83545 From Universal Price Subsidies to Modern Social Assistance: the Political Economy of Reform 1. Introduction The efficiency and effectiveness of generalized subsides as part of a safety net system are questionable. In the first place, generalized subsidies tend to create distortions in the food and fuel markets 1. Secondly, as evidence from several countries shows, generalized subsidies are likely to be regressive and to suffer from significant leakage of benefits to the non-poor, diluting their impact on poverty reduction 2. Thirdly, subsides often have higher administrative costs than cash transfers, particularly where such programs involve food distribution and require supply side capacity and logistics that may be under-developed. Finally, subsidies also reduce spending available for other social services and infrastructure which could benefit the poor more directly. The concerns expressed in this report are certainly not new, and many countries in the world have already gone through significant attempt to reform at different points in time. Yet, two developments in particular provide a renewed opportunity to look at the subsidy issue: 1. the present economic juncture in the region and elsewhere, with increasing fiscal pressure due to rising food and fuel prices, combined with 2. the new momentum in strengthening social assistance systems throughout the region The policy response to the recent crises in the region highlighted emerging interest for a transition from food and fuel subsidies toward greater reliance on targeted cash transfer programs in several countries. A review undertaken in the aftermath of the food and financial crises 3 showed that while some countries provided incremental resources to food-based programs in response to the crises, this approach was less widespread or pronounced than might have been expected. Malaysia and Cambodia significantly increased their food subsidies in late 2008 and early 2009 as part of the GFC/food crisis response. In contrast, in Indonesia, a targeted cash benefit (a revamped version of the BLT cash transfer which had been used in 2005 to mitigate the social impact of fuel price reforms) was introduced in 2008. While the rice subsidy program has been retained, the marginal increments in spending in social assistance crisis response have been through cash. In the Philippines, rice subsidy spending was ramped up in 2008 in response to the food crisis, but then reduced and offset by a targeted cash benefit in 2009 during the GFC response. The transition from a strong reliance on generalized subsidies to more modern and administratively more sophisticated social protection has been witnessed in much of the developing world as countries move from lower to middle income status. As countries experience economic growth, they often also 1 According to an IEG review of climate change and energy policy reforms, studies have found that a 10 percent increase in energy prices would be expected to reduce long-run demand by 7 percent, on average (IEG, 2009) 2 For example, Jha (2010) reviews the case of food subsidies in Philippines and India, showing very high exclusion rates of the poor (75% in the Philippines, 70% in India), as well as inclusion errors of around 70 percent in urban areas. 3 ASEAN and World Bank (2010) 1 tend to have higher inequality and greater variation in household and individual vulnerability profiles. In such cases, targeting is harder (errors can be greater) but also yields highest returns. As countries acquire greater administrative capacity, and deeper understanding of evolving poverty and vulnerability profiles, greater emphasis on cash transfers and other forms of household assistance and becomes an important policy alternative. This is a relevant issue for many countries in East Asia today. However, cutting back subsidies has social as well as political implications. Removing subsidies and raising prices needs to be well managed. For one thing, in many cases social assistance programs will need to be strengthened so as to help poor and vulnerable households weather the price shock 4. Another is to increase public understanding and support for subsidy reform by having a transparent and evidence-based discussion and scrutiny of subsidies: the full cost of the subsidy, the distribution of the subsidy and who is benefiting from the subsidy, and the implications for public spending on priority areas. This paper draws upon some of the most recent efforts to look at the political economy of subsidy reform (see Commander 2012 among others) and, in particular the socio-political risk of cutting subsidies in the context of modernizing social assistance. The shift in dialogue over subsidy reform, from an almost exclusive focus on the fiscal cost of subsidies to broader considerations of what kind of compensation should be offered in the context of reform make it particularly important to shed more light on the perceptions of society and how these views can contribute/hinder reform progress. Moreover, such an approach has the potential to provide practical information that reform champions within and outside the government can act upon through changes in communication strategies and policy design. After outlining the main factors behind the persistence of costly and ineffective subsidies, therefore, the paper chooses to focus the attention on the often overlooked dynamic link between perceptions and policy, and does so from the particular angle of social protection by considering the implications of moving from a universal social protection mechanisms to a more modern one relying on some degree of targeting. To do so, it draws upon literature on factor shaping preferences for universal versus targeted transfers, and on the qualitative findings of two background studies – in Indonesia and the Philippines – where a series of focus groups were conducted to understand citizen perceptions regarding existing price subsidies and their effectiveness, as well as expectations regarding social protection measures and possible compensation packages. These qualitative findings are placed within the institutional and, to some extent, political context to understand how these perceptions can contribute to shape the political economy of reform. 2. Price subsidies as a social protection mechanism Objectives and evolution Price subsidies are policies aiming at modifying consumer or producer prices in order to pursue given economic or social outcomes and/or overcome specific market failures. Thus, price subsides of various kinds are used primarily to alter consumption or production patterns in order to achieve particular policy objectives (such as increasing economies of scale, addressing externalities, stimulating chosen industries, etc.) or to redistribute resources and benefit certain groups, such as the poor. In many cases, 4 The consensus on the importance of considering implications of changes to the subsidies in terms of the overall safety net is growing, see World Bank. OECD and OPEC Joint Report (2010), UNDP (2012), Coady et al. (2010), Commander (2012), etc. 2 they are introduced as temporary income buffer and become persistent over time. They can also be used for ideological reasons, to pursue self-sufficiency or to share national natural resource. Consumer price subsidies in particular are commonly justified as support for parts of populations that could not afford the full market price of particular goods. The introduction of food subsidies was also used to finance industrial growth through implicit transfers of income from producers to consumers. In East Asia, as in other regions such as the Middle East, over time consumer price subsidies have become a key component of public spending and of the social contract 5. Consumer fuel subsidies in Indonesia date back to the 1960s, when the Government used them to mitigate the effects of its increased developmental expenditures and subsequent rampant inflation on households’ budgets 6. The current rice subsidy program, Raskin, was first introduced in response to the Asian Financial Crisis in 1997. Of the different crisis-initiatives, only the rice subsidy continued to receive significant budget allocations and has become a permanent feature of the Indonesian safety net. In 2010, Raskin represented 53 percent of total social assistance expenditure 7. Similarly, the origins of the Philippines National Food Authority 8 agency can be traced back to a critical rice shortage in the early 70s. Today, rice is often referred to as a political commodity in the Philippines with the price of rice playing significant role in election results 9. The same patterns can be found in the Middle East, where the surge in subsidies was fuelled by the oil price shock of 1973. Government revenues, inflated by oil exports, were used to finance and justify the expansion of subsidies to protect consumers from the inflationary effects. From a political and ideological perspective, universal consumer price subsidies allowed governments to claim that national wealth was being shared. In Egypt, the state has made explicit its mandate to ensure basic food supplies since the Nasser era in the 1950s. Over time, the provision of subsidized bread has become a powerful symbol of the broader social contract in a political system where political participation is highly limited10. From a social protection perspective, we are most interested in the role price subsidies play in protecting the population against price fluctuations of goods that are thought to be essential. Controlling prices of important commodities such as staple food items and fuel is crucial both to protect consumption levels of the poor and to reduce the political risk that traditionally comes with higher prices. In this sense, price subsidies are a risk reduction strategy for population and for that reason uncertainties related to exclusion from possible transfers that are proposed to replace them are the key factor that can undermine public support to reform. Cost and Effectiveness (distributional impact, etc): the case for reform The fiscal burden of subsidy programs is significant and it has been growing. Food and fuel prices have been very volatile in the last XX years, but remain high (see Figure XX). In the absence of reform, higher prices naturally lead to higher spending on consumer subsidies, as countries are reluctant to pass on the 5 More and more African countries are now also subsidizing fuel prices reluctant to pass on the full cost of rising prices to their citizens. See for example World Bank (2012a) 6 In 1965, fuel subsidies were around 20 percent of GDP (Beaton and Lontoh, 2010 – IISD Indonesia) 7 Indonesia SAPER (2012). In 2010, Raskin aimed to supply over 3 million tons of rice to 17.5 million poor and near-poor households 8 As will be outlined in more detail below, NFA is mandated to procure to (a) support prices for rice farmers and (b) provides subsidies to the general consumers through the sale of cheap and subsidized NFA rice. 9 Intal and Garcia (2005) 10 IFPRI (2000), Gurtner (1999) 3 increases in international prices. Measuring and comparing subsidies as a share of GDP across countries is not straightforward, but estimates from the IMF database show that fuel subsidies represented, on average, over 2 percent of GDP in East Asia, and 3.5 percent in Eastern Europe and almost 5 percent of GDP in the Middle East and North Africa between 2002 and 2008. 11 As a share of GDP, six economies spent more than 2 per cent of their GDP subsidizing fossil fuels in 2010: Brunei Darussalam (2.6 per cent), Indonesia (2.3 per cent), Malaysia (2.4 per cent), Russia (2.7 per cent), Thailand (2.7 per cent) and Vietnam (2.8 per cent). In 2010-11 over half of all African countries had some subsidy in place for fuel products, and these in turn cost on average, 1.4 percent of GDP in public revenues. These costs have grown sharply in some countries in recent years 12. The cost of food and fuel subsidies has become significant, and poses an increasing fiscal risk in many countries. Table 1: Estimated global energy subsidies, 2007–2010 (US$ billion) 2007 2008 2009 2010 Fossil fuels 342 554 300 409 Oil 186 285 122 193 Gas 74 135 85 91 Source: GSI (2012) Subsidies consume valuable government resources and crowd out much needed investments in human capital and infrastructure. In most cases, the share of government expenditure on fuel subsidies already exceeds social spending on pro-poor sectors such as health and education. The large increases in spending to finance price subsidies translate into expenditure cuts for needed infrastructure investment as well as social sectors such as health and education. Achieving reform could allow governments to invest more actively in specific priority sectors. The opportunity cost of subsidies in terms of development policies is big, and it is growing fast. Moreover, generalized subsidies are often not effective in reaching the poor, with most of the benefits accruing to the richest quintiles. Several studies have looked into the effectiveness of generalized subsidies on protecting incomes, and in particular the incomes of the poor (Coady et al. 2010, Alderman 2002, Arze del Granado 2010, IMF 2008 among others). The evidence is that, with some differences related to specific types of fuels such as kerosene distribution of price subsidies tends to be regressive 13. On average, the top income quintile receives about six times more in subsidies than the bottom quintile (Figure XX).This is also often true for food subsidies. Coady, Grosh, Hoddinott (2002) for example look at targeting performance of safety net programs and find that universal food subsidies are rarely progressive. Only 3 out of 15 programs for which they had evidence on targeting performance were found to be progressive, and the median targeting performance was regressive (with the poorest 30 percent receiving only 29 percent of total transfers). However, the design of the program can make a big difference for the extent of leakage and the effectiveness in reaching the poorest. The choice of 11 Commander (2012) 12 WB, Africa’s Pulse (2012) 13 Furthermore, when differences between market and regulated prices are greater, consumption of subsidized goods, and divergence to not-intended beneficiaries increases 4 which goods to subsidize, of packaging and of distribution channels among other things can impact the outcome. Figure 1: Share of total subsidies captured by each quintile 70 60 50 Q1 40 Q2 30 Q3 20 Q4 10 Q5 0 Gasoline Kerosene LPG Source: Arze del Granado (2010) Many countries in the region have fairly under-developed or nascent social assistance programs, but over time, an inclusive and functioning safety net is a more effective social protection instrument. Rapid and shared growth after the 60s reduced the need/urgency of a social contract that included a safety net. In the fastest growing economies, increased labor mobility and rapidly rising incomes generated strong expectations of future improvements in welfare. According to Birdsall and Haggard (2000), lack of political pressure for a social contract also reflected the political strength of the business community, on which the governments relied for growth. In the same study, the authors note how export-orientation made businesses particularly sensitive to taxes and labor costs, in such a way that while the private sectors were in favor of greater social investment in education (which enhanced labor productivity), they were typically hostile to a broader social agenda. Moreover, the restricted political space and at times outright repressive nature of some regimes constrained any nascent political demand for state-run social safety nets, unless used by the government to reduce social pressure during periodic political or economic crises. 3. Subsidies persist for political reasons Despite their high cost/low effectiveness, price subsidies can be extremely persistent and hard to remove for a number of important reasons. None of these reasons can usually be considered single- handedly responsible for their persistence, but each can play a role over time in the onset and evolution of the subsidy system.  Concern with social and political risk vis-à-vis the removal of blanket subsidies and substitution with alternative measures of social protection. The political cost is immediate and certain, the benefits are uncertain and accrue over a longer time horizon 5  Misunderstanding/misperception about the benefit incidence of the current policy and uncertainty regarding the distribution of gains and losses from reform. This is true both on the policy side as well as the general public  Persistence of an ideological content, such as a desire to reach food self-sufficiency, or to share existing national natural resource wealth  Weak capacity to implement alternative, more targeted, mechanisms to protect income  Rent-seeking opportunities develop over time and firmly entrenched in the current system  Finally, all of the above can be exacerbated by individuals and/or institutional actors with limited direct stake in the subsidy reform using the issue for political reasons 14 Because politicians may not be able to get credit for initiating important reforms but have to sustain the cost of unpopular decisions, they focus on policies that can be delivered quickly with little opposition. Decisions about reforming subsidies and transitioning towards more modern forms of social assistance are influenced by many of the same factors affecting the political incentives of social expenditure decision: (i) the extent of access to relevant information on the part of the public, (ii) collective action problems, (iii) lack of credibility of political promises, (iv) time-horizon of policies and (v) practices of client-patron politics. The uncertainty of a concrete return for the politician makes far- reaching reforms extremely unlikely. In Indonesia, subsidies remain persistent so political cost was achieved with only a short-term fiscal gain. Depending on each country’s specific objectives, the calculation and decision on how to proceed with subsidy reform will involve considering the trade-off between more economically efficient options that are seen as having low political acceptability, and politically acceptable options that may have high cost implications. Different combinations of these options can be designed to reflect the ways that policy goals are prioritized. Expanding consumer subsidy programs is always politically easier than reducing them. Gutner (1999) mentions a high number of attempts to food subsidy reform that sparked violent protests, often seen by policymakers as politically destabilizing 15. However, Bienen and Gersovitz (1986) argue that, while “subsidy cuts have political salience […] it is not clear that [they] pose special long-run problems for political stability compared to cutting other public expenditures, devaluations, and import restrictions.” They also suggest that “the short-run problems associated with subsidy cuts can be mitigated, though not overcome, by tactical proficiency” and that “even violent reactions to such cuts are not decisive for political stability”. More recently, episodes of violence associated to rises in fuel prices have taken place 14 For example in the Philippines some of the opposition to removing rice subsidies and is coming from opposition politicians who simply seem to be "inciting" negative perceptions on the CCT by leveraging the “dependency” and “patronage” arguments to criticize the government use of CCT. The unconditional cash transfer (BLT) which was used twice to compensate households for rises in fuel prices in Indonesian is also being used in similar ways. 15 Increases in food prices have resulted in violent protests in countries including Dominican Republic (1994), Egypt (1977), Indonesia (1998), Jordan (1989, 1996), Liberia (1979), Morocco (1981, 1984), Peru (1978), Sudan (1979, 1982), Tunisia (1984), Yemen (1998), Zambia (1986), and Zimbabwe (1988). 6 in Africa (Mozambique, Nigeria 16), and Indonesia decided to halt its plans to proceed with new fuel price increases after civil unrest in anticipation of the parliamentary debate on reform 17. Cuttings subsides has a social cost that if not properly accounted for, can be large. Even though the majority of benefits of universal subsidies accrues to richer segments of the population, the impact of higher prices can be devastating for the poor and therefore, protective measures that can shelter the most vulnerable during a period of reform are essential. The cases of Jordan, Iran and Indonesia fuel subsidy reform are particularly instructive in this instance. Also Tunisia food (Tuck and Lindert. 1996). At the same time, global experience also points to the fact that providing compensatory measures may not be sufficient for governments to get the political and general public support for reforms, especially when compensations are targeted to the poor. Political and institutional factors will directly affect the space for reform. In a recent paper developed for the World Bank, Commander et al. 18 look into political and institutional factors associated with energy subsidies. They identify a statistically robust association between energy subsidies – as determined by the subsidy to GDP ratio – and political systems, as well as quality of institutions 19. In particular, greater use of energy subsidies appears to be associated with weaker institutions. In turn, institutional frailty appears to be associated with the type of political system, with less competitive and less democratic regimes displaying weaker institutions. The author also points to the fact that non- democratic polities tend also to display high inequality, combined with resistance to implement redistributive policies such as higher taxes on the better. In order to lower risks of social turmoil governments may therefore choose to resort to some form of income (not asset) transfers to the rest of the population, such as consumer price subsidies. This is consistent with the view expressed in focus groups that while not particularly useful, consumer subsidies are at least providing a tangible – and hence credible - transfer to the population20. The author also points out that the status-quo bias may be even stronger in less democratic regimes, where the absence of a political mechanism for reducing uncertainty about outcomes (by punishing non-performing incumbents) over time may further reduce credibility and undermine reform. In her analysis of consumer price subsidies and options for reform in the Middle East, Kelly (1999) highlights as barriers to further reform the possibility of social unrest and the region’s poor experience with targeted cash transfers. The author describes how inadequate compensation or ineffective communication have hindered reform and suggests that because the small transfer programs in the Middle East have lacked effective targeting and delivery mechanisms, an already skeptical public viewed them with suspicion. Therefore, a successful transition to a cash transfers based system would require a significant improvement in administrative capacity, and a process of confidence building between the 16 In 2011, Mozambique raised fuel prices twice, by 10 and 8 percent respectively, and Ghana by 30 percent. Guinea also introduced measures to reduce the fuel subsidy. On January 1, 2012, the Nigerian government removed the fuel subsidy on gasoline, an action that sparked violent protests, after which a portion of the subsidy was reinstated (WB, Africa Pulse 2012) 17 The Indonesian government was planning to increase fuel prices on April 1, 2012, which led to large protests across the country. Indonesian Police recorded there were at least 168 demonstrations on the last working day before the parliament announced its decision on the fuel price increase proposal (Lontoh et al., 2012). The political battle in the parliament eventually ended up with a conditional suspension of the planned fuel price increase. 18 Commander et al. (2012) 19 The variable “Polity2” is a composite measure of the political system developed by the authors, while institutional quality is captured through the World Governance Indicators (Government Effectiveness, Rule of Law, Regulatory Quality and Freedom from Corruption) 20 Commander (2012) mentions focus groups conducted in the Middle East (Morocco and Egypt), but similar views have been found in both our country cases (Indonesia and Philippines) as well as in sub-Saharan Africa (Cameroon) 7 government and society. The administrative and political contexts can influence what kind of program is feasible. High/low administrative capacity, the structure of political institutions and strength of interest groups, need for parliamentary support, link to health and education policies on which stronger views by public and interest groups Within a given political and institutional context, reform strategies can be strengthened to mitigate the social and political costs. Political will and institutional capacity to implement reform are natural prerequisites for moving away from entrenched price subsidies. However, success requires an ability to understand views and expectations of key stakeholders, including the constraints posed by public opinion and attitudes. Reviews of global experiences with subsidy reform (Commander 2012, Kelly 2009, but also Tuck and Lindert 1996 for Tunisia for example) point to three key elements that can strengthen the governments approach to reform: (i) timing, (ii) an appropriate compensation package that takes into consideration the impact on different groups and (iii) communication. Table 2 highlights the possible use of compensation packages and communication strategies in addressing some of the main reasons for persistence of subsidies. Table 2: Role of compensation and communication in easing resistance to reform Reasons for Persistence of Role of compensation Role of communication subsidies package strategy Concern with social and political risk of Yes, very important. Yes, can be influenced by a reform. The political cost is immediate Need to understand how communication strategy and certain, the benefits are uncertain different groups will be and accrue over a longer time horizon affected and provide appropriate  Misunderstanding/misperception compensation/protection about the benefit incidence and the cost of the current policy  Uncertainty regarding the distribution of gains and losses from reform Weak capacity to implement Maybe. If compensation Maybe. Communication alternative mechanisms to protect package is designed in a strategy may help with the income and lack of vision way that makes it easier element of trust in to implement and more government capacity, but credible not with the fundamentals of implementation Ideological content is persistent (ex. Not much Yes, can be influenced by a Food self sufficiency, entitlement to communication strategy natural resources) 8 Rent-seeking opportunities in the Maybe, though needs to Yes, can be exposed by current system be carefully assessed to appropriate communication limit cost strategy. Early communication can also help create coalitions which may overcome rent-seekers Individual and/or institutional actors No Maybe. Early with limited direct stake in the subsidy communication/explanation reform per se using the reform for may limit the opportunity other – often political – reasons A careful analysis of the winning and losing groups can shed light on the best mechanism to use in order to ensure that new transfers reach the right people without compromising the reform effort. To the extent that subsidies reach the poor, any alternative instrument must consider the impact of their withdrawal on poor households. At the same time, a viable solution must take into considerations all “losers” who have been benefiting from universal subsides and in many cases these are not poor. A well designed compensation package should be based on solid poverty and social impact analysis to determine distribution of losses, but also to better understand the social impact and what compensation could be effective, based on people’s needs, perceptions and expectations. It should also be part of a clear vision of where to reallocate budgetary savings to crucial public investments (education, health, priority infrastructure) or even tax reform Extensive efforts to publicize and explain reforms to the public, presenting the shortcomings of the subsidy systems and detailing a vision for reform can help credibility and address misunderstandings and perceptions of inequity. Public acceptance is more likely if the rationale is explained well in advance. Early communication and explanations may also prevent some groups from using the reform to make unsubstantiated claims that may support a hidden political agenda. Finally, communication and publicity can contribute to overcome information asymmetry that hinders the formation of coalitions of public support that could offset opposition from established interest groups. 9 Box 1: Determinants of support for redistribution Countries differ widely in their views about the government’s role in the provision of public services and on the extent of the need and responsibility to provide for those unable to provide for themselves. The establishment of a permanent welfare system requires the development of a sustainable social contract, and public choices about allocation of scarce resources. In developing countries, social assistance programs are often financed with support from external funds in times of crisis, and may therefore not require the kinds of political economy choices that are necessary to establish permanent systems for social assistance (Graham, 2002). Some issues may perhaps be considered generic to the political economy of social policies across countries, as they shape the political cleavages around which positions are formed. o Targeted vs. universal o Temporary vs permanent o Education and human capital vs. social assistance o Deserving vs. “undeserving” poor o Racial heterogeneity and immigration o Tradeoffs between objectives (i.e. discourage dependencies, improve life for families and children) Following Graham (2002) these issues will hinge to a large extent on who the public deems deserving of social assistance and to what extent that choice is dictated by similarities or differences between those needing 1 assistance and the general public. Grosh and al (2008) look at evidence from the WVS and similar surveys and show correlation between higher social welfare spending and the belief that poverty is caused by forces outside of an individual’s control, rather than individual behavior and laziness. Existing literature focuses on the United 1 States, Europe and Latin America . The evidence found underscores the importance of attitudes regarding redistribution and who are the “deserving” poor in shaping the social contract and influencing government’s choices. Of course, societal attitudes are only one of the elements influencing welfare policy choices. In the same study, Grosh and al (2008) look at the relationship between spending patterns and factors such as income levels, 1 inequality, voice, democracy and attitudes about inequality . According to many, universal schemes enjoy more political support than targeted schemes and are, therefore, more likely both to be successful, and to survive. Segal (2012) discusses some of the literature on the move from general to targeted subsidies. Cornia and Stewart (1993, p. 473) find that the move towards targeted subsidies in Zambia and Sri Lanka was accompanied by a reduction in real value of the subsidy over time, and link this to a reduction in political support for the targeted scheme. Similarly, Skocpol (1991, p. 263) argues that in the United States “the most successful measures— Civil War pensions and social security—have been those that ensured entitlements to broad categories of people.” More recently, Gelbach and Pritchett (2002) formalize this discussion with a model of the government budget where the executive chooses a universal or a targeted scheme, subject to majority voting. In equilibrium the majority assign so little of the budget to a targeted scheme that the poor are better off under a universal scheme. The outcomes in practice of such situation will be highly contextual but it should be kept in mind that electoral politics matter and doing the best for the poor requires taking a view on (i) whether the support for a particular program will depend on the way in which it is targeted, and (ii) in what way these preferences will influence political decisions. At the same time, the political choices that need to be highlighted may not always be between targeted and universal services. For example the distinction between deserving and non-deserving poor may be a more important determinant of public support for welfare than is that between targeted and universal programs, which has received more attention in academic literature (Graham, 2002) – in this case, substituting universal for targeted scheme would not be the main determinant of support for redistribution. For example, in cash transfer programs in Latin America the main reason for adding joint responsibility in building human capital between a program and its beneficiaries was to maximize the program’s impact on intergenerational poverty. However, this feature also proved critical in changing the political acceptability of SSN programs in Latin America (Grosh et al., 2008). 10 4. An application: Philippines and Indonesia We now turn to looking more in detail at the cases of energy subsidy reform in Indonesia, and rice subsidy reform in the Philippines. The choice of such different contexts and different types of goods (with all the implications for the political economy of the sector, as well as the perceptions and values attached to the goods) speaks to the particular angle pursued in this report. As mentioned in the beginning, the attempt is to focus the attention on one dimension which is particularly important when thinking of subsidy reform as a package including the implementation of social protection measures: people’s perception regarding existing policies and social protection expectations. are critical to inform (i) the design of an appropriate compensation package and (ii) the design of an effective communication strategy. It is therefore particularly useful to focus attention on countries where the government is at present considering options for reform, such as Indonesia and the Philippines. From a political economy perspective there are several interesting dimensions that could emerge from this comparison between food and fuel: (i) different drivers behind the policy choice of using subsidies as a response mechanism, (ii) different impact of their removal and the implications for the definition of a compensation package and (iii) consumer attitudes and expectations may be different depending on how entitled they feel towards the subsidy and the role it plays in their consumption patterns. However, the situation presents some important similarities for our purpose:  Objective of price stabilization, coupled with  Objective of supporting more vulnerable of populations that can’t afford the full market price of the given good  Implicit targeting of the poor through quality (for rice in the Philippines) or moral persuasion (for energy in Indonesia), but absence of real restrictions on access  Long-standing subsidies which people feel entitled to,  Various reform attempts over time, and on-going dialogue on reform, but only partial progress In addition, the choice of countries is also instrumental to consider the political economy implications of moving from untargeted to targeted transfers. In both countries targeted transfers have been introduced in the past, but with important differences: - In the Philippines, the CCT is a well established program, which was introduced as a pilot in 2007 and has since expanded quite rapidly (from an initial 360,000 beneficiary household to 2.3 million, representing roughly 44 percent of poor households). The program’s first large expansion was in 2008, in response to the food and fuel crisis. - In Indonesia, the temporary unconditional cash transfer (BLT) was first introduced in 2005 to mitigate the impact of fuel price increases, and covered 35 percent of the population. The poorest deciles received 21 percent of the benefits, and deciles 2,3 and 4 captured 40 percent. The BLT was then repeated in 2008-2009 when the food and fuel crises was once again pushing the fiscal cost of the subsidies and forcing the Government to raise prices. 11 We will now briefly describe the situation in each country before moving to a discussion of the qualitative findings and how these can contribute to shape the political economy of a reform strategy. The subsidy Four fossil-fuel products are the focus of price subsidies in Indonesia: premium-brand gasoline, solar- brand diesel, kerosene and liquefied natural gas (LNG). The Ministry of Energy and Mineral Resources, in communication with the Ministry of Finance are responsible for setting the fuel price, taking into account an international benchmark price. The exact formula and considerations for price-setting are not known. The subsidy is paid to Pertamina, the sole distributor of subsidized fuel products in the country, every three months to reimburse it for the below-market products it has sold during this time. In order to receive this payment, Pertamina provides the MoF with monthly reports on the volume of subsidized fuel it sold, as well as the difference in price. Adjustments to the size of the payment can also be made, but have to go through parliament. The process is audited once a year by the Audit Board of the Republic of Indonesia (Badan Pemeriksa Keuangan or BPK). At the pump, consumers can choose among three different prices (and products), with richer consumers “encouraged” not to buy the subsidized products, which are meant for the poor. In 2008 spending on fuel subsidies was four times greater than spending on targeted safety nets at US$ 23 billion, equivalent to 4.5 percent of GDP on energy subsidies and less than 1percent on social assistance. This represented as much as 32 percent of total central government expenditure. Even once fuel prices decreased, in 2010 2 percent on subsidies and 1 percent on social assistance 21. Moreover, energy subsidies do not target poorer segments of the population, as most benefits accrue to commercial users and non-poor car-owning households. Throughout 2010 and 2011 there has been considerable debate in the press, particularly regarding option for fuel subsidies’ reform. 22 However, reform has stalled. 21 Indonesia Household Social Assistance Public Expenditure and Program Review 22 For example, the current administration is considering a plan to prohibit private cars from using subsidized gasoline, and this has attracted considerable press coverage and spurred debate. WB Indonesia Economic Quarterly, March 2011 12 Figure 2: Indonesia: spending on subsidies and spending on SA, % of GDP, 2010 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Social Energy Rice subsidy Assistance subsidies In the Philippines, consumer rice subsidies are administered by the National Food Authority (NFA) as part of its mandate to support price and supply stability. Basically, the NFA buys high (from farmers), sells low (to consumers), and stores rice to create buffer stocks. Operationally, the NFA’s programs are linked to another objective: to provide low-cost rice to the poor. Policies to sustain rice production and limit exposure to the international rice market, such as import quotas, are maintaining prices artificially high, so the NFA has to ensure availability of lower-than-market-price rice through its accredited retailers in addition to its own outlets. The amount of the subsidy varies depending on international prices, but can be significant. There is no restriction as to who can buy NFA rice and how much rice one can purchase 23. However, NFA rice is of lower quality and implicitly targets the poor. Discussion over inefficiency of NFA and reform of rice subsidies has been ongoing in the Philippines for the past thirty years. In December 2011 retail price of NFA rice was increased from Php 25/kg to Php 27/kg. Fernandez and Velarde (2012) provide a full analysis of the incidence and cost-effectiveness of the NFA rice program. 23 During the food crisis the Government made a rapid effort to target the poor by issuing “Family Access Cards” (FACs) to poor households in Metro Manila. The government relied on local government officials to identify the poor, and around 315,000 such cards had been issued to poor families by the end of 2008. Today, however, the rice cards are no longer required. 13 Figure 3: Philippines: spending on subsidies and spending on SA, % of GDP, 2008 0.6 0.5 0.4 0.3 0.2 0.1 0 Social Assistance Rice subsidy The following sections will illustrate people’s perceptions, as uncovered through the two background studies in the Philippines and Indonesia. The Focus Group Discussions conducted between April-May 2012 had the objective of assessing public perceptions regarding subsidies and the prospect of higher prices as a result of subsidy reform, and to provide some insight into the preferences and needs for alternative social protection measures. The study also attempted to understand whether the following factors played a role in shaping these preferences: (i) previous experiences with targeted transfers or other benefits, (ii) income levels and utilization of subsidy, (iii) awareness and knowledge of subsidy mechanisms. Utilization of subsidy and impact of higher prices, coping strategies When faced with energy price increases, the poor cut expenditures, seek additional work and resort to negative coping strategies such as decreasing food consumption, and taking additional loans. As expected, in the rural areas, people use firewood as a substitute for gas as much as possible, but this is rarely possible in urban areas. The first coping strategy naturally is to cut other expenditures, but the room is not much. For example, education related expenditures make up a fourth to a third of the household expenditure for most participants, and many are concerned of being unable to offer opportunities to their children. Participants claimed to minimize the use of utilities and water, their food intake and to seek alternative employment or mobilize more family members to work. These extra workloads often placed a significant burden on the participants and are not usually very remunerative. The poor also often resort to borrowing money from their families, employers, friends, and loan sharks to cope with the increase of fuel prices. In the case of the Philippines coping strategies aim at sustaining current levels of consumption of quality rice even at the cost of sacrifices in other areas. For poor respondents, rice is their biggest household expense, taking up at least 30 to 50% of their food budget. The availability of quality rice is what defines the quality of their meal (“You can cost cut on the dish as long as you have good and tasty rice”). Increases of a 50-centavo to Php 1 per kilo in the price of rice would raise concerns, but still be manageable. A Php 2-4 per kilo increase would entail sacrifice in order to maintain high levels of (still commercial) rice by cutting costs of other food items such as beef, pork, and fish and, if this is not enough, by mixing available rice with root crops, or cooking it so it goes a longer way, i.e. in porridge form. Only an increase of Php 5 and higher per kilo would push them to really cut consumption of rice. Some participants would turn to NFA rice or purchase corn or root crops in its place. NFA rice is usually a last resort because of its perceived low quality. 14 Even the poor do not report consuming subsidized rice, with few exceptions. Poor quality, small price difference with current commercial rice and lack of accessibility are the main reasons for not resorting to NFA rice. NFA rice was consumed in Albay, but not in Pampanga, NCR, Butuan and Surigao. Albay is a very poor province that is only now beginning to recover from two "If [there is a] Php 5 increase, it devastating typhoons. In other regions, where the poor would depend. I would rather run more easily have access to cash, such as in NCR and after cheap and stored NFA rice than Pampanga for example, they are willing to use it to pay for to see my child starve to death.” better quality rice. Moreover, at current price levels, subsidized rice only costs Php 1 to 3 less than the cheapest commercial rice 24. Finally, in many cases, participants report having to travel far distances and entail transportation expenses in order to reach points of sale. In fact, the distribution of NFA rice does not appear to be sensitive to poverty levels. 25 The National Capital Region (NCR) and Central Luzon (Region 3) are two regions that stand out for having received a much higher share of subsidized rice than their respective poverty share would have suggested (Table 3). Table 3: Distribution of NFA Rice, 2009 Region Poverty Rate Distribution of Rice Share to rice (% poor HHs) (Metric Tons) distribution (%) Region 1 17.8 111,201 6 Region 2 14.5 56,335 3 Region 3 12 230,480 12.4 Region 4 13.7 224,392 12 Region 5 36 173,430 9.3 Region 6 23.8 72,426 3.9 Region 7 30.2 94,742 5.1 Region 8 33.2 88,495 4.7 Region 9 36.6 72,287 3.9 Region 10 32.8 60,066 3.2 Region 11 25.6 104,132 5.6 Region 12 28.1 40,179 2.2 NCR 2.6 431,875 23.2 ARMM 38.1 56,311 3 CARAGA 39.8 46,851 2.5 Philippines 20.9 1,863,202 100 Sources: Fernandez and Velarde (2012). Based on NSCB for poverty rates and NFA for the rice distribution. Fuel consumption patterns are highly dependent on occupation and location; therefore some groups are more vulnerable to changes in fuel prices. For example in Indonesia, fishermen may consume up to 30 liters of diesel for one single trip, short-range buses and taxi drivers consume an average 30-40 liters of subsidized gasoline per day, and food vendors may consume 12 kilogram of cooking gas per week. Some individuals also have long commutes to reach their jobs. Construction workers in Babelan, for 24 Php 28 for NFA rice versus Php 29-32 for commercial rice 25 Fernandez and Velarde (2012). See also Jha and Mehta (2008) 15 example, spend 2-3 liters of gasoline everyday (equivalent to 50-80 kilometers of commuting with motorcycle to reach their workplaces and conduct everyday activities (travel to the market, school, etc). They admitted that having motorcycle is a must-have asset if someone does not want to get stuck within the locality. For those whose employment is not only highly dependent on fuels, but also highly competitive, such as street vendors and taxi drivers, fuel price increases were seen as a threat to their financial stability, as they simultaneously increase their cost and reduce their market, as clients start looking for cheaper alternatives. In the past, the practice of mixing automotive fuel (benzene and diesel) with cheaper, subsidized kerosene was also common, but this practice was deserted since the elimination of the kerosene subsidy 26, which put the kerosene price higher than the price of automotive fuel products. Fishermen and taxi drivers would be more cautious when planning their work (e.g. taxi drivers preferred to station in one area instead of roaming around the streets of Jakarta, whereas fishermen would plan more wisely their trips to the sea to avoid inefficient use of fuels) and also sought other options such as loans or additional work. However, in contrast to the other professional groupings that have higher than average fuel consumption, fishermen 27 were concerned about frequent diesel shortages even more than higher prices. Even without a price increase, fishermen already have difficulties in obtaining diesel and frequently face shortages (see Box XX). Box 2: High vulnerability of fishermen in Indonesia Fishermen live in seaside areas that are often distant from gas stations. Since they cannot bring their boats to gas stations, they use plastic containers to store diesel, a practice that is prohibited by the government for other fuel consumers to avoid illegal stacking (penimbunan). As part of an effort to support a category with difficult access to public services, which started in 2010 and aims to cover 3 million fishermen in four years, the Indonesian Government has issued identity cards for fishermen so they are able to acquire diesel with containers. In an attempt to improve targeting, in February 2012 the government ruled that subsidized diesel could only be purchased by fishermen whose boat size does not exceed 30 gross tones. These solutions, however, are not sufficient and shortages often occur. Interviewed fishermen suspect that their designated diesel supply is flowing to other players who are willing to spend higher prices on subsidized diesel. They believe that they are competing for the diesel with industrial businesses, which are not eligible to buy subsidized fuel according to the law. The oil and gas downstream regulatory body BPH Migas has also mapped this pattern of illegal subsidized fuel transaction (Hilir Migas 2011). BPH Migas noted that in Cilincing, an area where one of the FGDs was undertaken, subsidized fuel often runs out by about the third week of each month, and for the rest of the month the fishermen were forced to buy fuel at higher, non- subsidized prices. 26 The kerosene subsidy for industrial sector was lifted in 2001, and in 2007 Indonesia started a program to encourage the use of gas to replace the use of kerosene in household activities, which was followed by a drastic limitation of subsidized kerosene distribution in the areas where the conversion program took place (Beacon and Lontoh, 2010). 27 Fishermen are particularly vulnerable to fuel price increases for two reasons: (1) their primary skill is not transferable; (2) there is no substitution to diesel, the primary fuel product for their boats In 2008, the International Labour Organization (ILO) recorded there were 1.8 million people who worked as fishermen in Indonesia. A report in April 2012 from the People’s Coalition on Fishery Justice (KIARA) estimated that the number of Indonesian fishermen fell from 4 million in 2004 to 2.2 million in 2012 (Lontoh et al., 2012). 16 Awareness of subsidy and perception The FGD discussions both in Indonesia and the Philippines suggest surprisingly low levels of awareness regarding existing subsidy policies. In both countries, consumer price subsidies are well established and because of the attempts at reform over the past years, also object of significant debate in the press. In Indonesia, the energy price increases of 2005 and 2008 where accompanied by compensation packages and communication campaigns. When asked about these subsidies, however, the majority of participants had limited awareness of government’s policy and, less surprisingly, very superficial understanding of it. For example, in Indonesia people were unaware of the different values between subsidized and market prices, despite the fact that information concerning the changes in market prices is relatively easy to access (e.g. through visiting a local gas station or by listening to government’s price policy announcements). Certain occupations, such as taxi drivers and industrial laborers, had higher levels of knowledge. One possible reason is that taxi drivers and industrial laborers have greater mobility which provides opportunities to interact with people from different backgrounds or outside their occupation groups. In contrast, farmers, fishermen and small business owners are more secluded. Moreover, the freedom of association and politicization of industrial laborers generally enhances their access information concerning highly politicized issues, such as fuel subsidy and social protection programs. Most participants believe that the government can control prices of food and fuel, and have no sense of “The government can lower the price the cost of this policy. In the Philippines, poor and non- of NFA just to pacify the angry poor alike were unfamiliar with the notion of subsidy in people. They lowered the price of general, and unaware that rice subsidies entailed NFA to pacify people. The budget allocations from government funds. Most people government has the power to lower took for granted the fact that NFA rice was cheaper any commodity in the market.” because the government sold it cheaply. People strongly feel that the government should make cheap and good quality rice widely available. This is tied to the belief that the government controls the price of rice and that the bad quality is rooted in the fact that the NFA stores the rice for a long time or that traders hoard it in warehouses Participants generally feel that the main beneficiaries of subsidies are not the poorest. In Indonesia, FGD participants were aware that fuel subsidies are intended to improve access to fuels for poorer consumers, but they found such a distinction to be relative. Many pointed out that government should not subsidize the rich and people using cars, especially luxury cars. In the Philippines, participants recognized that NFA rice served the very poor (though very few of the interviewed poor admitted to using it), but at the same time saw it as more of a source of profit for big businessmen, given that “They hoard the rice until the price becomes high and then they sell it. They get richer that way,” and for rice traders that mix subsidized rice with commercial rice and sell it as “pure” commercial rice at higher prices. The poor consider the subsidy system prone to abuse and corruption, leading to bad quality, higher prices and benefits to middlemen instead of the poor. Furthermore, subsidized products are not always accessible by those who need them the most. Another reason for the ineffectiveness of the NFA rice program in the Philippines appears to be the unavailability of the subsidized rice (see Table 3). This is particularly felt during lean or non-harvest 17 season in Albay, as well as in Surigao and Butuan, where it is available only in the main markets, requiring significant travel costs to purchase. In the NCR, many participants did not know where to purchase it. Respondents in Pampanga complain that affordable NFA rice is no longer available. Similarly, fishermen in Indonesia complain of shortages of diesel, and of subsidized shares destined to them going to industrial business (Box XX). Despite this perceived lack of effectiveness, subsidies are a certainty and people resist changes. All respondents in the Philippines objected to the removal of the NFA rice from the market even if they all object to the quality and price of the rice and claim that they do not use it. Nevertheless, they appreciate the continuous existence of the subsidy because they feel that it should be there when they are in dire straits or when there is a rice crisis. It is considered a lifeline by poor and non-poor alike. Even when it was explained that subsidizing the NFA rice could mean less allocation for other government programs, the poor respondents still wanted the program to be continued28. Similarly, in Indonesia, the priority for many was to improve and maintain subsidies that were important to their livelihoods. For instance, women in Makassar, saw that their daily spending on energy was mainly on cooking-gas rather than automotive fuel, and identified this fuel as a priority, while the motorcycle taxi drivers in Jakarta thought the opposite way. Most of the participants said that subsidy reform could only be accepted if the price of daily needs, especially food, would not be affected. What kind of support would be appreciated? In both countries, participants agreed the government had a responsibility to assist the poor in time of difficult economic times and viewed social protection mechanisms positively. The government is the primary institution from which they expect assistance, and people believe that it is the duty of the government to help the people. In the Philippines, it was clear that participants did not feel entitled to such support, but recognized their marginalization and understood that there are institutions that are able to help them access their needs. Government assistance is not a means for creating greater equity or a way of addressing injustice, but a way for them to meet their daily sustenance, health, education and housing needs. Table 4: Participants’ preferences on social protection programs - summary Livelihood Education Health Transfers Other Philippines *** ** ** * * Indonesia ** *** ** *** As can be expected, previous access to social protection programs and social services in general contributes to shape preferences for possible compensation programs in case of removal of subsidies. In the Philippines, sampled CCT beneficiaries clearly stood out in their preference for cash. They called for an increase in coverage, explaining that they can see the benefit and wish it to reach the poorest sectors. Respondents in the capital region, in Pampanga and farmers overall were clearer than average about what kind of government assistance they need for livelihood (i.e. loans, farm inputs, training, etc.). This appeared to be due to the fact that they had more livelihood opportunities concretely available to them and were therefore more aware of the possibilities. In Indonesia, Income levels made 28 Only the capital region respondents believed that the budget currently received by the NFA would be better used for fuel subsidies, for example. 18 a difference. Multiple explanations are possible, of course, but one significant reason for this is likely to be the correlation of income with access to services. The poor suffer from more significant cash constraints while at the same time having less access to services. Most poor groups expressed preferences for cash, while the non-poor tend to prefer services: education, healthcare and microcredit. Another interesting element to note is the strong and generalized preference for health and education support through universal health cards, and extension of free education (particularly in terms of additional and informal fees associated with students’ registration into a school and purchases of school material). It is reasonable to assume that such preference may be a reflection of a number of factors. First of all, notwithstanding free education, education related cost represent a very significant share of expenditure after food and energy for many poor households. This, combined with the high value attached to education in order to offer opportunities to their children, weighs heavily on people’s minds. Moreover, the preference for inclusive types of support is also representative of a tendency to “play it safe” in terms of making sure that something will be available, and that no connection or “qualification” will be required to access it. It is generally acceptable that the poorest receive more support and that there is a difference in support measures between the poor and the better off, but more inclusive programs are preferred. Most of the respondents in the Philippines, poor and non-poor, accept the necessity of some programs being targeted to the poorest-of-the-poor, as long as they are not implemented corruptly. However, they would prefer the implementation of programs that are more inclusive, such as cheap rice (as long as the program is improved), or universal health cards as in this way they feel certain that assistance programs would not exclude certain sectors and they will have a chance to access government resources 29. In Indonesia the discussion revolved around previous round of compensation, and there was no evidence to suggest that the participants felt the program should be broadened or narrowed. Those that had not benefitted from the BLT did not see themselves as rightful compensation recipients (vis-à-vis the compensation package) because they admitted that there were more people in their localities that lived in worse conditions. They naturally felt affected by higher fuel prices and cared about compensation, but they are looking for other programs suitable for their needs, such as education and healthcare. Stronger communities, with higher social capital, adapted programs to better suit their sense of equity and fairness in Indonesia. As expected, the BLT was the best-known compensation program, and it was also the subject of some criticism. Complaints had to do with the insufficient size of the transfer, targeting (BLT threatened community solidarity since its implementation lacked transparency and created dissatisfaction and jealousy among members of communities) and did not lead to long-term improvements in the community 30. However in some cases social capital helped people overcome the perceived weakness of compensation mechanisms. For example, some communities decided to rearrange the distribution mechanism of cash aid that was given for them. They set up a community assembly, identified recipients and the amount that each of them would receive in light of their own criteria of fairness. This practice required a solid coordination and leadership within the community. In contrast, the focus group discussions found that those with weaker social networks were more likely to 29 For example, while the CCT has been acknowledged by many participants as contributing significantly in alleviating the plight of poor households, the fact that it excludes many potential beneficiaries (as opposed to the NFA rice program’s inclusiveness) has caused some resentment against it 30 Industrial labors were particularly skeptical, and in large measure this was due to the linkage between the BLT program and fuel subsidy reform, as they saw it as a part of government’s effort to increase fuel price. 19 seek programs that they can access individually, reducing the need for interaction with people they do not know. These findings are consistent with what suggested by an analysis of household surveys in previous years 31. The main social protection programs (BLT, Raskin, and Askeskin) are generally viewed as mostly fair and transparent. Regarding targeting, a higher satisfaction with Raskin and Askeskin relative to BLT was reported, which could be related to the greater degree of community involvement in their targeting, but also to poor socialization of the BLT procedures. Moreover, the higher dissatisfaction with BLT targeting may indicate a link between complaints and the size and nature (cash) of the benefit, compared to Jamkesmas which was contingent on illness, and Raskin, where actual benefits are highly diluted, as well as the program’s coverage 32. Institutional capacity Setting the stage: WGI tables (voice, government effectiveness, political stability), WVS o Income inequality as a ROUGH proxy for social polarization. o the credibility of proposed options for reform and compensation measures, as influenced by:  Strength and credibility of institutions  pre-existence of SP mechanisms which can increase the likelihood of targeted compensation measures being effective Table 5: provides a snapshot of the dimensions described above, for our countries. Income Strength and Pre-existence of targeted SP i inequality credibility of mechanisms (main programs) ii institutions Large UCT (BLT), with 18.7 million households covered in 2010 round Indonesia (237 36.7 48th percentile Jamkesmas (health insurance for the Million) poor), with 18.2 million households covered in 2010 Flagship CCT program (4Ps), with 2.3 Philippines 42.9 51st percentile million households covered in 2012 (44 (92 Million) percent of poor households) I Gini index ii Government effectiveness, WGI Percentile rank among all countries, from 0 (lowest) to 100 (highest) 31 World Bank (2012b) 32 Ibid. 20 Figure 4: Government effectiveness: Indonesia, Philippines and comparator countries, 2010 In Indonesia, for example, the focus groups revealed that the issue at stake is not so much about participants’ distrust of government-related institutions, but more on the lack of information available with regard to reform policy and its accompanying compensation programs. While it was true that much of the policy debate around reform options and compensation packages received media attention, much of the information provided was, at best, fragmented and influenced by political interests. Many of the participants highlighted the need for better information concerning the scale of the aid provided to them, the schedules for the distribution and the way in which it would be distributed. Some participants also emphasized the importance of mechanism to relay complaints back to government, in order to channel their concerns about the implementation of the compensation packages distribution. Participants also wished for a more transparent and open to see the decision-making process at the local level. Media play an important role in promoting and leading debate on various public policy issues of importance both in Indonesia and the Philippines. In both countries, the debate over subsides and social protection (or compensation packages) is quite frequently reported on television, radio and newspapers. Thus, in addition to formal and informal socialization, public perception and buy-in of programs are also driven by the issues and sentiment the media convey in their coverage. A media analysis of the main social programs in Indonesia revealed that BLT in particular had the highest and most politicized media profile 33. In the Philippines, iT is interesting to note how much perceptions about 33 Indonesia media study, 2007-2009 [Mediatrac (2009) “Media analysis on program buy-in and socialization for BLT”, highlights report, background analysis for]– The most intensive coverage of BLT was in 2008, when rumors began of government plans to raise fuel prices and re-implement BLT as compensation, but the high coverage continued through program implementation in the second half of 2008, followed by politicized discussions in relation to the parliamentary and presidential electoral campaigns in early 2009. During the campaigns, many media articles on BLT were mostly politically related, as parties attempted to exploit BLT’s popularity as it provided short-term, just-in-time cash assistance for nearly a third of Indonesian 21 the rice subsidy program have been shaped from the media. For example, respondents from every sector know about the bad quality of rice because it is being hoarded by traders and stored in warehouses for a long time. They believe these things because there was media buzz about this for a while. 5. Conclusions The study also attempted to understand whether the following factors played a role in shaping these preferences: (i) previous experiences with targeted transfers or other benefits, and access to services (ii) income levels and utilization of subsidy, (iii) awareness and knowledge of subsidy mechanisms. Below we reflect on whether compensation packages including social protection measures and communication strategies play a role in influencing these perceptions once they are formed. The findings confirm, first of all, that where citizen have little trust in government, subsidies are a certainty, provide a well-known and reliable support. Reforming subsidies in context where government credibility is low can become complex. Citizen, even the poor who are not benefiting from the subsidy as much as the rich will not trust the government to spend the savings from subsidy cuts in ways that will favor them. For example, while the rice a subsidy in the Philippines is recognized by almost all respondents as having a very limited effectiveness for them; it serves a symbolic value that is important for the poor by providing a safety net. This preference may also have something to do with the fact that for most of the people, this is the only kind of national government assistance that they have ever been able to access. NFA rice represents the most concrete experience they have had of government assistance in their lives. The value of subsidized rice for society in the Philippines is very clear, and not dependent on income class, perceptions of the subsidy, or utilization of it. Subsidized rice stands for what the government can and must do for the marginalized. The rice subsidy exists as a safety net, actual or symbolic, of how when they fall short of their needs because of their inability to access resources from the mainstream systems, the government will fill the gaps. It also stands for this: the kind of government assistance that is tangible. At least it is a thing that is there and that they can access, as opposed to other forms of assistance that they can only access if they have a particular status or if they have special connections to the implementers. Secondly, It appears that the “pre-existence” of – and access to – social protection services can influence perceptions and choices regarding the compensation package. In the same way that institutions and systems affect the choice of a particular policy instrument such as the introduction of universal subsidies, they affect the feasibility of reform or changes in that policy. • Knowledge about the policy, trust in government institutes, and experience from previous implementation, social capital, economic status and fuel dependency are identifiable factors that contribute to the participants’ views on the compensation programs. households. Policy-related issues, including controversy on the effectiveness of programs as poverty reduction instruments, the need for complementary programs, and program exit strategies were also present, but gradually decreased and were replaced by coverage of implementation problems. 22 • In general the participants said that, should the government raise the price of fuel, it should also take steps to reduce the burden it would impose on vulnerable consumers. The compensation packages used in previous reforms and the broader social protection programs currently in place, where largely seen as inadequate. • One of the most common critiques of previous compensation packages in Indonesia was the lack of information and transparency on how they would be implemented and who should benefit. Participants also said that they lacked ways to provide feedback to government on the weaknesses of these programs, which stresses the importance of communication • In the Philippines, it appears that the main reason respondents have no clear and strong preferences regarding government assistance is that hardly any of the poor can concretely imagine what kind of programs the government can implement. This is especially true since the poor do not know if they can trust the government to deliver these possible or future benefits. Thirdly, awareness of subsidies and understanding of government’s efforts to support prices is low, so the first aim should be to inform citizens of the existence, intentions and scope of program measures that affect them. Effective communication can help citizen understand the burden of subsides and make a case for reform. Presenting evidence on why the current situation is not sustainable and on plans for a more effective use of public resources – including the introduction of compensation package – is the first step. Building awareness of the need for reform, as well as communicating a vision for a better country is important, but bridging the communication gap will be a challenge. The focus group discussion uncovered some important elements of concern, as well as beliefs and expectations that can help shape messages within a communication strategy. For example in Indonesia, the most common concern over fuel subsidy reform was the inflationary impact, especially its effects on food prices. Most participants feared fuel subsidy reform for this reason. Information campaigns showing clearly the beneficiaries of government programs may help. When asked directly about utility of universal subsidies to them in Indonesia, participants pointed to the fact that they do not benefit as much as others from specific policies to control prices (i.e. cooking oil versus fuel for cars, too low quality of rice), but the desired response is to attempt to improve the program rather than to imagine alternative compensation. In the Philippines, the government will face sizable much opposition from the public unless they can find ways to convince them that they are not being deprived of affordable rice. Therefore, the most important preparation for the potential removal of the subsidy would be to create a general awareness regarding the problems of NFA, the ineffectiveness of rice subsidies, but also about the cost for government budget, etc. Secondly, it would be important to raise awareness on government programs that are having a greater impact on poverty and livelihoods, such as the CCT or KALAHI [check]. Creating a buzz in the public mind to have open discussion regarding their concrete impact of these and how they could be expanded if only they were funded more. Finally, the government should be clear how it will manage the need for affordable rice and make this message clear. If this will be achieved through improved local rice production, or imports, then this must be shown in a credible way. Not all sources of information are equally trusted. In the Philippines, the poor in particular trust the local barangay sources of information and news the most, so the awareness raising should begin at this 23 level. But a supporting campaign should be present in the media as well in order to broaden the reach of the message and support it. Open questions, discuss • Value system and preferences for redistribution can change over time– is Asia different? Priors point to yes, very little literature and availability of survey data on value systems for systematic analysis (i.e. WVS only has very few data points, etc) • Keep in mind: Role of exogenous forces: high global prices, fiscal crisis, political crisis • What is the role of sequencing of reform? Quick vs. gradual approach to reform. Many questions remain unanswered and ultimately navigating the political risks of implementing consumer price subsidy reform will depend on a number of external factors. In order to develop a reform strategy, governments will need a sound evidence base and, in many cases, an open discussion around the merits of reform in order generate acceptance, if not consensus. Initially, a solid quantitative analysis and simulation of poverty impact to determine distribution of losses. A deep understanding of the social impact in order to better design compensation measures to protect form the rise in prices. The development of an appropriate communication and outreach strategy outlining the rationale for reform. Finally, timing and sequencing are key. 24 Annex Lessons from experiences with price subsidy reform and implementation of social protection measures This annex provides a brief overview of some examples (positive and negative) of subsidy reform strategies. There are no clear and easy criteria against which to measure success of reform in the case of removal of universal subsidies. In most cases, reform has been partial and setbacks are frequent. Nonetheless, these examples provide some valuable lessons learned in terms of how to go about a reform strategy once the political decision has been made. The likelihood of success (as measured by budget savings, social stability and implementation of more effective policies) is greatly increased if policies are developed on the basis of: • Solid quantitative analysis and simulation of poverty impact to determine distribution of losses and serious effort to understand the social impact. • Introduction of compensation measures to protect form the rise in prices. Social protection for the poor, compensation measures for specific groups • Development of an appropriate communication and outreach strategy outlining the rationale for reform • Timing and sequencing are key. Some external events will not be able to control for (but at the same time, they may create windows of opportunity) The information is provided in tables and as a quick summary, for more detail on energy reform in particular see Commander (2012) and Nikoloski (2012). On food subsidies, useful examples are Kelly (2009), Tuck and Lindert (1996) and also Alderman (2002) among others. 1. Timing - Good timing depends on context – know the situation: • Unexplained, one-time large changes in many goods may be too much • Gradual or staggered approaches can build resistance and stall reform, and also increase risk of hoarding/speculation Tunisia (food) Bolivia (energy) First attempt (1984) through price increase of FAILED: In 2010, the government increased many commodities at the same time was gasoline and diesel price suddenly and by large unsuccessful; then in 1989 – Government amounts (+73%, +57%, +83%). Price increases staggered price increases in summer months while were not communicated in advance, the process students were on vacation, also providing enough was highly politicized (President and VP directly time for advance communication involved in price setting, and process not transparent). Reform was unsuccessful and prices were reversed 25 Jordan (energy) Gradual increases between. In 2008 switched to monthly automatic pricing regime, based on formula that follows changes in international prices. Kerosene +76%, Diesel +76%, Gas +53% , etc 2. Compensation. Must be tailored to the impact and nature of the change on different groups • Poor people that are affected by high prices (mostly poor consumers of food and fuel commodities) • People most directly affected by the reform process, those for whom commodities represent a large share of production cost (farmers using fuel/fertilizer, taxi drivers, etc.) • Groups with a strong political voice, such as trade unions and other organized groups (civil servants; civil society, etc.) Indonesia (energy) Nigeria (energy) FAILED: no clear mitigation measures, only vague Untargeted cash transfers in 2005 (2 rounds, 35% promise to “use well the money saved” – of population covered, poor target ppl 28%) and expectations of corruption and lack of trust. then 2008. Reform had to be pushed through and partially Other measures included health insurance, failed education, rice subsidy, rural infrastructure, low interest rates for small entrepreneur Tunisia (food) Algeria (energy) Cost of subsidies from 4% to 2% of GDP FAILED: No clear direction, concern for social Measures for the poor - increase assistance impact and claim that mitigation measures could transfers, improve targeting of food distribution not be easily introduced (government lacks programs (limit to commodities consumed by administrative capacity and information); no poor – Percentage of food commodities reaching reforms implemented the poor going from 8 to 21%) Others: Students (direct aid for cafeterias), organized labor and middle class (raise construction workers’ wages, etc.) Jordan (energy) Mozambique (energy) Measures for the poor - cash transfers FAILED: Riots after fares of minibuses increased Others: increase in public sector wages and 35-50% following a 14% increase in the price of pensions diesel. Government agreed to supply diesel to Assistance for small farmers and key minibus operators at the old prices, but many services/SMEs, taxis/public transport that faced were not able to claim the subsidy because they direct price increase of inputs had not paid taxes. Iran 26 One of the most ambitious attempts at energy reform (large size of required adjustment, prices increased up to 20 times aiming at savings of 15% GDP) explicitly linking reform to compensation through transfers. Government committed to redistribute to households 50percent of fiscal savings initially, as well as a further 30 percent to firms. To build credibility, cash transfers were made to households prior to reform, and about 80 percent of population received the transfer.The effort was accompanied by an extensive public relations campaign (rationale for reform and budget cost; compensation package; guidance on environmental issues and energy efficiency). Mobilization of media, businessmen, academics, political and social leaders. Presence of President and top Government officials, emphasizing social inequality. What is important in the Iranian context however, is that, the communications and outreach strategy also emphasized the compensatory aspect of the reform 3. Effective communication can ease reform – a strong communication package should contain: • Reasons why is the current situation is not sustainable and not effective (not reaching its objectives) • Clear plans for more effective use of public resources • Compensation package 27 Ghana Kyrgyzstan Coordinated strategy (2005). Analytical work provided compensation package (minimum (PSIA), campaign kick-off with the Minister of income, pensions, cash transfers), but very poor Finance making a radio broadcast to explain the communication. Had to reverse reform, but unable situation and to announce mitigation measures. to eliminate new compensation measures Radio interviews with government officials, advertisements in newspapers. Creation of independent agency responsible for pricing (fuel prices +50%, new pricing formula - temporary) Tunisia Bolivia awareness campaign, reform as “strategic compensation measures announced by President necessity”; focus of the President’s independence Morales were not attractive and were not day speech; inequitable aspect highlighted as announced in advance . Lack of transparency found to be important for Tunisian society; following increases, newspapers continued to support “the need to do what is right/necessary”; polls (change in commodities, staggered increases) 28 Bibliography ASEAN and World Bank (2010). “Regional and Country Reports of the ASEAN Assessment on the Social Impact of the Global Financial Crisis,” Jakarta: ASEAN Secretariat Alderman, Harold (2002). “Subsidies as a Social Safety Net: Effectiveness and Challenges,” Social Protection Discussion Paper Series No. 0224 Arze del Granado, J., Coady, D. and Gillingham, R.(2010). “The Unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries,” IMF Working Paper WP/10/202, Washington, D.C. Beaton, C. and Lucky Lontoh, (2010). “Lessons Learned from Indonesia’s Attempts to Reform Fossil-Fuel Subsidies,” International Institute for Sustainable Development (IISD), Trade, Investment and Climate Change Series Bienen, H. and Gersovitz, M. (1986). “Consumer Subsidy Cuts, Violence, and Political Stability,” Comparative Politics, Vol. 19, No. 1, pp. 25-44 Birdsall, Nancy and Stephan Haggard (2000). “After the Crisis: The Social Contract and the Middle Class in East Asia,” Carnegie Endowment for International Peace, Washington, D.C. Birdsall, Nancy and Francis Fukuyama (2012). “The Post-Washington Consensus: Development after the Crisis - Working Paper 244,” accessed on the Center for Global Development website: http://www.cgdev.org/content/publications/detail/1424882 Coady, D., Grosh, M., and Hoddinot, J. (2004). “Targeting of Transfers in Developing Countries: Review of Lessons and experience,” Washington D.C. Coady, D. et al. (2006). “The Magnitude and Distribution of Fuel Subsidies: Evidence from Bolivia, Ghana, Jordan, Mali, and Sri Lanka,” IMF Working Paper WP/06/247, Washington, D.C. Coady, D. et al. (2008). “Evaluating Alternative Approaches to Poverty Alleviation: Rice Tariffs Versus Targeted Transfers in Madagascar,” IMF Working Paper WP/08/9, Washington, D.C. Coady et al. (2010a). “Petroleum Product Subsidies: Costly, Inequitable, and Rising,” IMF Staff Position Note, SPN/10/05, Washington, D.C. Coady et al. (2010b). “The unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries,” IMF Working Paper WP/10/202, Washington, D.C. Commander, Simon (2012). “A Practitioner’s Guide to the Political Economy of Reform,” mimeo Faustino, J. and Rood, S. (2012) “How Do You Get Reform in a Country Like the Philippines?”, Notes from the Field, Asia Foundation - http://asiafoundation.org/in-asia/authors/jaime-faustino/ 29 Farrar, Curtis (2000). “A REVIEW OF FOOD SUBSIDY RESEARCH AT IFPRI,”IFPRI IMPACT ASSESSMENT DISCUSSION PAPER NO. 12, 2000 Fattouh, Bassam and Laura El-Katiri (2012). “Energy Subsiides in the Arab World,” UNDP Arab Human Development Report Research Paper Series Fernandez, L. and Velarde, R. (2012). “Who Benefits from Social Assistance in the Philippines? Evidence from the Latest National Household Surveys,” World Bank Philippine Social Protection Note, No. 4 Gelbach, J. B. and Pritchett, L. (XXX), Is More for the Poor Less for the Poor? The Politics of Means- Tested Targeting Global Subsidies Initiative, IISD (2012). “Phasing Out Inefficient Fossil-Fuel Subsidies to Reduce Waste and Limit CO2 Emissions while Protecting the Poor,” Prepared for the APEC Energy Working Group Grosh, M et al. (2008). For Protection and Promotion: the Design and Implementation of Effective safety Nets. Washington D.C.: World Bank. Graham, Carol (2002). “Public Attitudes Matter: A Conceptual Frame for Accounting for Political Economy in Safety Nets and Social Assistance Policies,” World Bank Social Protection Discussion Paper Series No. 0233, Washington, D.C. Gutner, Tammi (1999). “The Political Economy of Food Subsidy Reform in Egypt,” IFPRI FCND Discussion Paper No. 77, Washington D.C. Institute for Philippine Culture (2012). “Final Report for the Social Protection and Rice Subsidy Project,” background paper International Monetary Fund (2008). “Fuel and Food Price Subsidies: Issues and Reform Options,” IMF Fiscal Department, Washington, D.C. IFPRI (1988). “Food Subsidies in Developing Countries: Costs, Benefits, and Policy Options,” edited by Per Pinstrup-Andersen Kelly, Elaine (2009). “Pricing for Prosperity: Consumer Price Subsidies in MENA and the Prospects for Reform,” University College London & Institute for Fiscal Studies Lontoh, L. et al. (2012). “Public Perceptions of Fuel Subsidy Reform in Indonesia,” International Institute for Sustainable Development, background paper Mediatrac “Media analysis on program buy-in and socialization for BLT, 2007-2009”, highlights report, background analysis for “Suggested Communication Strategy to Support Program Buy-in and Socialization for Poverty Reduction Programs” Muller, C., (2007) “Anti-Poverty Transfers without Riots in Tunisia,” Document de travail DIAL 30 Nikoloski, Zlatko (2012), “Energy Subsidies: A Practitioner’s Guide to the Political Economy of Reform. Country Narratives,” mimeo Niles, Kimberly (1999). “Economic Adjustment and Targeted Social Spending: The Role of Political Institutions (Indonesia, Mexico, and Ghana),” Background paper for World Development Report 2000/2001 U.K. Department for International Development, London; and World Bank, Washington, D.C. Pritchett Lant (2005). “The Political Economy of Targeted Safety Nets,” World Bank SP Discussion Paper No. 0501, Washington, D.C. Segal, Paul (2010), “Resource Rents, Redistribution, and Halving Global Poverty: The Resource Dividend,” World Development Vol. 39, No. 4, pp. 475–489, 2011 Tuck, L. and Lindert, K., (1996). From Universal Food Subsidies to a Self-targeted Program. A Case Study in Tunisian Reform,” World Bank Discussion Paper 351, Washington, D.C. Victor, David (2009). “The Politics of Fossil-Fuel Subsidies,” Global Subsidies Initiative World Bank, OECD, OPEC (2010). “A Roadmap for Phasing Out Fossil Fuel Subsidies,” The World Bank Group’s contribution to the 2nd IEA, World Bank. OECD and OPEC Joint Report, Prepared for submission to the G20 Summit Meeting in Seoul, 11-12 November 2010 World Bank (2011). 2008 Again? Indonesia Economic Quarterly March 2011 World Bank (2012a). Africa’s Pulse, Volume 5 World Bank (2012b). “Targeting Poor and Vulnerable Households in Indonesia,” World Bank, Jakarta World Bank (2012c). “Protecting Poor and Vulnerable Households in Indonesia,” World Bank, Jakarta 31