management, institutions and the political implementing a medium-term fiscal rule, economy of the Budget. In Indonesia, there such as moving from an annual fiscal is little doubt that the fiscal rule serves as a deficit limit of 3 percent of GDP to a rule vital anchor for policy, but this should not that the fiscal deficit not exceed 3 percent preclude examining whether the country of GDP across a number of years. If could build further on this success. appropriately implemented, such a rule would increase budget flexibility, The specific issue of arrears building up especially enabling firm multi-year capital due to pressures to delay spending in order spending allocations. For such a medium- to stay within the fiscal deficit limit could term rule to continue to enshrine fiscal be resolved by the government moving its discipline, it will need to be credible. fiscal reporting to an accrual basis (that is, Future years’ projections for both revenue reflecting costs and income when they are and expenditure would need to be based on actually incurred, not when the cash realistic, technocratic forecasts, not on changes hands). Line Ministries began desired outcomes (i.e. “targets”), and any reporting their accounts on an accrual basis deviations from the plan would need to be from 2015. Building on this by moving the clearly set out and explained. The change in full fiscal accounting system to an accruals rule would also need to be clearly basis would remove any incentives to communicated to financial markets, to generate arrears to comply with the fiscal ensure that investors saw the change as a deficit limit, since delaying cash payments positive step forward in fiscal management, have no effect on the recorded fiscal deficit given the need to plan infrastructure under a full accruals system, as well as spending on a multi-year basis, and not a conferring other benefits. strategy to raise the trajectory of borrowing. A medium-term fiscal rule Ultimately, adopting a multi-year would therefore need to be implemented as budgeting cycle is the natural solution to part of a comprehensive, rolling medium- the problem of the current annual fiscal rule term budgeting framework, underpinned by subjecting capital budgeting and spending credible, transparent forecasts (for a recent to excessive uncertainty and in-year review of the international experience and adjustments. Adopting such a medium- evidence regarding pre-requisites for term approach to Budgeting would need to success, see Brumby et al 2012). This is a ensure that the significant benefits of the challenging fiscal reform agenda, the current annual fiscal rules are preserved. implementation of which will take time. Consequently, reform objectives would need to include (i) ensuring the ability to Conclusion continue to tightly control the overall fiscal position and net financing needs, and Indonesia has established a strong track ensuring that they do not blunt incentives to record of overall public borrowing restraint continue improving overall efficiencies; (ii) and enjoys a low debt burden with minimal creating an environment that is conducive risks of debt distress. Its clear fiscal rules to high quality public investment (i.e. that have underpinned this success. However, limits the need for in-year adjustments); historically, the fiscal rules have not been (iii) is relatively easy to explain and for costless, and they do not operate in a May 2016 · Number 13 · 11 vacuum. Rather, they have interacted with About the authors: politicized budget assumptions, volatile energy subsidy spending and volatile Alex Sienaert, Senior Economist, World Bank’s Macroeconomics & Fiscal Management resource-related revenues, to make Global Practice (GMFDR) discretionary, and especially infrastructure, Email: asienaert@worldbank.org spending highly volatile, both across and Ndiame Diop, Lead Economist, World within annual budget cycles; infrastructure Bank’s Macroeconomics & Fiscal Management became something of a residual spending Global Practice (GMFDR) Email: ndiop@worldbank.org item and was frequently squeezed when circumstances caused there to be a risk of Arvind Nair, Economist, World Bank’s Governance Global Practice (GGODR) breaching the fiscal deficit limit. Email: anair3@worldbank.org Now, Indonesia’s government is Jamie Carter, Public Financial Management Advisor, World Bank’s Governance Global determined to implement a major Practice (GGODR) infrastructure spending program. This has fiscal policy implications that need to be confronted and managed. In particular, delivering on infrastructure commitments while respecting the fiscal rule will require that energy subsidy spending be kept low and ultimately eliminated; the quality of infrastructure spending needs to be improved by planning and implementing multi-year projects; and revenues need to be increased. In parallel, fiscal reforms could work over time towards developing a solid medium-term fiscal framework based around a credible, sustainable borrowing trajectory, that would ultimately increase overall budget flexibility, capital budgeting certainty and, potentially, also the level of infrastructure spending. May 2016 · Number 13 · 12 References Ahern, M., Ihsan, A., Purnomo, H, Sienaert, A., Thomas, T. and Taylor, A., 2012, “Identifying Constraints to Budget Execution in Indonesia’s Infrastructure Sector”, available at: http://documents.worldbank.org/curated/en/2015/08/24881575/identifying-constraints-budget- execution-indonesia%E2%80%99s-infrastructure-sector Blöndal, J., Hawkesworth, I. and Choi, H., 2009, “Budgeting in Indonesia, OECD Journal on Budgeting, available at: http://www.oecd.org/indonesia/45362389.pdf. Bova, E., Kinda, T., Muthoora P., Toscani F., 2015, “Fiscal Rules at a Glance”, IMF background document, available at: https://www.imf.org/external/datamapper/FiscalRules/Fiscal%20Rules%20at%20a%20Glance%20 -%20Background%20Paper.pdf Brumby, J. et al, 2013, “Beyond the Annual Budget: Global Experience with Medium-Term Expenditure Frameworks”, World Bank Directions in Development Series Calitz, E., Siebrits, K. and Stuart, I., 2013, “Enhancing the credibility of fiscal forecasts in South Africa: is a fiscal council the only way?”, Stellenbosch Economic Working Papers: 25/13 Diop, N., 2014, “Why is reducing energy subsidies a prudent, fair and transformative policy for Indonesia?”, Economic Premise, World Bank, available at: http://siteresources.worldbank.org/EXTPREMNET/Resources/EP140.pdf Frankel, J., 2011, “Over-optimism in forecasts by official budget agencies and its implications”, Oxford Review of Economic Policy, vol. 27 no. 44, pp. 536-562. Ihsan, A., “Identifying the constraints to budget execution in the infrastructure sector”, in World Bank, July 2012, Indonesia Economic Quarterly “Rising to Present and Future Challenges”, available at: http://siteresources.worldbank.org/INTINDONESIA/Resources/Publication/280016- 1328062662077/8406495-1342061334479/IEQ-JULY-2012-English.pdf Ihsan, A., “Piecing together the picture of Indonesia's infrastructure investment trends”, in World Bank, March 2013, Indonesia Economic Quarterly “Pressures Mounting”, available at: http://www.worldbank.org/content/dam/Worldbank/document/EAP/Indonesia/IEQ-MARCH- 2013-English.pdf IMF 2015, Country Report 15/74 “Indonesia 2014 Article IV Consultation – Staff Report”. IMF 2007, “Manual on Fiscal Transparency”, available at: http://www.imf.org/external/np/fad/trans/manual.htm Infrastructure UK (HM Treasury), 2010, “Infrastructure Cost Review”, available at: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/192588/cost_review _main211210.pdf Jakarta Post, December 24, 2014, “Jokowi aims for 7 percent annual growth”, available at: http://www.thejakartapost.com/news/2014/12/24/jokowi-aims-7-percent-annual-growth.html May 2016 · Number 13 · 13 Schaechter, A., Kinda, T., Budina, N., and Weber, A. 2012, “Fiscal Rules in Response to the Crisis—Toward the ‘Next-Generation’ Rules. A New Dataset”, IMF WP/12/187. World Bank, October 2013, Indonesia Economic Quarterly “Continuing adjustment”, available at: http://www.worldbank.org/content/dam/Worldbank/document/EAP/Indonesia/IEQ-Oct2013- ENG.pdf World Bank, December 2014, Indonesia Economic Quarterly “Delivering change”, available at: http://www.worldbank.org/content/dam/Worldbank/document/EAP/Indonesia/IEQ-DEC-2014- ENpdf.pdf. World Bank, December 2015a, Indonesia Economic Quarterly “Reforming amid uncertainty”, available at: http://pubdocs.worldbank.org/pubdocs/publicdoc/2015/12/844171450085661051/IEQ-DEC-2015- ENG.pdf. World Bank, July 2015b, Indonesia Economic Quarterly “Slower gains”, available at: http://www.worldbank.org/content/dam/Worldbank/document/EAP/Indonesia/IEQ-JUL-2015- english.pdf. May 2016 · Number 13 · 14