WPS6450 Policy Research Working Paper 6450 Drawing a Roadmap for Oil Pricing Reform Masami Kojima The World Bank Sustainable Energy Department Oil, Gas, and Mining Unit May 2013 Policy Research Working Paper 6450 Abstract In 2011, the median oil imports rose to 5 percent of The path to market-based pricing depends on the starting gross domestic product for net importers. In the past conditions: the gap between current and market-based several years, many governments have not passed through price levels, the level of public awareness about the the world oil price increases to consumers fully. As a sign extent of departure from market prices, the degree of of divergent pricing policies, the retail prices of gasoline, market concentration and competition in downstream diesel, and cooking gas in January 2013 varied by a factor oil, the subsidy delivery mechanism where subsidies are of 190, 250, and 70, respectively, across developing provided, the robustness of social service delivery, and countries. Policies to keep oil product prices low to the perceived credibility of the government. The evidence benefit the economy and protect the poor have had a presented in this paper suggests that pricing reform number of unintended negative consequences, including often does not have a clear end and should instead be flourishing corruption in the oil sector and entrenchment viewed as a continuous process of adjustment and search of monopoly operators or inefficient firms through which for mechanisms that take into account the country’s subsidies are channeled, stifling competition and raising institutions and political system, and the oil sector’s costs. market structure, infrastructure, and history. This paper is a product of the Oil, Gas, and Mining Unit, Sustainable Energy Department. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at mkojima@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Drawing a Roadmap for Oil Pricing Reform Masami Kojima 1 JEL codes: D40, H20, H30, Q48 Key words: subsidies, social protection, gasoline, diesel, kerosene, liquefied petroleum gas, transparency, competition, efficiency Sector Board: Energy and Mining 1 World Bank. The author is grateful to Osvaldo Victorino João of the International Monetary Fund; Saidamon Bodamaev, United Nations World Food Programme; and M. Khaliquzzaman, Veasna Bun, Salam Falah Almaroof, Keomanivone Phimmahasay, Bayartungalag Gursanjid, Olayinka Olufunke Babalola, Sylvie Nenonene, Amir Mokhtar Althibah, all of the World Bank, for providing price information for Angola, Tajikistan, Bangladesh, Cambodia, Iraq, the Lao People’s Democratic Republic, Mongolia, Nigeria, Togo, and the Republic of Yemen, respectively. This paper benefited from helpful comments by Punam Chuhan Pole, Marianne Fay, Paolo Verme, Ruslan Yemtsov, and Ariel Yepez, all of the World Bank. All remaining errors are the sole responsibility of the author. The findings, interpretations, and conclusions are the author’s own and should not be attributed to the World Bank, its Executive Board of Directors, or any of its member countries. 2 1. Context After averaging US$18 a barrel throughout the 1990s and rising to an average of US$27 in the first four years of the last decade, the price of crude oil began to soar in 2004. The prices of oil products followed similar trends, reaching historic highs in mid-2008, collapsing, then rising again. The prices of liquefied petroleum gas (LPG) reached a new high in 2012 before halving in the following months. Figure 1 shows the history of monthly average free-on-board (FOB) prices of gasoline, diesel, and LPG since 2004. Figure 1: International Prices of Gasoline, Diesel, and LPG since 2004 $1.40 Regular gasoline, $/liter Diesel, $/liter LPG, $/kg $1.20 $1.00 $/liter or kg $0.80 $0.60 $0.40 $0.20 $0.00 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Sources: Industry sources. Note: Gasoline and diesel prices are from Singapore. LPG prices are average Saudi Aramco contract prices for propane and butane. Kerosene prices are not shown because they track diesel prices closely. Earlier in history, the annual average spot price of three benchmark crudes (Brent, West Texas Intermediate, and Dubai Fatah), expressed in 2012 U.S. dollars, reached a historic high of US$89 a barrel in 1980. This record price was surpassed in 2008 when the annual average price soared to US$103 a barrel. The prices in 2011 and 2012 were slightly higher, averaging US$105 (Figure 2). Figure 2: Annual Average Spot Prices of Three Benchmark Crudes in 2012 U.S. Dollars 120 100 $/barrel 80 60 40 20 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Sources: World Bank 2013 for the prices of Brent, West Texas Intermediate, and Dubai Fatah; Federal Reserve Bank of St. Louis for the implicit price deflator for U.S. gross domestic product. 3 Government interventions in the oil industry and pricing were common around the world until three decades ago. World oil prices were opaque and were not uniform in the 1950s and 1960s. Vertically integrated companies dominated the oil sector, with opaque transfer pricing of crude oil to their own refineries. Transport costs were high, accounting for almost one-third of the landed cost of oil imported from Saudi Arabia to New York in the early 1970s. As a result, instead of one global market, there were many distinct regional markets each with its own prices (Stevens 2010). The oil price shock of 1973 prompted governments to regulate prices and domestic oil supply even in developed countries. The Emergency Petroleum Allocation Act of 1973 authorized the U.S. government to exercise price and allocation controls in the oil industry until 1981. The Canadian government regulated crude oil prices between 1974 and 1985. The effects of these price controls were not always positive. A U.S. Federal Trade Commission report (Harvey and Roush 1981) concluded that the federal price controls in the United States had led to adoption of higher cost production methods as well as sporadic shortages that limited the days filling stations were open, restricted amounts and days in which consumers could purchase gas (alternate days by odd-even license plate numbers), resulting in long waits and travel from station to station for fuel. Analysis of these limitations in “The Welfare Costs of Rationing by Waiting” (Deacon and Sonstelie 1989) suggested that the queues cost some consumers—those who were fully employed—more than they saved as a result of price regulation. Efforts to hoard gasoline and an increased hazard of car fires from gasoline stored in containers in vehicles were among the secondary effects (Bradley 1995). Although world oil prices doubled between 1978 and 1979 and remained twice the 1978 price level for the next three years, members of the Organisation for Economic Co-operation and Development (OECD) began to dismantle price regulation in the 1980s. Several factors contributed to this policy change. The commercialization of very large crude carriers reduced transport costs beginning in the late 1960s, and after 1973 the share of transport in the landed cost of imported oil declined precipitously, helping to equalize prices across different regional markets. The globalization of oil markets in the 1970s was further assisted by the rise of paper markets. Heating oil in 1978 became the first widely traded oil financial contract sold through a regulated exchange. By 1990, there were 10 active oil futures contracts trading worldwide (Medlock and Jaffe 2009). Today crude oil and oil products are commodities bought and sold in one international market; the only exception among the oil products considered in this paper is LPG, which is derived more from natural gas production than from oil refining and for which distinct regional markets still exist. Instead of regulating prices or allocation, governments in high-income OECD countries over the past three decades have focused on creating market conditions conducive to healthy competition, ensuring fair trading and acting on anti-competitive behavior. The UK Office of Fair Trading, for example, has acted twice in the last two years to boost the transparency of heating oil prices for end-users (OFT 2011, 2012). Even so, as recently as 2003, the government of Portugal was still setting price ceilings for gasoline and automotive diesel (IEA 2004), and the U.S. State of Hawaii, fearing a lack of adequate competition, imposed ceilings on wholesale gasoline prices from September 2005 to May 2006. Producer subsidies also continue, although at relatively low levels. For example, according to the OECD database on fossil fuel subsidies, between 2009 and 2011, the largest annual producer subsidies in the oil sector among its members were found in the United States, averaging US$2.4 billion a year, followed by US$1.4 4 billion in Canada (OECD 2013), although these particular producer subsidies have little effect on global or domestic oil prices. In contrast to the oil sector deregulation in high-income OECD countries today, many developing country governments continue to be involved in price regulation, which has been made difficult by the steep world price increases in the last decade. The nature of price regulation has varied from setting price ceilings for one or more fuels based on a market-based formula at one end of the spectrum to pan-territorial pricing with frozen retail prices of all key oil products for more than a decade at the other end. Price regulation in recent years has resulted in price subsidies in many countries. In addition to subsidies for oil products, electricity and natural gas are two other large sources of subsidies in the energy sector. However, of the three, only oil is a commodity. Electricity cannot be traded globally, and natural gas (because it is much more expensive to store and transport than oil products) has regional markets with sharply different prices. For example, in the first quarter of 2013, natural gas prices in Europe were more than triple those in the United States, and the prices in Asia were even higher, about four-and-a-half times the U.S. prices. Because natural monopolies exist in some or all segments of the supply chain, electricity and natural gas are subject to economic regulation, whereas oil products in many markets are subject only to competition and not to economic regulation other than anti-trust. Connection charges for consumers can be high for electricity and natural gas, sometimes several tens of times the monthly use charge, whereas there is no equivalent of connection charges for oil products other than LPG, for which the connection charge may be several times―not several tens of times―the cost of monthly consumption. Electricity and natural gas consumption by each user can be metered, but virtually no country has a system in place to meter all oil consumption by each consumer. These differences have significant implications for developing a pricing regime. Subsidized connection charges and subsidies based on monthly consumption (for which low consumption is a reasonable surrogate for low-income households) merit consideration in setting tariffs for electricity and natural gas, but they are far less applicable to subsidies for oil products (see “Targeted Price Subsidies for Liquid Fuels” in section 5). The price of oil is denominated in U.S. dollars and hence domestic prices are affected by currency fluctuations. Globally, the median increases from 2003 to 2012 in the nominal prices of gasoline, diesel, and LPG in local currency in 172 countries were 230, 250, and 170 percent, respectively. For gasoline and diesel, the median increases were lower than in the United States, because the U.S. dollar depreciated in slightly more than half the countries in the sample. International prices of gasoline, diesel, and kerosene were comparable in all major refining centers, but the LPG price movements depended strongly on the market, with North American prices much lower than those in the rest of the world in recent years (Kojima 2011). Therefore, price increases in the Americas tended to be smaller. (More details are provided in appendix A.) Expressed in real terms after accounting for changes in the consumer price index (CPI) in each country, the price increases between 2003 and 2011—the last year for which CPI data are available for most developing countries—were predictably much smaller. The median price increases for gasoline, diesel, and LPG were 110, 130, and 80 percent, respectively. Importantly, only a dozen countries saw their currencies depreciate against the U.S. dollar in real terms during the period. This meant that, for gasoline and diesel, most countries experienced smaller price increases in real terms than the United States. 5 Table 1 summarizes the results for diesel by income. High-income countries are further split by membership in the OECD. The results for individual countries and the summary statistics for gasoline and LPG are provided in appendix A. In nominal terms, high-income OECD countries had the lowest increases and low-income countries had the highest increases for all of the three oil products. In real terms, the highest price increases were found in high-income OECD countries for each of the three oil products. Despite the highest price increases in relative terms, none of these countries has reintroduced price controls. Lower-middle-income countries had the lowest price increases in real terms compared to other income groups, followed by low- income and upper-middle-income countries, which had approximately the same average as well as median price increases. Table 1: Increase in World diesel Prices in Local Currency, 2003–12 (nominal) and 2011 (real) Income: All Low Lower Upper High middle middle All Non- OECD OECD Percent increase in nominal prices in local currency, 2003–12 Minimum 110 230 110 110 140 180 140 Maximum 1,440 1,250 690 1,440 520 280 520 Median 240 330 260 280 230 240 230 No. of countries 172 32 48 47 45 14 31 Percent increase in real prices in local currency, 2003–11 Minimum –47 26 –47 19 66 89 66 Maximum 330 200 180 330 240 210 240 Median 130 130 110 130 150 140 150 No. of countries 160 31 44 43 42 11 31 Ratio with price increase in the United States in real terms, 2003–11 Minimum 0.2 0.5 0.2 0.4 0.6 0.7 0.6 Maximum 1.6 1.1 1.0 1.6 1.2 1.1 1.2 Median 0.8 0.9 0.8 0.8 0.9 0.9 0.9 Sources: Author’s calculations. The average oil price having more than doubled in real terms since 2003 has posed considerable political challenges to governments that administer oil product prices. In the period leading up to mid-2008, in the face of rapidly rising world oil prices, a number of the governments that had been keeping domestic prices artificially low seriously explored options for price reforms. The fiscal pressure to press on with reforms subsided briefly after the global price collapse in late 2008, but those governments that had done little were caught by rising prices again soon thereafter. Where governments have passed on some or all of world price increases to domestic markets, soaring prices have led to calls for governments to take action, ranging from providing greater safety nets for the poor and reducing fuel taxes, to ordering oil companies to lower prices and granting outright price subsidies, even if temporary. Many interlinked developments have affected costs, availability, and prices paid by end- users with respect to oil products in developing countries in the last several years: 6 • High oil prices have exacerbated the poor financial states of the national oil companies in some countries with price subsidies, leading to their inability to procure oil products on time, acute fuel shortages, and high black market prices. • Fuel price subsidies have increased incentives for diversion to black markets and smuggling, pushing up domestic prices markedly above the official prices. • Power shortages in a number of countries have increased demand for diesel for emergency power generation, causing diesel fuel shortages and higher diesel prices in some markets. Aside from chronic inefficiencies and financial trouble in a number of power markets, a growing cause of power shortage is declining rainfall that has reduced hydropower generation in East Africa and elsewhere. Diesel and fuel oil link the oil market to the power sector, with the pace of power sector reforms affecting oil demand and domestic oil product prices. • Piracy in the Gulf of Aden and Indian Ocean has increased insurance costs, led to shipping delays, and at times has caused fuel shortages in East Africa. • The challenges to the authorities mounted by citizens across the Middle East and North Africa since 2010 have stalled and sometimes reversed oil price reforms in several countries, against the backdrop of perceived declining state legitimacy. This paper is a part of a larger study that examines, from the point of view of consumers, issues related to oil prices in the downstream oil sector and other sectors where oil is an important input. It does not consider macro-level policies (such as monetary or exchange rate policy) or the impact of oil price changes on the macroeconomic performance of countries, nor does it discuss management of windfall income by large oil exporters and the long-term economic consequences of revenue management. This paper pulls together the findings from two recent working papers in the same study (Kojima 2012, 2013), which had built on earlier publications (Bacon and Kojima 2006, 2008a, 2008b; Kojima 2009a, 2009b). It covers developments in recent years with a focus on developing countries and proposes a menu of options for drawing a roadmap for pricing policy reform for oil products. It takes events since 2009 and looks at how recent oil price movements have affected countries’ vulnerability to world oil price increases, how governments in developing countries have adjusted domestic fuel prices in response, the consequences of the policy responses, other coping mechanisms to deal with high oil prices and price volatility, the roadblocks to reforming pricing policy, and how to overcome the roadblocks. It complements recent publications and a Web-based database on this topic—such as IMF (2013a, 2013b); Vagliasindi (2013); Arze del Granado, Coady, and Gillingham (2012); numerous studies by the Global Subsidies Initiative (GSI); and the country factsheet section of the International Fuel Price page of Energypedia recently set up by the German development agency, Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ)— by focusing particularly on the workings of the downstream oil sector. Section 2 provides background on changes in the global oil market since the 1970s, how these changes have driven pricing policies in developed countries, and where many developing countries are in the evolution of pricing policies. Section 3 takes a brief look at vulnerability to changes in oil prices in recent years. Section 4 compares domestic oil product prices in a sample of developing countries four years apart, in January of 2009 and 2013, to examine the degree of pass-through of world price increases to the domestic market. Section 5 provides an overview of 7 pricing policy options and likely consequences. Section 6 touches on complementary policies to help cope with high oil prices and price volatility. Section 7 concludes with observations and lessons. All medians in this paper are unweighted to avoid having large economies dominate the statistics. The paper uses the numerical notation adopted in the United States and defines a billion as a thousand million (109), in contrast to other parts of the world where a billion is a million million (1012). 2. Drivers for Price Controls and Subsidies With the globalization of oil markets, oil today is just like any other commodity, and government interventions in the pricing of oil products do not confer benefits if there is healthy and fair competition in the market. Importantly, nothing drives efficiency improvement as much as relentless competition, which also ensures that efficiency gains be passed onto consumers in the form of lower prices. The proper role of the government is to update regulations as needed; establish sound regulations and ensure their effective monitoring and enforcement; monitor for evidence of, and act upon, price collusion and other forms of anti-competitive behavior; provide market and other information—such as prices at different points along the supply chain; a breakdown of taxes and other charges imposed by different levels of the government; sales volumes; supply sources ; all regulations, ordinances, decrees, and any other decisions affecting the downstream sector; and instances of violation of regulations, including the names of the companies found guilty and the charges—so that suppliers and consumers can make informed decisions. In an emergency situation, such as natural disasters disrupting fuel supplies, a government may intervene with supply allocation, pricing, or both, on a temporary basis. For example, oil product prices in the Philippines have been deregulated since 1998, but when a series of typhoons devastated Luzon in 2009, the government imposed price ceilings between October 23 and November 16. In small markets vigorous competition may not be possible because of large economies of scale in fuel supply. If there is inadequate competition, governments may wish to set price ceilings to prevent marketers’ profit margins from rising too high. Even in such markets, it is important to encourage price competition—the level of which can be gauged from the degree of divergence of the prices on the market from the price ceilings—through information dissemination and other means. If the government deems price competition to be insufficient, it may set price ceilings, but such economic regulation need not involve subsidies. The objective of the pricing policy to address inadequate competition would be to establish price levels that would have prevailed in competitive markets. Depending on the trade status of a given fuel, the government may start with an import-parity or export-parity price equivalent, and add taxes, other government charges, costs, and reasonable profit margins. A point of departure from a competitive market is that profit margins tend to be static and always positive. The government may wish to impose cross-subsidization—for example, to make the prices to households more affordable—without incurring net subsidies. A scheme to impose cross-subsidization, however, frequently ends up accumulating large deficits, even when subsidies are not intended to do so. Cross-subsidization occurs particularly with LPG, used 8 widely for cooking, for which some governments impose the same unit price independently of quantities sold. There are greater economies of scale with LPG supply infrastructure than with other liquid fuels because LPG must be kept under pressure at all times—requiring metal containers throughout the supply chain—and it is typically sold in small cylinders to households. Absent government intervention, households would pay much higher unit prices than larger consumers because LPG in small cylinders is markedly more expensive (Kojima 2011). Forcing cross-subsidization, however, can distort market incentives, reduce competition, and compromise efficiency. The frequent rises and falls of world oil prices since 2004 have intensified suspicions about “rockets and feathers”—when oil prices rise, retail prices rise quickly like a rocket, but when oil prices fall, retail prices fall slowly like a feather (Bacon and Kojima 2010b). These and other such concerns have prompted some governments to tighten price control or launch investigations into price collusions. Kenya and Tanzania, having earlier deregulated domestic prices, began setting price ceilings by geographical location in 2010 and 2009, respectively. The U.S. State of Hawaii illustrates the challenges of monitoring price collusion in small markets. The state has a total population of 1.4 million, of which about 1 million live on the island of Oahu. At about 100,000 barrels per day (bpd) of oil product consumption (EIA 2013c), the market is not small by developing country standards. The state has two refineries and about a dozen oil marketers. Concerned about high oil product prices from a lack of adequate competition in this small island economy, the State in the past decade has experimented with price ceilings as well as a requirement for all marketers to report detailed information for transparency. In 2002, after the state had settled a price-fixing lawsuit with the two refiners for US$22 million, the legislature passed an act, which set ceilings on pre-tax wholesale and retail prices for regular gasoline. Before this act was enacted, it was replaced by Act 242, which set ceilings on wholesale prices of all grades of gasoline (the consumption of which was about 30,000 bpd at the time), linked to the average of the spot prices in New York, Los Angeles, and the U.S. Gulf Coast. Act 242 act was enacted in September 2005 and suspended eight months later in May 2006. Unfortunately, the timing of the start of the price ceilings coincided with high price volatility on the U.S. Gulf Coast from Hurricanes Katerina and Rita, and about the same time a major refinery in Los Angeles suffered a fire. Opponents objected to linking the state’s prices so closely to the three markets and expressed frustration with high price volatility from week to week; proponents argued that the scheme should have been given more time (PUC 2006). The price ceilings were replaced by the Petroleum Industry Monitoring, Analysis, and Reporting Program, intended to increase the transparency of pricing across the supply chain. The program required oil companies to submit weekly reports to the Public Utilities Commission on the volume, acquisition costs, margins, and sales prices at every stage in the supply chain— starting with crude oil acquisition by the refineries or oil product imports—by fuel type and consumer category on a confidential basis, and required the Commission to publish summary statistics within 14 days of receipt of these data. This program found that wholesale gasoline and diesel prices were higher than the price ceilings that would have been imposed under Act 242, and—perhaps reminiscent of rockets and feathers—retail margins for gasoline at filling stations rose during the sudden decline in global oil prices in 2008, while retail margins for diesel fuel were extremely high in several markets to mid-2009 (PUC 2010). For both price ceilings and price and margin monitoring, it was difficult to determine what would be a fair price, while the inability to analyze margins by company on account of confidentiality limited the usefulness of 9 the information published under the Petroleum Industry Monitoring, Analysis, and Reporting Program. As one lawmaker put it, “We do have some information now on margins, but we don’t know what profits are. Is there excessive profit or not? We need that information and we don’t have it” (AP 2007). The program was repealed in 2010. Many governments in developing countries keep retail prices of one or more oil products artificially low, thereby subsidizing them. Oil was considered a strategic good and price controls were common worldwide decades ago. While many governments today treat oil no differently from other commodities, a large number of governments have not moved from the era of price controls to deregulation. In some countries, consumers have become accustomed to relatively low prices, and in the last decade a confluence of food crisis, oil price shock, financial crisis, and, in some countries, political crisis has made it politically difficult to press on with price reforms. Reasons cited by the opponents of price reforms include protecting the economy from high world oil prices, curbing inflation, promoting economic development by making relatively cheap energy available, and making modern energy affordable to them. In major oil producing countries, low oil product prices are regarded as windfall dividends from oil revenue to which many citizens consider themselves entitled. Opponents of subsidy reduction often argue that higher oil prices hurt the poor more than the rich. This may be true when both the direct and indirect effects on household expenditures are considered. For example, a detailed study of households in Madagascar found that a 17- percent price rise would increase household expenditures by 2.1 percent for the bottom quintile, compared to 1.5 percent for the top. About three-fifths of the increase in expenditures was due to indirect effects, mostly through higher food prices. At the same time, because the poor spend far less on energy and food in absolute terms, the top quintile captured 40 percent of the total benefits of lower oil prices, whereas the bottom quintile captured only 9 percent (Andriamihaja and Vecchi 2008). Most poor do not own motorized vehicles or backup electricity generators and hence they may spend less than the rich on oil products not only in absolute terms but also as a percentage of total household expenditures. However, in countries where the poor rely on kerosene for lighting, they may be the largest spenders on oil products in terms of the percentage of total household expenditures, as shown by an analysis of household expenditure surveys in Bangladesh, Cambodia, and Uganda (Bacon, Bhattacharya, and Kojima 2010). What these observations suggest is that, even purely from the standpoint of equity, there is a cost associated with reforming oil price subsidies, and that cost should be weighed against the opportunity cost of maintaining the subsidies. Some may argue that a fuel price subsidy is justified when the fuel is a merit good, such as a clean cooking fuel that causes much less pollution and has health benefits. Given cash instead of price subsidies, households may use the cash for other purchases while reverting to traditional fuels that are cheaper or collected for free, with attendant air pollution and health problems. As section 5 shows, however, it is difficult to target price subsidies for oil products, and universal price subsidies have large leakages. Price subsidies for oil products are provided today because they are easier to administer than social services or social protection and are visible and politically popular. Although they benefit the politically powerful disproportionately (both legally and illegally through diversion and smuggling), subsidies can be billed as a pro-poor policy if necessary. All too often, if the government in power is attempting to reform universal price subsidies on the grounds that the bulk of the total subsidy goes to the rich, the opposition would nevertheless argue that the 10 government is intent on harming the poor who would starve (for lack of cooking fuel), freeze, or live in the dark (for lack of kerosene for lighting). Once introduced, subsidies tend to get locked in: the costs of organizing for subsidy reform are very high. In short, price subsidies are provided because they are easy to introduce and politically expedient (Victor 2009). The ease of subsidy administration may be compared to the ease of downstream taxation of oil products. Regression analysis indicates that the share of tax revenue derived from oil products rises with decreasing GDP per capita (Bacon 2001). Fuel taxation is important, especially for low-income countries, because the points of tax collection are fewer than for income tax or sales tax for general goods, a situation comparable to administering fuel subsidies. What is good for fuel taxation, however, is bad for fuel subsidies in that consumption of fuels as a group is strongly income elastic and only weakly price elastic (see Dahl 2012 for price and income elasticities for gasoline and diesel in 124 countries), ensuring buoyant government revenue as income rises and fuel tax rates are increased, but leading to rapidly growing subsidies if the pace of domestic price increases lags behind that of world price increases. The ease of administering price subsidies for oil products is all the more attractive in countries where the government has a poor track record of delivering social services. This is also the reason why high-income OECD countries, with well-developed administrative systems in place for delivering essential social services and social protection for the vulnerable, have done away with price subsidies for oil products. Ironically, the presence of subsidies often entrenches the conditions that make it difficult to deregulate the downstream sector. In some countries, such as India and Indonesia, subsidies are channeled exclusively through state-owned oil companies or refineries. This subsidy delivery mechanism inhibits the entry of other players, thereby discouraging competition. And where price levels themselves are controlled, no price competition is possible by definition. Absent a competitive market, prices cannot be deregulated overnight; interim price regulation may be necessary as a transition measure, but continuing government involvement in price regulation politicizes price adjustments. Ultimately, subsidies persist in large part because of weak institutional capacity. Establishing and enforcing sound regulations creating market conditions that promote healthy and fair competition in the downstream oil sector, delivering essential social services efficiently, protecting the poor and the vulnerable effectively, encouraging citizen participation in decision making, and responding to legitimate demands from citizens all require strong institutions, good governance, and capacity to deliver. Facing elections, political instability, or both, governments without such capacity use broad-spectrum subsidies—which are blunt instruments but are popular—especially when governments have few other administrative tools in their arsenal (Victor 2009). Kojima (2012) found that the greater the vulnerability of a country to changes in oil prices as consumers, the more likely the government would be to consider passing on price increases on the international market to domestic consumers rather than keeping domestic fuel prices artificially low. The next two sections review the findings related to vulnerability and passing through of price increases. 11 3. Oil Price Risks This paper takes one measure of vulnerability to changes in oil prices and defines it as the ratio of the value of the net volume of traded crude oil and oil products to GDP. This metric does not show the distributional or fiscal effects of a change in oil prices. By definition, vulnerability so defined is negative for net oil exporters and positive for net oil importers. For net importers, the larger the vulnerability index, the more vulnerable the country is to oil price increases. For net exporters, the higher the price of oil, the more negative the vulnerability index, everything else being equal. While a highly negative vulnerability index would signal large oil revenue flowing to the country, such an economy would suffer from an abrupt drop in the world oil price, as occurred in late 2008 and 2009. Therefore, any country with a large vulnerability index (with a positive sign for importers and negative sign for exporters) is highly vulnerable to oil price shocks, with net importers adversely affected by price increases and net exporters by price decreases. Changes in vulnerability can be linked to several factors through an identity that forms the basis for a decomposition analysis that allocates the change in vulnerability to changes in the different factors in the identity. A refined Laspeyres index enables decomposing vulnerability to a sum of consumption terms and production terms, which in turn consist of products of several factors (Kojima 2012, 4–5): Thus ∆V = consumption terms – production terms ≡ ∆C – ∆P = (oil price effect through consumption + oil share in energy effect + energy intensity effect ∆C oil intensity effect + real exchange rate effect ) – ∆P (oil price effect through output + oil production effect + effect of the inverse of current GDP in U.S. dollars). One advantage of this approach to decomposition is that the individual factors are additive. The sum of the effects of the oil share of energy and energy intensity of GDP is the effect of oil intensity of GDP. A comparison of the oil intensity effect with the oil price effect through consumption can indicate how much declining oil intensity has offset, or rising oil intensity has amplified, the adverse effects of increasing oil prices on consumers. Kojima (2012) provided the results of decomposition analysis for changes in vulnerability between 1999, which was chosen as the base year, and 2008 and 2009 as years with high and low world oil prices, respectively. The limitations of the methodology and of the underlying data are discussed in Kojima (2012). Between 1999 and 2008, when oil prices in U.S. dollars quadrupled in real terms, more than half the countries in each income category (low, lower-middle, upper -middle, and high) and more than two-thirds of all countries combined reduced the oil intensity of GDP. This occurred more as a result of reducing energy intensity than reducing the oil share of energy, although half the countries with declining oil intensity reduced both. High-income countries had the largest proportion of countries with declining oil intensity. About two-fifths of low-income and lower-middle-income countries saw their oil intensity rise during the same period. As expected, net oil importers were somewhat more likely 12 to reduce oil intensity than net exporters. Upper middle-income countries had the largest proportion of countries, one-third, in which declining oil intensity had offset one-fifth or more of the oil price effect through consumption. Among net oil exporters, nearly half had declining oil production between 1999 and 2008. Despite rising oil prices, oil exports as a share of GDP declined in one-fourth of the exporters—that is, the oil price effect through output had been offset by a fall in oil production combined with the effect of the inverse of GDP. Although the full set of data to carry out decomposition analysis is not yet available, preliminary data are available from the U.S. Energy Information Administration (EIA) to compute vulnerability for 2011, when the average price of oil reached the highest level in history in real terms; the average price in 2012 was slightly lower. The data available as of March 2013 were used to calculate vulnerability in 2011 as well as the change in vulnerability between 2009 and 2011, during which period the world price of oil in U.S. dollars increased by more than 60 percent in real terms. The EIA database is periodically revised and further revisions are expected; hence these results should be viewed as preliminary. Table 2 provides summary statistics by income, region, and oil trade status. The results by country can be found in appendix B. More than two-thirds of the countries in the sample reduced oil intensity between 2009 and 2011. The median vulnerability for net importers in 2011 was 5 percent of GDP, and that for net exporters was -16 percent. The median increase in vulnerability for net importers between 2009 and 2011 was 1.5 percent of GDP, and the median decrease in vulnerability for net exporters was 3.1 percent of GDP. Table 2: Vulnerability in 2011 and Change in Vulnerability 2009 to 2011 by Income, Region, and Oil Trade Status (% of GDP) Classification Oil trade status No. of countries Median V Median ∆V All income categories Importers 132 5.0 1.5 Exporters 40 -16 -3.1 Low income Importers 30 5.7 1.9 Exporters 3 -2.0 -3.3 Lower-middle income Importers 35 5.2 1.5 Exporters 12 -3.9 -2.5 Upper-middle income Importers 33 6.7 2.2 Exporters 14 -16 -2.8 High income Importers 35 3.4 1.2 Exporters 11 -32 -3.6 13 Classification Oil trade status No. of countries Median V Median ∆V Developing countries only East Asia and Pacific Importers 14 5.9 4.1 Exporters 3 -3.0 0.9 Europe and Central Asia Importers 17 5.0 1.8 Exporters 5 -15 -2.7 Latin America and the Caribbean Importers 23 8.0 5.9 Exporters 7 -2.7 -3.4 Middle East and North Africa Importers 9 5.4 6.4 Exporters 3 -31 -1.2 South Asia Importers 8 5.4 3.0 Exporters 0 n.a. n.a. Sub-Saharan Africa Importers 34 5.4 4.0 Exporters 11 -19 -3.8 Sources: Author’s calculations based on data from EIA (2013a) and WDI. Notes: Results by region exclude high-income countries. V = vulnerability; ∆V = change in vulnerability; n.a. = not applicable. Among net importers, high-income countries had the lowest median vulnerability and the lowest median increase in vulnerability. Upper-middle-income countries had the highest median vulnerability and increase in vulnerability. By region, Latin America and the Caribbean had the highest vulnerability and second highest increase in vulnerability. Among exporters, as expected, the Middle East and North Africa, with some of the major oil exporters, had the largest median vulnerability in magnitude. These countries also dominated the vulnerability index for high- income net oil exporters. Figure 3 illustrates the distribution of vulnerability in 2011 by oil trade status, and Figure 4 provides the distribution of the change in vulnerability between 2009 and 2011. Figure 3: Countries’ Vulnerability by Oil Trade Status in 2011 40 40 38 38 35 Net importers Net exporters 35 32 35 % of countries 30 30 25 25 20 17 20 18 15 15 15 10 10 8 5 2 5 0 0 0 to 2.5 2.5 to 5 5 to 10 10 to 20 Above 20 Below -30 -30 to -15 -15 to -5 -5 to 0 Vulnerability as % of GDP Vulnerability as % of GDP Source: Author’s calculations based on data from EIA 2013b. 14 Figure 4: Change in Vulnerability between 2009 and 2011 by Oil Trade Status 70 66 50 45 Net importers 45 Net exporters 60 40 % of countries 50 35 40 30 25 23 20 30 24 20 20 15 10 8 10 5 3 5 2 1 5 0 0 Below 0 0–2.5 2.5–5 5–7.5 7.5–10 Above 10 Below -10 -10 to -5 -5 to 0 0 to 5 Above 5% Vulnerability as % of GDP Vulnerability as % of GDP Source: Author’s calculations based on data from EIA (2013b). 4. Changes in End-User Oil Product Prices This study collected information on retail prices of gasoline, diesel, kerosene, and LPG in January 2012, July 2012, and January 2013, and calculated the degree of pass-through to consumers of world price increases since January 2009. Because prices at different stages in the supply chain are not available for most developing countries, a simplified methodology, described in the next paragraph, is used, yielding meaningful results only if the price difference between the two time periods is large. January 2009 was selected as the initial period for two reasons: first, the oil product prices in January 2009 were as low as in the latter half of 2004, and second, intervals of three to four years were deemed long enough to provide sufficient time to adjust to world price movements. The pass-through coefficient for each fuel is based on the ratio of the difference in domestic retail prices to the difference in free-on-board (FOB) prices of the same fuel in one of the four major refining centers relevant to the domestic market. As in Kojima (2013), this paper makes one adjustment to the methodology used in Kojima (2012): FOB prices in the denominator are taken from one month earlier to account for the findings in Meyler (2009). That detailed study of the transmission of world oil price increases in major 12 EU countries found that the increases were passed on to consumers within three to five weeks, or about a month. These countries can be safely assumed to be passing on the changes in world oil prices to their domestic markets in full. This paper therefore takes FOB benchmark prices from the previous month. As an illustration, the coefficients between January 2009 and January 2013 are calculated by taking the following ratio: (Retail fuel price 2013 – Retail fuel price 2009 ) . (Benchmark FOB fuel price 2012 – Benchmark FOB fuel price 2008 ) It is important to recognize the limitations of this simplified approach, as discussed at some length in Kojima (2012, 7–8), and not to regard a pass-through coefficient of 1.0 as the threshold level below which a market failed to pass through the increase in world oil prices; because of various factors, a fully deregulated and competitive market may pass through less than or more than 100 percent of the increase in the benchmark FOB prices. However, a pass- 15 through coefficient smaller than 0.8 in times of rising prices is likely to suggest attempts to keep domestic prices artificially low, or else domestic prices were very high initially and price stability was maintained against volatile world prices (as seems to be the case with LPG in Japan). The starting point for computing pass-through coefficients is collection of retail prices. Retail prices of gasoline, diesel, kerosene, and LPG were collected in January 2009, January 2012, July 2012, and January 2013. Figure 5 shows the distribution of the prices in January 2013. The prices in U.S. dollars are given in Table B.2 in appendix B. The prices of gasoline across the countries varied by a factor of 190, diesel by a factor of 250, kerosene by a factor of 10, and LPG by a factor of 74. The large differences between the minimum and maximum prices are because of exceptionally low prices of gasoline and diesel in República Bolivariana de Venezuela, and very low LPG prices in the Arab Republic of Egypt and República Bolivariana de Venezuela. Without these countries, the price differences are reduced to a factor of about 10. Of the four months for which retail prices were collected, these variations were largest in January 2013. When distribution curves from the three months in 2012 and 2013 are overlaid, they match between the 20th and 95th percentiles for gasoline, 25th−95th for diesel, virtually the entire range for kerosene, and 0−45th for LPG. World prices in U.S. dollars were comparable in the three designated months for gasoline, diesel, and kerosene, but the price of LPG fell by one-third in the Americas between January 2012 and January 2013. Figure 5: Distribution of Fuel Retail Prices in January 2013 in U.S. Dollars 4.50 4.00 Gasoline, $/liter 3.50 Diesel, $/liter Kerosene, $/liter $/liter or kg 3.00 LPG, $/kg 2.50 2.00 1.50 1.00 0.50 0.00 100 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Percentile Sources: See Table A1.2 in Kojima 2012. Table 3 provides median prices by income, oil trade status, and region. For gasoline and diesel, high-income OECD countries had the highest median prices on account of high fuel taxes and provision of no price subsidies. At the opposite end of the spectrum, lower-middle-income countries had the lowest median prices. As seen in all earlier publications, net oil exporters had much lower prices than net oil importers. Median gasoline prices were comparable across the regions, with the exception of the Middle East and North Africa, a region dominated by net oil exporters. The median prices of other fuels showed greater variation across the regions. 16 Table 3: Median Retail Prices in January 2013 in U.S. Dollars and Sample Size Category Gasoline Diesel No. of Kerosene No. of LPG No. of $/liter $/liter countries $/liter countries $/kg countries Income Low 1.34 1.22 15 0.98 12 1.77 10 Lower middle 1.12 1.02 24 0.84 17 0.80 22 Upper middle 1.21 1.08 24 1.15 12 1.04 18 High 1.97 1.52 8 a 1 a 2 Oil trade Net importers 1.28 1.22 51 0.99 32 1.24 35 Net exporters 0.87 0.93 20 0.56 9 0.61 15 Region (excluding high-income countries) East Asia and Pacific 1.25 1.03 9 1.11 4 1.20 8 Europe and Central Asia 1.16 1.23 4 a 1 a 1 Latin America and the Caribbean 1.24 1.14 16 1.18 8 1.01 14 Middle East and North Africa 0.67 0.44 8 0.41 4 0.40 7 South Asia 1.25 0.91 5 0.86 5 1.44 5 Sub-Saharan Africa 1.24 1.22 21 0.96 19 0.97 15 Sources: See Table A1.2 in Kojima (2012). Note: Median prices are followed by the number of countries in each category. Gasoline and diesel have the same number of sample countries. a. Too few countries in the sample for meaningful statistics. Price Movements between 2009 and 2013 Table B.3 in appendix B shows pass-through coefficients for January 2009−January 2012 (first time interval) and January 2009−January 2013 (last time interval). Although the results for the first time interval were previously published, the revised results take benchmark FOB prices from December 2008 and December 2011 to be consistent with the revised calculation methodology. Figure 6 shows the distribution of the coefficients for the four fuels for the last time interval, 2009–2013. Figure 6: Distribution of Pass-through Coefficients, January 2009–January 2013 300 Gasoline Pass-through coefficient, % 250 Diesel 200 Kerosene 150 LPG 100 50 0 -50 100 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Percentile Sources: Author’s calculations using data from sources cited in Table A1.2 in Kojima 2012. 17 When the distribution curves for the earlier end points (January 2012 and July 2012) are overlaid, the plots overlap except at the lower and upper ends of the range for gasoline and diesel. The distribution plots for kerosene, and especially LPG, in the second time interval (January 2009−July 2012) show divergence from those for the other two time intervals. The 50th percentile (that is, the median) did not reach 100 percent for any fuel in any of the three time intervals. The lowest median pass-through coefficient was 60 percent for LPG in the last time interval, and the highest was 95 percent for gasoline in the first time interval. Kojima (2012) showed the medium to strong correlation between the pass-through coefficients—that is, the degree of pass-through of world price increases tended to be comparable between gasoline and diesel, and so on. There was some correlation between pass- through coefficients and the country’s oil import status as well as the vulnerability index. The correlations with other parameters examined—currency appreciation, current account balance, months of import cover, GDP per capita, and the logarithm of retail prices in the starting month (January 2009)—were weak or nonexistent. Table 4 summarizes median pass-through coefficients by income, oil trade status, and region. Although there was no correlation between the pass-through and GDP per capita, when median coefficients were examined, they increased with income for gasoline and diesel. Among net importers, the median pass-through coefficients doubled going from low-income to upper- middle-income countries for gasoline, diesel, and kerosene. Sub-Saharan Africa had the second lowest coefficients after the Middle East and North Africa for all four fuels. Earlier, South Asia had the second lowest coefficients for gasoline and diesel during the first time interval and for gasoline and LPG during the second time interval. 18 Table 4: Median Pass-through Coefficients, January 2009–January 2013 Category Gasoline Diesel No. of Kerosene No. of LPG No. of countries countries countries Income Low 57 57 15 54 9 95 6 Lower middle 66 79 24 51 17 30 22 Upper middle 116 116 25 108 14 86 20 High 157 127 8 a 1 a 2 Oil trade Net importers 105 99 53 83 31 74 35 Low income 63 60 14 58 8 95 6 Lower-middle income 87 96 17 84 12 49 15 Upper-middle income 126 121 15 109 10 94 12 High income 169 130 7 a 1 a 2 Net exporters 57 42 19 17 10 21 15 Region (excluding high-income countries) East Asia & Pacific 99 99 9 59 4 57 8 Europe & Central Asia 111 145 4 a 1 a 1 Latin America & Caribbean 117 113 17 111 10 117 16 Middle East & North Africa 22 4 8 22 4 3 7 South Asia 65 72 5 88 5 84 5 Sub-Saharan Africa 57 49 21 42 16 31 11 Sources: Author’s calculations using data from sources cited in Table A1.2 in Kojima (2012). Note: Median coefficients are followed by the number of countries in each category. Gasoline and diesel have the same sample countries. Three figures are retained for coefficients exceeding 99% for formatting reasons only and are not intended to signify the number of significant figures. a. Too few countries in the sample. The foregoing results suggest that domestic prices have been lagging the price movements on the world market for more than half of the countries studied. The next section looks at the pricing policies that have led to this outcome. 5. Pricing Policy for Oil Products The wide range of prices in section 4 mirrors the diversity of approaches to government pricing policy for oil products. Each has its own history and reflects the state of oil supply and consumption in the country, as well as the political economy of the downstream oil sector and the state of social service delivery by the government. This section reviews different pricing policies followed by governments in developing countries. Unless indicated otherwise, the materials are drawn from Kojima (2013), which contains case studies of 65 countries. Range of pricing policies Various options have been adopted by governments to help reduce price volatility, keep prices low for one or more user categories, or do both. Measures include setting prices or price ceilings, providing targeted or universal price subsidies, setting up a fund for smoothing prices across time or freight cost equalization across the country for pan-territorial pricing, reducing 19 taxes, requiring oil companies to bear some or all subsidy costs, and imposing high export tariffs or export bans to keep domestic prices low. Combinations of these measures have enabled domestic prices to be frozen for months or even years at a time in dozens of countries. Countries that have frozen prices of gasoline, diesel, or both, for several months or longer in the past three years include Angola, Bangladesh, Bolivia, Cameroon, Côte d’Ivoire, Egypt, Ethiopia, Ghana, India, Indonesia, Islamic Republic of Iran, Iraq, Jordan, Kazakhstan, Madagascar, Malawi, Malaysia, Morocco, Mozambique, Nepal, Niger, Nigeria, Russian Federation, Rwanda, Sri Lanka, Syria, República Bolivariana de Venezuela, and the Republic of Yemen. Malawi is rare in that it suspended automatic price adjustments in 2004 but resumed them in June 2012. Jordan also resumed monthly price adjustments in December 2012 after suspending them at the end of 2010 except for LPG. More generally, about two-thirds of the study countries have kept domestic prices below market-based levels for one or more fuels in the past three years, subsidizing consumers. The government pays in every case directly or indirectly—through budgetary transfers, tax expenditures, or lower corporate tax collection—on account of financial losses suffered by oil companies. Many countries have universal price subsidies and quite a few have subsidies targeting certain consumer categories, but price subsidies targeting the poor are quite rare. Typology of pricing mechanisms Table 5 provides a typology of different pricing mechanisms, examples of countries where these mechanisms have been operating in recent years, and their respective advantages and potential disadvantages. The mechanisms are not mutually exclusive and a single market may exhibit features of several of them. At one end of the spectrum are those countries where fuel prices are deregulated, subject to anti-trust legislation. The Philippines deregulated pricing in 1998 and Turkey did so in 1989, and both have maintained deregulated fuel prices. The Government of the Philippines, however, has been more active in influencing fuel prices: it has negotiated diesel price discounts with oil companies for transport operators and at times required oil companies to justify price increases in writing. At the opposite end of the spectrum are those where pricing is ad hoc for one or more fuels with no clear rules for when and how much to adjust prices. Many countries fall under this category. In countries where prices are frozen for years at a time, when they are finally adjusted, price increases can be very large. As a result, price adjustments in some markets have come to be synonymous with large price increases, giving price reform a bad name. Table 5: Examples of Different Pricing Mechanisms Mechanism Advantages Potential problems Deregulate, subject to anti-trust Minimizes market distortions, no Downstream oil sector needs to be regulations (Philippines, Turkey) subsidies, gives appropriate price competitive or else consumers may be signals to consumers, competition can charged high prices; world price drive down costs and prices by driving volatility immediately transmitted inefficient firms out of business Adjust based on some link to world (1) Tracks world prices well while (1) World price volatility quickly prices and domestic costs: providing some measure of stability, transmitted (1) Frequent adjustments based on limits scope for mounting subsidies (2) World and domestic prices can be world prices averaged over 1–4 (2) Prices are more stable moving in the opposite direction, weeks (Dominican Republic, (3) Stability within the price band potentially large scope for mounting South Africa) (4) Large price changes avoided subsidies 20 Mechanism Advantages Potential problems (2) Frequent adjustments but based (3) If X is relatively large, potentially on world prices averaged over 1 large changes could occur when month or longer to smooth prices adjustments are made; possibility of (3) Adjustments made when world losses exceeding savings within the prices change by more than ±X% price band (Malawi, Togo) (4) Can lead to large subsidies unless (4) Price flotation within a price price bands are frequently adjusted band, changes smoothed outside (Chile for small and medium consumers, Peru) Steadily increase price at regular Each price increase is small and Could lose political commitment over time intervals until cost-recovery predictable and is not affected by time, and invite resentment if world levels are reached: sudden price spikes and collapses prices are falling; if the increases are (1) By a predetermined monetary regular but small compared to world amount (LPG for vehicles and price increases, subsidies could continue industry in Thailand) for years (Mexico) (2) By percentage (Mexico) Lower domestic prices by imposing No setting of domestic prices by Friction with oil companies; if large export tariffs or export quantity government, depoliticizes effective restrictions are sufficiently large, they restrictions such as export bans government price control for consumers create fuel shortages over the long run, (Argentina, Bolivia, diesel in China, because incentives to invest in oil Kazakhstan, Russian Federation) production and refining decline and investors move to other markets Stabilize prices through funds and Prices are smoothed. Mechanism (1) can Mechanism (1) is seldom, if ever, self- other means: be self-financing in principle. financing because of the arcsine law: (1) Based on the principle of over- Mechanism (3) helps deal with large even when prices move in a random recoveries offsetting under- price shocks while limiting the period of fashion, a period of under-recoveries recoveries (Ghana, India, Nepal, artificially low prices. can last a very long time, creating a Nigeria, Vietnam): domestic serious cash flow problem for a prices are kept higher than stabilization fund. Mechanism (2) is a market-based levels in times of tax expenditure, which is less low world prices, and over- transparent than subsidies financed out recoveries are saved; prices are of the budget because tax expenditures kept lower than market-based are not subject to annual budgetary levels in times of high world scrutiny by the parliament. Mechanism prices, and the savings from (3) faces political pressure to extend the over-recoveries are used to cover phaseout date repeatedly (Chile, Peru, the under-recoveries Thailand), potentially resulting in a (2) Tax adjustments (based on growing budgetary outlay or loans taken clearly defined rules in Chile for out by the fund. small and medium consumers, temporary diesel tax reduction in Thailand, frequent adjustments of import tariffs in Vietnam) (3) Establishing a temporary stabilization fund (Chile, Peru) with an initial transfer to cope with a sudden increase in world oil prices Cross-subsidize certain fuels (Ghana Possible to target net zero subsidy Net subsidies often exceed the zero and Nepal using gasoline to cross- threshold under political pressure. Inter- subsidize other fuels, Thailand using fuel price differences are amplified, its oil fund to cross-subsidize distorting incentives. 21 Mechanism Advantages Potential problems ethanol-gasoline blends today and LPG in the past) Deregulate prices for higher grade End subsidies to the rich, who are the Various market distortions as a result of fuels (Egypt, India, Indonesia, main consumers of higher-grade fuels, growing price differences for similar Malaysia) or certain fuels (diesel in or end subsidies to less politically fuels, fuel switching by users from Nigeria) sensitive fuels higher-grade to cheaper fuel, adulteration of higher-grade fuels with subsidized fuels Ration heavily subsidized fuels, Limit subsidies Diversion of rationed fuels to black charge higher prices outside the markets or smuggling quota (kerosene and LPG in India, gasoline and diesel in Iran) Set different prices depending on Limit subsidies and protect vulnerable Selling the same product at different user category (Angola, Argentina, groups prices invites corruption, starting with India, Indonesia, Iran, Kazakhstan, diversion to consumers who are not Malawi, Malaysia, Nepal, Nigeria, entitled to the subsidized fuel Panama, Russia, Thailand, Tunisia) Shift subsidy from one product to Subsidy for one product is eliminated Could lead to a growing subsidy on the another (kerosene-to-LPG product to which the subsidy is shifted conversion in Indonesia) (Indonesia) Set different pricing rules depending Limit subsidies to times of high world Unless price bands are adjusted from on world oil price ($80 and US$130 prices time to time, if world prices remain per barrel in China) high, subsidies could grow Establish the total subsidy envelope Limit the total subsidy bill Politically difficult to raise prices when for the fiscal year and adjust prices, money runs out volume, or both, accordingly Ad hoc: No clear rules; prices may Stable prices between changes Price changes can be large when be frozen for months or years at a adjustments are finally made, and time for one or more fuels (Angola, adjustments become synonymous with Bangladesh, Bolivia, Côte d'Ivoire, (large) price increases; tendency to Egypt, Gabon, India, Indonesia, delay price increases; lack of Islamic Republic of Iran, Iraq, predictability; likelihood of mounting Jordan, Malaysia, Morocco, subsidies; politicization of price Mozambique, Nepal, Niger, Nigeria, adjustments; hoarding in response to Syrian Arab Republic, Tunisia, rumors of imminent price increases and República Bolivariana de fuel shortages Venezuela) Source: Table A1.1 in Kojima (2013). The size of oil product subsidies can grow to be several percentage points of GDP, particularly in countries with ad hoc pricing. Two recent examples are subsidies of US$11.4 billion for gasoline and kerosene in Nigeria in 2011, amounting to 4.7 percent of GDP, and subsidy of US$22.6 billion for gasoline, diesel, kerosene, and LPG subsidies in Indonesia in 2012, amounting to 2.6 percent of GDP. Particularly in countries with ad hoc pricing, serious efforts to address subsidies are needed: these countries tend to face greater political challenges in reforming subsidies, fiscal outlays supporting subsidies are generally larger, and there tend to be greater market distortions, less transparency, and more commercial malpractice. Government control of fuel prices requires several decisions before criteria for setting prices can be selected. The first is whether to set price ceilings or price levels. The second is 22 where along the supply chain prices or price ceilings will be set. The third is whether prices or ceilings will be uniform throughout the country or will vary by location. The pros and cons of different options are outlined in Table 6. The inability to promote price competition is one of the disadvantages of controlling price levels. Table 6: Decision Parameters for Government-Controlled Prices Mechanism Advantages Potential problems Price ceilings Provide some scope for price competition; If price ceilings are too high, there is little divergence from ceilings suggests incentive to improve efficiency. If they are too emerging competition, with less need to low, some or all segments of downstream oil get the prices “exactly right” than may cease to be financially viable. controlling price levels Price levels Greater control Same as price ceilings, and in addition there is no scope for price competition. Control at retail Easy for consumers to check compliance More assumptions are needed to calculate prices than controlling prices upstream of retail. Compliance is more difficult to monitor because the number of points to be checked is the largest at retail. In the extreme, wholesale could be higher than retail (Kazakhstan). Control at More transparent because of greater If competition is inadequate, margins could grow wholesale or correlation with benchmark international and retail prices could be markedly higher than elsewhere upstream prices, easier to monitor compliance otherwise. If upstream prices are set too low, oil of retail because there are fewer points of sale. companies may try to recover losses by There is no need to try to estimate increasing retail prices to compensate. transport costs and retail margins throughout the country. Uniform prices Sense of national unity: one country, one Freight equalization introduces additional scope price. Easy for consumers to check for inefficiency as well as corruption. Cross- compliance. subsidization could increase to the point of making the cost of compliance unacceptably high. Pricing by location Costs are better reflected. Consumers in remote areas may compare themselves to those in major cities and feel a sense of injustice. If the cost of serving remote areas is too high, some may not be served. Source: Author. Cost of smoothing prices One of the common objectives for government control of prices is to smooth the international oil price volatility on the domestic market. There is an economic cost to price volatility, and a vast literature exists on whether it would make sense to intervene to try to reduce price volatility, particularly in agriculture. Keynes (1942) argued, “One of the greatest evils in international trade … [is] the wide and rapid fluctuations in the world prices of primary products... It must be the primary purpose of control to prevent these wide fluctuations.” One systematic approach to smoothing oil price volatility on the domestic market would be to set the domestic price by averaging past, and possibly futures, prices over a fixed period of time. Figure 7 shows a hypothetical smoothing scheme that uses moving averages of spot prices from the previous nine months as the starting point for constructing domestic prices. The plot uses heating oil (a type of diesel) in the New York Harbor, for which both spot prices and futures 23 contract prices (which are used in the next illustration) are available publicly. Smoothed prices significantly reduce price volatility, although world oil prices and domestic prices move in the opposite direction half the time―that is, domestic prices (black solid line in the figure) are rising when world prices (gray solid line) are falling, and vice versa. The plot also shows over- and under-recoveries in each month (dotted line). Figure 7: Comparison of Monthly Average Spot Prices and Nine-Month Average Prices for U.S. Heating Oil $1.20 $1.00 Spot price 9-month average price $0.80 Over/under-recovery $/liter $0.60 $0.40 $0.20 $0.00 -$0.20 -$0.40 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Source: EIA 2013b and author’s calculations. Table 7 summarizes the results of an analysis for different averaging periods—ranging from taking the average price from the previous month (third column) to averaging prices from the previous nine months and the next three months (last column)—using the historical spot and futures contract prices of U.S. heating oil since January 2003. The analysis is based on returns, which are differences in successive data entries. When logarithms of prices are taken, as in the table, returns approximate fractional changes in prices from month to month, so that multiplying returns by 100 approximates percentage changes; a return of 0.05, for example, represents a price change of about 5 percent. The hypothetical smoothing schemes maintain the same price for a month at a time and make monthly adjustments. Cumulative losses between January 2003 and March 2013 for smoothing prices for 10,000 bpd are also shown. 24 Table 7: Comparison of Returns and Cumulative Losses Using Different Averaging Periods for January 2003−March 2013 for Heating Oil Parameter No 1-0 3-0 6-0 9-0 1-1 3-3 6-3 9-3 averaging Minimum return -0.11 -0.27 -0.24 -0.16 -0.13 -0.23 -0.21 -0.16 -0.09 Maximum return 0.11 0.22 0.15 0.09 0.07 0.14 0.12 0.10 0.08 Coefficient of 48 7.1 5.1 3.7 2.9 5.9 4.9 3.6 2.8 variation Cumulative loss, US$ n.a. 27 58 102 147 5 13 57 103 million for 10,000 bpd Sources: EIA 2013b and author’s calculations. Note: Except for the column labeled “no averaging,” for which daily spot prices are used, all other prices are monthly average prices of spot and futures prices for no. 2 heating oil in the New York Harbor. The first number, “n,” in “n-m” stands for the number of past months for averaging prices, and the second the number, “m,” of months in the futures contracts. For example, “6-3” is the average of the past six months and the average of the next three months, taking the futures contract prices for one, two, and three months ahead. Returns = differences in logarithms of successive prices; coefficient of variation = ratio of standard deviation to average; n.a. = not applicable. The results show that, as expected, increasing the averaging period reduces price volatility: the coefficient of variation, which is one measure of volatility, is largest when daily prices are taken with no averaging, and smallest when the prices from 12 months—the previous nine months and the next three months—are averaged. Cumulative losses increase with decreasing volatility. Inclusion of future prices, however, reduces cumulative losses markedly without sacrificing the benefit of volatility reduction. The number of months when world and domestic prices move in the opposite direction—which could be politically awkward—is also reduced substantially when futures contract prices are included. For example, adding the futures prices for the next three months to the past nine months in the averaging basket reduces the percentage of the time when world and domestic prices move in the opposite direction from 50 to 30 percent. Although not shown, adjusting prices daily rather than monthly increases volatility and decreases the cumulative loss for the same averaging period. When there is no strong trend in the underlying international prices, smoothing schemes can operate without incurring an excessive fiscal burden for the government in the long run. Even in that case, however, the pattern of oil price changes can result in the scheme running a deficit for a lengthy period. The choice of moving average is important. The longer the moving average, the lower will be the volatility of the regulated price but the more vulnerable the scheme will be to periods of sustained price increases. There tends to be a trade-off between cumulative losses and the extent of smoothing. The incorporation of several futures prices into the moving average may help reduce volatility without incurring an additional fiscal burden. However, futures contract prices are available largely for crude oil, and much less for oil products. Based on the belief that the economic cost of volatility could be reduced with price stabilization and other price control mechanisms, many governments tried to dampen price volatility after the first oil shock of 1973. As the U.S. experience demonstrates, however, government interventions to smooth price volatility on the world market proved costly, and were soon abandoned in a number of countries. The economic cost of oil price volatility is little researched and difficult to quantify. Recent experience suggests that no smoothing scheme in the past decade has managed to achieve substantial smoothing without a corresponding fiscal burden or taking out large loans (Kojima 2013, Table A1.1). As such, there are large opportunity costs associated with government interventions targeting significant price smoothing, and the costs and 25 the perceived benefits of such smoothing schemes should be weighed against those of other alternative uses of the same financial resources. Further, as section 6 shows, if the objective is reducing periods of high oil prices and price hikes on the domestic market, there are non-fiscal means of doing so. In a detailed assessment of commodity price stabilization, Newbery and Stiglitz (1981) concluded, “The major result of our analysis is to question seriously the desirability of price stabilization schemes, both from the point of view of the producer and of the consumer.” Targeted price subsidies for liquid fuels Universal price subsidies for liquid fuels are widely recognized as being generally regressive and captured disproportionately by the rich and businesses (Coady et al. 2010; Bacon and Kojima 2010a, 75–79; Kojima 2011, 28–30; Arze del Granado, Coady, and Gillingham 2012). Large universal price subsidies lead to large fiscal burdens, losses for suppliers, or both; can decapitalize the downstream oil sector; and weaken governance by giving strong incentives for smuggling and other forms of commercial malpractice. These observations underscore the importance of moving away from universal price subsidies as rapidly as possible. As Table 5 shows, many governments have attempted to target price subsidies to limit the size of the subsidy bill and help primarily those who are vulnerable. Examples include subsidizing or offering larger subsidies for, and sometimes rationing, kerosene for household use in Angola, India, Indonesia, Malawi, Nigeria, and Tunisia; subsidizing LPG for the poor in Argentina and Panama and for household use in India, Indonesia, Morocco, Thailand, and Tunisia; offering price discounts or larger price subsidies for gasoline, diesel, or both for certain user categories in Ghana (fishing), Iran (certain vehicle categories for gasoline and diesel), Kazakhstan (farmers), Malaysia (certain vehicle categories and fishing boats), the Philippines (price discounts negotiated with oil companies for transport operators), and the Russian Federation (farmers). With the exception of premix in Ghana, which is gasoline pre-mixed with a small quantity of lubricant for two-stroke engines, the identical fuel is sold at different prices depending on the user in all other cases. There are segments of the energy sector where it is possible to design and implement targeted subsidies effectively. Electricity and natural gas are suited for this purpose because consumption by each consumer can be metered precisely, consumption is typically correlated with the user’s income level so that the poor can be relatively easily targeted, and a properly designed tariff structure can limit the amount of subsidized energy sold to the poor and achieve cost recovery through cross-subsidization. Importantly, it is difficult to divert electricity or natural gas delivered to the poor on a large scale to businesses. The difficulties associated with diversion help maintain the integrity of multi-tier pricing. In contrast, liquid fuels are very easy to store and transport, making it virtually impossible to prevent diversion. An attractive destination for diversion is arguably the automotive sector, and the most common diversion is that of kerosene for household use to the automotive diesel sector. Kerosene is not produced to diesel specifications, but can be added in large quantities to diesel without immediate detection by the driver. As a result, the level of diversion tends to increase with increasing price difference between kerosene and diesel. Apparent kerosene and diesel consumption in Nepal illustrates this point. Two types of kerosene were sold in Nepal until 2006, so-called quota kerosene, which was rationed, and open kerosene. Both were priced much lower than diesel. By 2006, the quota kerosene had been withdrawn from 26 the market, but kerosene continued to enjoy a significant price advantage over diesel. The government eliminated the price difference in November 2008, at which point apparent consumption of diesel rose sharply (Figure 8). Similarly, when Vietnam in March 2005 lowered the price of kerosene by D 600 ($0.038) a liter below that of diesel—against the historical difference of D 50 ($0.003)—the demand for kerosene immediately rose by 30–40 percent (Asia Pulse 2005). In response, the government equalized the prices of these two fuels in July 2005. Figure 8: Kerosene and diesel pricing and consumption in Nepal 700 14 600 12 500 10 Rupees per liter Million liters 400 8 300 6 200 4 100 2 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Kerosene consumption Diesel consumption Diesel-open kerosene price difference Diesel-quota kerosene price difference Source: Sales statistics from the Nepal Oil Corporation. Note: The years are fiscal years in Nepal, starting on July 16, and each year represents the year in which the fiscal year ended. For example, fiscal 2000 started on July 16, 1999 and ended on July 15, 2000. The price differences are averaged over each fiscal year. Multi-tier pricing of gasoline and diesel, targeting specific businesses, is equally problematic. One example is premix in Ghana, which is sold at less than one-third of the price of gasoline to help the fishing industry. Although premix is not an ideal diluent for gasoline, it is diverted on a large scale to adulterate gasoline, leading to frequent complaints of premix shortages (Modern Ghana 2012). Eliminating subsidies for high-octane gasoline which is used in high-performance cars typically owned by the rich and widening the price difference between different grades of gasoline similarly risks diversion of cheap gasoline to adulterate high-octane gasoline, or fuel switching from high-octane to lower-octane gasoline. Subsidized diesel set aside for farmers and fishing boats can also be diverted, either by suppliers before the fuel reaches the intended beneficiaries, or, especially in the case of fuel for fishing, by the beneficiaries themselves. Fishing boats are particularly suited for transporting large quantities of fuel and hence for diversion. It is for this reason that the government of Malaysia in August 2012 announced a new requirement for fishing boats whereby each catch would be verified by a government representative to ensure that the subsidized fuel is used for fishing and not for some other purpose (Borneo Post 2012). LPG is more difficult to store and transport because it is a gas at room temperature and atmospheric pressure and needs to be kept under pressure. Nevertheless, selling LPG in small cylinders for household use at lower unit prices than that in large cylinders invites diversion from households to businesses and industries, as in Indonesia and Thailand, or to the automotive sector, as in Ghana, if the price difference between gasoline and LPG is large enough. Even 27 without taking diversion into account, LPG price subsidies for households are regressive and benefit the rich disproportionately (Kojima 2011). In all cases, diversion makes less subsidized fuel available for the intended beneficiaries, enables black markets to flourish, and all too often forces the poor—especially in rural areas—to pay more. In the extreme, the prices paid can be markedly higher than what would have prevailed in a deregulated market. An example is kerosene in Nigeria, for which the official price has been ₦50 ($0.32) a liter for years but which has been known to cost as much as six-times the official price in some regions (All Africa 2011). Large enough financial incentives may attract well-organized mafias, who may use physical force and violence to silence investigators and whistle-blowers (Hindu 2011). To stem diversion, smuggling (into or out of the country, depending on price levels in neighboring countries), and tax evasion, some countries require the use of chemical markers or dyes (Kenya, Senegal, South Africa, Thailand, Togo, Tanzania, Turkey, Uganda), tagging of vehicles (Repúblicana Bolivariana de Venezuela), and even electronic tracking of delivery vans (Tanzania). But when the financial incentives are powerful, those engaged in commercial malpractice find ways to get around these restrictions over time. A clean-burning fuel for cooking and heating may be a merit good with health benefits not fully recognized by the users. One way of ensuring continuing consumption of such a merit good is to ear-mark cash transfers for the purchase of that specific fuel. In this way, market prices are not affected, and the poor receiving cash can use it only to buy the clean fuel. This is the principle behind bonogas cards for LPG in the Dominican Republic. In this regard, recent developments in biometric databases of personal information merit attention; presumably it would be more difficult to escape detection loaning cards with biometric information than loaning cards without. The government of India is proceeding with what it calls end-to-end computerization of the targeted public distribution system, which distributes subsidized food and kerosene. Some states are now piloting distribution of subsidized food and fuel using smart cards containing biometric information about beneficiary families (India 2013). Egypt is planning to introduce a nation-wide smart card system for rationing subsidized gasoline starting in July 2013, and eventually subsidized diesel except for agriculture and food industries (Reuters 2013). While it is always possible to sell the fuel purchased in this way and obtain cash, the transaction cost of doing so would presumably discourage resale on a large-scale. Unintended consequences of keeping domestic prices low Rising claims on the budget or growing tax expenditures, the public being shielded from sharp price rises, and weak price signals reducing incentives for fuel conservation are among the predictable consequences of policy of keeping domestic prices artificially low. Less expected are growing fuel shortages, protection of inefficient fuel suppliers, and rising demand with rising prices once subsidies are reduced or removed. Fuel shortages are common in a number of markets with low prices. In Egypt, LPG shortages in recent years have been so serious that several consumers have been killed in scuffles over LPG cylinders and the army and the police have provided armed guards for LPG transport. Shortages occur because of commercial malpractice, implicit rationing, and adverse effects of subsidies on the operational and financial performance of oil companies. Subsidized fuels are diverted to black markets or smuggled out of the country, reducing fuel availability in the formal sector. Even when there is no commercial malpractice, as in the United States in the 1970s, price 28 controls have led to sporadic fuel shortages and long queues, with costs to many consumers of queuing and other inconveniences outweighing the benefits of low prices, as described in section 2. Many governments cannot cover the full costs of subsidies for apparent demand, leading to implicit rationing of subsidized fuel. Where oil companies have to shoulder the costs of subsidies partially or fully, they may cut back on refining or export refined products as in China and Russia, imports, or both, and shut down filling stations—some 3,500 filling stations in Argentina closed between 2005 and 2010 because of poor profitability, and in India two private oil companies have closed thousands of filling stations, Essar Oil in 2005 and Reliance Petroleum in 2008 (Essar has since re-opened its stations). In extreme cases, financially cash- strapped oil companies cannot purchase crude oil for refining or import oil products, because banks are reluctant to issue letters of credit. Egypt, Senegal, and the Republic of Yemen have been in that situation in recent years. Years of subsidies can also decapitalize the downstream oil industry, leaving refineries and other infrastructure in disrepair and creating fuel shortages. Where there are state-owned oil companies, it is common to channel all subsidies through them. As a result, these companies end up attaining a monopoly or near-monopoly status. But years or even decades of not having to face any competition make these operators opaque and inefficient, raising costs. Higher costs in turn raise the cost of subsidies to the government. If there are frequent acute or widespread fuel shortages, it is even possible that removing subsidies and raising prices will free up supply constraints—such as by operating refineries at higher utilization rates and importing more oil products—to meet suppressed demand. In such circumstances, the elasticity of demand with respect to price may be positive rather than negative initially while consumption rises to match supply. Shift to market-based pricing The foregoing suggests that price subsidies for liquid fuels, however “targeted,” are not efficient means of achieving the stated objectives―controlling inflation, helping the poor, making modern energy services affordable, and assisting agriculture, transport, or fisheries―because of various adverse consequences stemming from poor targeting, diversion, and market distortions. Even smart cards to restrict fuel purchase to the intended beneficiaries have been known to be abused; the targeted subsidy for fishing boats in Malaysia discussed above is administered through smart cards. Upward price trends on the world oil market in recent years have meant that keeping domestic prices artificially low or smoothing volatility significantly has been costly to the government, oil companies, or both. Given the weight of evidence, governments should pursue policies to make the downstream oil sector competitive and deregulate it, and achieve the objectives of price controls through means other than exercising control over pricing and fuel allocation. For the purpose of helping the poor cope with high oil prices, the long-term goal should be to replace fuel price subsidies with effective social service delivery. The most efficient and least distorting approach is arguably to transfer cash as part of an integrated, comprehensive poverty alleviation program; government interventions to keep prices low for each good and service through sectoral subsidies are generally suboptimal. The path to price deregulation (or market-based pricing in small markets with inadequate competition) depends on the starting conditions. Several aspects of the starting conditions 29 influence the design and timing of price reform: (1) the gap between the current and market- based price levels, (2) the market structure, (3) the subsidy delivery mechanism where there is a price subsidy, and (4) the mechanism for delivering social protection. The rest of this section discusses these four aspects briefly followed by comments on the potential impact of price reform on the quality of service, the role of communication, and the transparency of the pricing policy. Restructuring of the downstream oil sector is beyond the scope of this paper. Bridging the price gap If current prices are low, they may need to be raised several-fold. Examples include all subsidized fuels in Repúblicana Bolivariana de Venezuela, Egypt, Indonesia, Iran, and Iraq; pre- mix in Ghana; kerosene in Angola, Bolivia, Ghana, India (for household use), and Nigeria; and LPG in Angola, Bolivia, Argentina (LPG for the poor), Morocco, Tunisia, and Panama (LPG for the poor). Implementing very large, one-off price increases is likely to be too disruptive and unlikely to find political acceptance. Large price increases also invite backlashes. A recent example is the price reform in the Islamic Republic of Iran in December 2010, in which fuel prices were at least doubled (for super gasoline) and increased by as much as 21-fold for diesel. That reform was remarkable in that such large price increases were not rolled back. Nevertheless, despite very large handouts to virtually the entire population for compensation at a cost that far exceeded the savings from the price increases, the parliament in 2012 amended the law reforming subsidies, turning future price floors in the law―intended to eliminate subsidies in five years―to price ceilings, thereby effectively entrenching subsidies (albeit smaller than at present). In these situations, the government needs to decide how long it should take to raise prices to market levels, and how large each price increase should be. This task is obviously made easier if price adjustments are started in times of low world oil prices, as the government of China did in January 2009, when it switched to a new regime requiring more frequent price adjustments based a 22-day running average price of a basket of crudes. Several of the countries above now requiring large price increases had retail prices that were close to market-based price levels in January 2009. Having missed that opportunity, and given that oil prices are unlikely to collapse as spectacularly as they did in 2008 and 2009, governments may consider raising prices in small increments regularly until market levels are reached, as the government of Thailand has begun to do with LPG prices. However, if the starting prices are very low―a cup of coffee in Repúblicana Bolivariana de Venezuela is a hundred times more expensive than a liter of premium gasoline―large initial price increases may be warranted and find political acceptance. Although rare, there are circumstances when overnight subsidy elimination may be feasible. One such case is a limited price subsidy program targeting only the poor, as with the LPG programs in Argentina and Panama. The Dominican Republic in September 2008 eliminated the LPG price subsidy and replaced it with bonagas cards issued to the poor; cash ear- marked for LPG purchase is transferred to bonagas cards as part of social protection. Another scenario where overnight subsidy elimination may make sense is where black market prices have become the de facto end-user prices and are close to what would have been market prices, as with subsidized kerosene in Nigeria. In these situations, it would be informative to conduct price and quantity surveys to identify how much consumers actually purchase and at what price, because the actual quantities purchased and prices paid determine the impact of higher official prices. 30 Market structure If there is a national oil company that dominates the downstream oil sector, competition is difficult to introduce, cost reduction from efficiency gains presents a greater challenge than otherwise, and price deregulation is not possible. Mimicking market conditions using an import- or export-parity pricing formula for ex-refinery prices and estimating costs and reasonable returns downstream of refining or importing would be an alternative until the sector is restructured and adequate competition is fostered. Even if there are several oil companies operating in a country, regional monopolies in different parts of the country could still exist. It is difficult to estimate what would be efficient costs in an uncompetitive market for the purpose of establishing a pricing formula, because the only readily available benchmark costs are FOB prices. Inefficiencies in importation, storage, transport, and retailing, and excess profits are difficult to estimate; doing so requires data collection from other markets that are competitive and similar in size and fuel consumption makeup. In large markets, where vigorous competition is possible in principle, domination by national oil companies may continue if subsidies are delivered through them―subsidies isolate the national oil companies from competition and entrench the monopoly structure. It is clear that prices cannot be deregulated in a highly concentrated market. Removal of subsidies and sector restructuring may need to proceed in parallel. Small markets pose special challenges, especially if they are isolated, as the experience in Hawaii in the last decade demonstrates. Even if there are several suppliers with no single company dominating, it may not be easy to gauge whether there is adequate price competition. As described earlier in this section, price ceilings may be used as a transition measure to avoid high prices from price collusion while giving some indication of the degree of competition in the market. Subsidy delivery How subsidies are delivered affects the steps for price reform and their sequencing. In some cases, crude oil (typically from domestic production) is provided at a heavy discount to domestic refineries, almost always state-owned. The crude oil price discounts also isolate the refineries from competition. It would not make sense to raise retail prices without adjusting crude oil prices, but that may mean that the budgetary outlay for subsidies increases for a while―previously oil producers were footing a portion of the subsidy bill out of the upstream rent, but now the government budget must cover the whole subsidy bill―before subsidies are eliminated completely. In other cases, price subsidies are channeled through state-owned oil companies. In both cases, subsidy delivery is closely linked to the market structure. Oil companies often end up bearing some of the costs of subsidies. Subsidies are seldom reimbursed on time, creating cash-flow problems even if they are fully reimbursed eventually. The government may freeze margins for a long time, effectively eroding their values. In countries such as Argentina, Brazil, and India, oil companies are being asked to bear some or all the costs directly. In such cases, oil companies would welcome elimination of subsidies. If there is much profiteering from subsidies—notably through diversion and smuggling but also from fraudulent claims on reimbursements, as in Nigeria in 2011—there is likely to be considerable political opposition to subsidy elimination, orchestrated through other stakeholders to disguise those who are benefitting through illegal means. Communication (see below) becomes all the more important. 31 Delivery of social protection One of the arguments for keeping prices artificially low is that, however large the leakage, such a policy does protect the poor. This argument is not invoked in high-income OECD countries where integrated social protection mechanisms already exist and there is no need to use price controls with attendant market distortions as social safety nets. There is nothing special about oil as a commodity, for which a single global market exists. Prices of commodities rise and fall all the time. It is inefficient to intervene in the market to protect the poor each time the price of something purchased by the poor increases. The government’s goal should be to develop a comprehensive and integrated social protection mechanism to help the vulnerable cope with rising prices of the basket of goods and services they purchase. In this respect, the highest priority for the government should be to work on two aspects of social protection: accurate identification of beneficiaries, updated from time to time, and a mechanism for efficient and timely delivery of cash and services to help pay for such essential needs as food, energy, health, education, transport, safe water, and sanitation. For the poor, the largest impact of higher prices of oil products may very well be through higher food prices, which is all the more reason to focus on the basket of goods they consume rather than on the prices of individual fuels. Developing a comprehensive social protection program takes time. But problems with fuel price controls and subsidies have also persisted for a long time, long enough to have developed a good social protection program by now. Governments accelerating efforts in this area today have a wealth of experience from other countries to draw from on the kind of safety programs that are suitable and how to develop such programs for maximum effectiveness (Grosh et al. 2008; Fiszbein et al. 2009). For countries with large subsidies, there could be a vicious cycle, whereby subsidies take up virtually the entire social spending by the government, which is left with few resources to begin to establish an effective social protection program. Breaking the logjam requires taking steps in parallel to chip away at reducing subsidies and shifting spending from subsidies to social safety nets. Higher oil prices may lower demand for labor, wages, or both in the short run, especially for unskilled workers (for example, see Keane and Prasad 1996). Such labor market adjustments could be painful for the poor and the near-poor. Unemployment benefits are essential to smooth labor market transitions, and yet they are underdeveloped or virtually nonexistent in rapidly urbanizing low-income countries. This highlights the importance of channeling more resources to strengthen unemployment policies in these countries. Good safety net programs require efficient and cost-effective systems for enrolling beneficiaries, making payments, and monitoring. Setting up a sound safety net program from scratch usually requires a minimum of four to six months, with a longer period for refinement. In the immediate term, the question is which existing programs can be scaled up while avoiding actions that will work against the medium- to long-run development of a sound social protection system. For example, if medium- or long-term plans envisage consolidating or closing specific programs, scaling these up as part of a short-run response will work against long-term reforms. In the near term, countries with sound and comprehensive safety nets already in place can increase the value of the benefits, coverage, or both. Examples include unconditional or conditional cash transfer programs. Targeted cash transfers of adequate coverage, generosity, and 32 quality are the best option. Increasing the benefits of social insurance programs not linked to earnings―such as social pensions, survivorship pensions, disability pensions, and unemployment benefits―can be helpful if they cover the poor. Near-cash instruments such as food stamps or vouchers for transport have slightly higher administrative costs than cash but can be politically popular. Fee waivers or vouchers for health and education help households maintain access to services even if households become poorer (World Bank 2008). Service quality Governments’ attempts to keep domestic prices artificially low can reduce the quality of service substantially in the downstream oil sector. Examples include long queues, flourishing black markets with much higher prices, adulterated fuels, and sometimes outright fuel outages with no fuels available at any price. Conversely, moving to market-based pricing could reduce or eliminate the financial incentives and causes of the low quality of service and transform the market. Some changes are instant; others take longer. Adulteration of automotive diesel with kerosene can be stopped instantly if the price of kerosene is raised to the price of diesel. Black markets may disappear if official prices are raised sufficiently; the size of a black market is inversely correlated with the price difference between the official fuel prices and prices on alternative markets, capped by import-parity prices. Fuel shortages may also disappear altogether if prices are raised to market levels. Short-selling is common in many markets, and is another way by which consumers pay more than the official prices. Raising official prices but clamping down on short-selling can reduce the adverse effects of price increases. Shortages may persist for a while in exceptional circumstances: if the supply infrastructure is in disrepair and needs significant refurbishment for normal operation or if suppliers are in financial distress after years of subsidies and need to be recapitalized. Normalization of fuel supply is readily observable and can be effected immediately, and hence a quick win for the government in price reforms. The more people are affected, the greater are the perceived benefits. Adroit handling of supply deficiencies therefore merits special attention. Consultation and communication Consultation and communication are essential to the transition to market-based pricing. Nearly all governments communicate about the regressive nature of fuel subsidies, often comparing the total subsidy bill with total spending on education or health. Governments usually go on to explain how the savings from subsidy reduction will be used more equitably. The content, the timing, and the manner of communication are all important. Resources should be devoted to developing an appropriate communication strategy. Analysis of winners and losers, effects on different segments of the economy as well as households, and how the savings from price reform might help offset the adverse effects of the reform should be undertaken ahead of time and communicated appropriately. There should be adequate time for the messages to be delivered and feedback to be received. Today, there is a variety of media for having a national conversation about price reform. An important new avenue is the electronic media, which give many people an opportunity to have their say and the government an opportunity to acknowledge their concerns. Aside from the internet, mobile phones have enabled citizens to voice their opinions. Although not related to fuel pricing, one such example is “Mali speaks” in a country where less than 3 33 percent of the population has access to the internet but two-thirds have mobile phones (Al Jazeera 2013). The government may also consider taking advantage of the fact that mainstream news stories are shared and commented on in the social media and set up a mechanism to listen and monitor the quantity and quality of the conversation to gauge where public opinions lie. Yet another option is to leverage existing organizations by formulating a questionnaire and asking civil society organizations located in remote areas to survey people in their areas. All these can complement more traditional forms of communication—TV, radio, newspapers, press releases, and stakeholder meetings. It is important to recognize that communication is only as effective as the credibility of the communicator. If the government has a poor track record of delivery, if there is a perception of widespread corruption, or if the subsidy delivery mechanism has large illegal leakages, the public rightly may question whether the savings from subsidy reduction will be used for their benefits, or whether the subsidies need to be reduced in the first place. As a former Nigerian trade union leader put it, the government can afford to continue giving a price subsidy, if only it can do so far more efficiently and without corruption (Guardian 2011). That is, however cleverly designed a communication strategy, it cannot compensate fully for perceived lack of credibility of the government in society at large. Lastly, working with the transport sector is important because higher transport costs are among the immediate consequences of higher fuel prices, readily noticed by the public. The timing of adjustments of controlled transport fares and tariffs often coincides with fuel price increases, with the result that the public assumes that the fare and tariff increases are solely due to the fuel price adjustments. When fuel prices are raised by 20 percent and transport tariffs increase by 30 percent, fuel price reforms are blamed, whereas the fuel price increases alone would account for only a fraction of the 30-percent increase. Communicating these aspects, and possibly reassessing the way transport fares and tariffs are set, are also important. Transparency One of the steps in fostering public acceptance of price reforms is to make fuel pricing policy transparent. Where there is a measure of price control, however limited, the government should establish, through regulations or laws, which agency is in charge of determining prices and the principles governing price control. The regulatory authority in charge of price control should be independent—it should not be a unit within the national oil company, for example— and the government should make public information about the process by which pricing principles and formulas are established, including all the parties involved. Costa Rica’s price- setting mechanism formally includes citizen participation, whereby objections can be lodged and considered by the regulatory authority. Careful consideration should be given to the degree of discretion embedded in the pricing policy. The greater the degree of discretion allowed, the more politicized fuel price setting is likely to be. Triggers for adjusting prices, the formula for the price buildup, and the frequency and rules for adjusting domestic costs—including profit margins, transport and storage, refining, and bottling of LPG—all determine the level of discretion accorded to the agency in charge of setting prices. Wide discretion may invite significant interference, but leaving virtually no flexibility may harm the downstream oil sector if certain cost estimates turn out to be too generous or too low. And, as the review of country experience in Kojima (2013) has shown, 34 political pressure and “extraordinary” world oil price movements can lead to suspension of automatic price adjustment mechanisms on a “temporary” basis with no clear timeline. The criteria for price adjustments, historical and current price calculations, and associated costs―for example, benchmark FOB prices in the relevant markets, exchange rates, various taxes and charges―should be available on the government Web site and through other forms of the media. It is important to disclose price controls at different points along the supply chain and the magnitude of under- and over-recoveries. Brazil, Ghana, and Thailand regularly report price structures but the subsidies at the refinery gate are not explicitly shown, giving a false impression to consumers. Where there is a price stabilization fund, flows in and out of the fund and the fund balance should be regularly reported; such reporting is rare. Equally important, information should be easily accessible, easy to follow, and timely. The pricing information should be consolidated in one place so that consumers are not forced to hunt for information that is scattered on different Web sites, some buried in hundreds of pages of government gazettes. In countries with price control, it should not be difficult to post information as soon as new prices are set. If prices are frozen, it is important to keep on reporting world oil prices in local currency units on a regular basis and the current as well as cumulative subsidies. 6. Complementary Policies to Cope with High Oil Prices Reducing consumption and reducing costs of supply are two means of responding to high oil prices. Reducing consumption There are two means of reducing consumption: fuel conservation and efficiency improvement, and fuel diversification. Fuel conservation Making cars, stoves, and heaters burning oil products more efficient; eliminating non- essential trips and tasks; and generally reducing the level of fuel-consuming activities reduce spending. Where electricity is generated in part from diesel or fuel oil, efficient lighting and other efficiency improvement measures for electricity can also conserve fuel. An important driver of fuel conservation is sending correct price signals to consumers. Keeping fuel prices and electricity tariffs artificially low militates against fuel conservation efforts. Apart from sending the right price signals, governments may in addition set efficiency standards, or encourage voluntary ones. Making information about efficiency performance widely available, such as through labeling, and verifying efficiency standards are also important. Fuel economy standards are still relatively rare in developing countries, in part because many do not manufacture vehicles and rely on imports instead. China has the longest history, having first introduced fuel economy standards in 2004. Power shortages can increase oil consumption as consumers turn to emergency diesel generation, which is an expensive and inefficient way of generating electricity. Significant diesel power generation often signals a sector in need of fundamental reform and plagued by chronic inefficiencies and financial trouble. In such cases, power sector performance is closely linked to the downstream oil sector. 35 Sometimes energy is wasted for lack of information or bad habits. Two classic examples are driving cars with improperly inflated tires and aggressive driving involving sudden acceleration and deceleration. Awareness-raising campaigns can help change behavior; some oil companies have actively promoted eco-friendly driving habits, such as Total Jordan (Trade Arabia 2011) and Shell’s FuelSave campaign (Shell 2013). Because transport fuel consumption is expected to grow rapidly in developing countries in the coming decades, it is particularly important to focus on curbing the growth of gasoline and diesel consumption. Fuel diversification In high-income countries with a well-functioning, market-based power sector, use of oil in power generation has been rapidly declining against the backdrop of widening gaps in cost between oil and its alternatives. Some small island economies and markets in other similar circumstances have little choice but to continue with oil power generation for the most part, and energy conservation remains the primary means of reducing consumption. In many other markets, optimizing investments in the power sector through long-term planning and improving the financial and operational performance of the power utilities through sector reforms can help substantially reduce reliance on power generation from oil. Gasoline and diesel account for about one-half of oil consumption in developing countries and substitution with natural gas or liquid biofuels can help with fuel diversification. Some countries have promoted compressed natural gas (CNG) as a substitute for automotive fuels, mainly gasoline and to a much lesser extent diesel. If there are abundant domestic gas supplies and an adequate gas distribution network in place, CNG may be an attractive option. The top five CNG markets in the world as of the end of 2011 were the Islamic Republic of Iran, Pakistan, Argentina, Brazil, and India (IANGV 2012). Relative availability of oil and gas can change over time—Pakistan and Argentina have been suffering from serious natural gas shortages in recent years and Argentina has even mounted a national program to shift away from CNG back to gasoline and diesel. Substitution of gasoline and diesel with bioethanol and biodiesel, respectively, is another way of reducing oil consumption. The cited objectives include substitution of oil with renewable energy, energy independence, and support for agriculture and rural development. Bioethanol is manufactured from sugarcane and starch crops (maize, wheat, cassava) by fermenting sugar. There are large economies of scale associated with ethanol production. Biodiesel, made from reacting methanol with plant oils (soy, palm, rapeseed, Jatropha curcas), is easy to make on a small scale, but there are economies of scale in making biodiesel that meets tight automotive diesel specifications. Argentina, Brazil, China, Colombia, Ethiopia, India, Jamaica, Malawi, Malaysia, Pakistan, Peru, the Philippines, Thailand, and Uruguay are among the study countries that have been blending biofuels for some time, and Mozambique began mandating blending in 2012. Brazil has the longest history, having authorized blending of 5 percent ethanol in gasoline in 1931 and mandating blending in 1938. In percentage terms, Brazil also has the highest share of ethanol displacing gasoline in the world. Factors contributing to the success of the ethanol program in Brazil have been described elsewhere (Kojima and Johnson 2005; Kojima, Mitchell, and Ward 2007). The economics of biofuels are not determined solely by the price of oil, but also by the price of the feedstock in the alternative agricultural crop market. The economics of ethanol from 36 sugarcane—technically the most efficient pathway to bioethanol production—have not been favorable (see Figure 2 in Kojima 2013). Despite high world oil prices, ethanol from sugarcane has not been economic since October 2008 (barring February 2013) because world sugar prices have soared. Taking world prices of sugar and gasoline since January 2000, ethanol from sugarcane has been economic only one-sixth of the time. Similarly, the economics of biodiesel have been challenging; the opportunity costs of one of the two principal feedstocks (plant oil, the other feedstock being methanol) have been higher than diesel prices before adding the capital, operating, and maintenance costs for the biodiesel manufacturing plant. The economics of biofuels are location- and feedstock-specific, and are more favorable where the costs of importing oil products are high and domestic production costs of biofuel feedstocks are low, such as a landlocked country importing oil and producing ethanol from molasses, a byproduct of sugar production. Malawi, which is landlocked, produces ethanol from molasses and fits this category. As a result of weak economics in many parts of the world, most liquid biofuel programs have required subsidies, mandates, or both. The subsidies can be substantial. In Thailand, the total reduction in taxes and charges on E10 for gasoline with 91 research octane number was more than US$2 per liter of ethanol blended in 2012 (EPPO 2013), or triple the FOB price of gasoline with 92 research octane number in Singapore. From the point of view of mitigating the adverse effects of high oil prices, a fuel diversification strategy requiring large subsidies to combat high oil prices raises questions about efficacy. Another concern is the competition for fertile land and water, which could contribute to food price increases. Concerns about food security have prompted China to ban grains for ethanol production, India to restrict ethanol production to molasses, and South Africa to exclude maize and Jatropha curcas (which is not edible but may nevertheless compete for water and land) from its national biofuel policy (South Africa 2007). Reducing costs of supply Hedging, bulk procurement, infrastructure modernization and expansion, reducing regulatory barriers, and price disclosure to promote price competition are among the means employed to reduce costs of supply. Hedging Hedging is a strategy intended to reduce the risk of adverse price movements—future oil prices rising for an oil purchaser, declining for an oil seller. Industries in which fuel costs make up a significant fraction of the total cost, such as aviation, have turned to hedging for years to mitigate fuel price volatility. Hedging risks were manageable until the last quarter of 2008, when world oil prices collapsed suddenly. Many airlines faced large losses from hedging in late 2008 and 2009, amounting to as much as US$1 billion for Air China. These losses have led to government interventions and investigations in some countries (China, Pakistan, and Vietnam), and even imprisonment of airline executives in Vietnam. The cost of hedging in a way that protects the purchaser against an oil price collapse is high, but not paying that extra price could lead to a situation similar to the one in Sri Lanka. Large hedging losses suffered by the national oil company in 2008 led to protracted litigation, the Supreme Court’s ordering the treasury to handle fuel imports between December 2008 and November 2009, and temporary suspension of the chairman of the national oil company. China restricted hedging in 2009 after large losses suffered by 68 state-owned companies. Pakistan’s 37 central bank in August 2009 rejected the finance ministry’s proposal to hedge oil prices for a year. Chile explicitly regulates hedging in its new price smoothing scheme for small and medium consumers. Ghana began to implement a government-led hedging strategy after the price volatility of 2008–2009, thereby avoiding the high risks associated with hedging in 2008 and 2009. Ghana’s cabinet in March 2010 approved a Commodity Price Risk Management Policy, paving the way for hedging oil products and crude oil. Ghana hedged half of its crude oil requirements in 2010−11. The International Monetary Fund reports that sizable hedging gains in the first half of 2011 allowed fuel subsidies to be covered through July (IMF 2012). Private companies engage in hedging at their own risk. In contrast, when state-owned companies engage in hedging and suffer large losses, the losses could become government liabilities. The costs and benefits of hedging are influenced by world oil price movements, which are extremely difficult to forecast. Rather than exploring hedging of oil product prices to reduce subsidies, governments would seem better off focusing on reforming the oil sector, its pricing policy, and social protection programs. Bulk procurement There are large economies of scale in importing crude oil and oil products. As mentioned in section 2, the introduction of very large crude carriers capable of carrying 2 million barrels slashed the shipping costs in the 1970s, and similarly importing products in large parcels reduce unit costs. Kenya, Mozambique, and Tanzania have instituted bulk procurement to take advantage of scale economy. Bulk procurement in these countries, however, has had mixed results. In Kenya, questions have been raised on and off about whether its bulk procurement system actually results in cost savings. In May 2011, the Ministry of Energy reportedly blamed high fuel costs on the refiner’s needing to go through a third party to procure crude oil and manipulation of the crude oil price by traders who alter the date when crude oil enters the country (BMI 2011). In Tanzania, questions have been raised about the same company winning the first three tenders in succession and importing gasoline failing to meet fuel specifications, which reportedly damaged vehicles (Tanzania Daily News 2012). A similar incident involving imports of contaminated gasoline prompted the government of Mozambique to give the state oil company a 51-percent stake in Imopetro in 2011 (All Africa 2012). While cost savings should be possible in principle, experience in East and Southern Africa suggests that reaping the benefits of scale economy through bulk procurement through government involvement is difficult. Infrastructure modernization and expansion Expanding the port capacity to enable large carriers to dock, increasing the speed of unloading, increasing fuel storage capacity, and enabling cheaper transport of oil products (by pipeline or rail rather than road transport) are some of the ways to lower costs and fuel prices. Shortening the time to unload imported fuels will slash demurrage charges and relieve port congestion, with positive economy-wide effects on both accounts. Some countries, especially those in Sub-Saharan Africa, suffer from inadequate fuel storage capacity, leading to fuel shortages and higher prices. Rail and pipelines tend to be underutilized due to the rundown state of the infrastructure, increasing transport costs. Both exacerbate the adverse effects of high world oil prices. 38 Table 8 shows examples of investments in infrastructure and related measures that can help reduce costs of supply. Hospitality arrangements and terminals that guarantee nondiscriminatory third-party access help minimize duplication of infrastructure, enable small players to take advantage of scale economy, and lower the barrier to entry, potentially enhancing competition. Port demurrage charges can add substantially to the supply costs; there are non- investment measures that can reduce demurrage charges significantly. Having sufficient storage capacity is important for avoiding fuel shortages. Table 8: Infrastructure Investments and Related Measures to Reduce Costs of Supply Objective Option Exploit economies of Hospitality arrangements, nondiscriminatory third-party access scale Adequate receiving and storage capacity for large vessels Closure of small refineries together with efficient import infrastructure Large refineries Minimize demurrage Rapid customs clearance charges Round-the-clock staffing by port authorities Adequate port receiving capacity Other means of reducing port congestion Improve transport Build or rehabilitate pipelines infrastructure Improve performance of freight rail Improve road conditions Minimize shortages Require minimum commercial and/or strategic stockholding in regulations Encourage hospitality and third-party access for pipelines and depots Ensure reasonable returns (through, for example, removal of universal price subsidies) to efficient operators to save enough earnings for construction of storage facility Source: Author. The operation of strategic reserves—as opposed to operational stocks—needs careful consideration. High-income OECD countries hold strategic reserves as an insurance against sudden supply shortages; these reserves have rarely been used. The costs of filling and running the strategic stocks that are rarely used may be affordable to high-income countries, but more challenging for developing countries. Countries now building strategic reserves may be considering their use for domestic price stabilization at times of unusually high prices without serious supply disruptions. Such a stabilization strategy is more likely to be cost-effective in a period of rising prices. Oil price movements are unpredictable, and correspondingly the likely costs and benefits of such a stabilization scheme are also very difficult to predict (Bacon and Kojima 2008a, chapter 6). Reducing regulatory barriers There may be regulations within and outside the oil sector that may be raising costs. As an example, the Federal Trade Commission testified before the state government of Hawaii that several regulatory features in the state are likely to be deterring competition and pushing up prices: Hawaii’s unusual landownership regime, which makes fee-simple land ownership difficult; rent control legislation, which limits the rent wholesalers can charge for leasing filling stations and which could reduce the number and quality of dealer-operated stations; and a law prohibiting encroachment, which limits the establishment of new filling stations (FTC 2003). 39 Consideration should be given to modifying regulations if there is a reasonable chance that their costs outweigh benefits. Price disclosure In markets with price competition, including those with price ceilings, price disclosure could aid price competition. The government can play an important role by collecting information on prices and making the information easily accessible. Among the most detailed, timely, and accessible is the online price database in Chile—mandated by a resolution issued in January 2012—which is also available on iPhone, BlackBerry, and Android. The database gives viewers the choice of displaying data in order of increasing or decreasing price and the address of each filling station, prices, and the date and time of the last price change (CNE 2013). The government of Guatemala highlights on its Web site the filling stations with the lowest prices in the Guatemala City Metropolitan Area with their addresses and street maps every week (Guatemala MEM 2013). The Nicaraguan Institute of Energy surveys about 75 filling stations in Managua every week and provides details on the three highest-priced and three lowest-priced filling stations (INE 2013). The Authority for Consumer Protection and Competition in Panama, proclaiming “an informed consumer has power,” collects and posts gasoline and diesel prices at filling stations every four weeks, highlighting those with low prices (Panama ACPC 2013). If only city-average or country-average prices are disclosed, if prices by filling station are published but one month later, or if prices are averaged by company across the country, the scope for promoting price competition is reduced. Aside from disclosing current prices by company and location, it is important to promulgate and enforce a rule that requires prices to be posted on display boards at readable heights that are clearly visible to drivers. For price competition to benefit consumers, there needs to be effective monitoring and enforcement of technical standards, or else a competitive fuel market could lead to partial or total product degradation, with a low-quality product (adulterated, mislabeled, or short-weighted) driving out a high-quality product. 7. Conclusions In competitive markets in which sound regulations are enforced, costs of price controls are likely to outweigh benefits. Enforcement of sound technical, environmental, health, and safety regulations is a goal in every market, whether or not there is adequate competition or price control. When enforcement is combined with competition, deregulation of the downstream oil sector including prices should be the goal. Absent adequate competition, economic regulation would be necessary, but the goal should be an automatic pricing mechanism that is market-based and that does not introduce market distortions, such as large price differences for similar fuels through inter-fuel tax differentials or subsidies. Significant government interventions in pricing and large departures from international prices are likely to signal other deep underlying problems: poor budget planning and execution, weak capacity to deliver essential social services, weak capacity to protect the poor and the vulnerable, rampant corruption in the oil sector and its opaqueness, a general lack of sense of security and fairness, or even challenges to the legitimacy of the government. These problems are interlinked, making it difficult to isolate and address the pricing policy for oil products alone. That does not mean that distortions in pricing policy cannot be addressed before all underlying 40 problems are resolved. Some problems, however, affect the likely success of reforming pricing policy more than others. International experience seems to point in particular to corruption in the oil sector and a lack of an adequate system for social protection. If corruption in the oil sector is legendary, widely perceived to reach the highest levels of the government and the national oil company, the public would not be sympathetic to the view that subsidies are hurting the government and oil companies. An opaque sector is rarely, if ever, efficient, and costs are generally higher. While this paper focuses on the downstream oil sector, very high rents in upstream oil (oil field development and production) are particularly susceptible to capture by powerful interest groups, and that colors the way the public perceives price subsidies for oil products in oil-producing countries. Corruption and opaqueness go hand in hand. Transparency throughout the supply chain greatly strengthens the government’s ability to carry through with price reform. In an oil- producing country, one way of making the sector more transparent is to join the Extractive Industries Transparency Initiative (EITI), which requires public disclosure of the revenues received by the government from oil producers, payments made by the oil companies to the government, and reconciliation of any material differences between the two (see eiti.org for more information). Refining dominated by one national oil company can also lead to inefficiencies and rent capture. It is important to take steps to break up its monopoly status, for example by breaking up the company; subjecting the company to competition from imports; facilitating new entry through third-party access and other means; and making information about costs, operations (such as refinery utilization rates), and prices public, and comparing them to international benchmarks. Although not necessarily related to corruption, the news of vertically integrated oil companies making record profits on the back of soaring oil prices, while consumers are being asked to pay more and more for oil products, also creates resentment and increases political pressure on governments to intervene. Such pressure has arisen worldwide in recent years, including in high-income OECD countries. Even in such cases, an appropriate policy response is not to start controlling fuel prices but to promote a tax system that is progressive, increasing government revenue markedly in times of high profits and using the extra revenue perhaps in areas that enjoy broad public support, such as delivery of essential social services, building human capital, and building infrastructure that will have economy-wide benefits. Different segments of society have vested interests in maintaining subsidies: • While no government welcomes the fiscal costs of subsidies, universal price subsidies benefit everyone and help buy political capital. Governments in competitive democracies may fear losing elections by acting decisively to reform subsidies; governments in other systems may fear challenges to their legitimacy. In hard economic times, when governments can least afford subsidies, demands for subsidies grow in many developing countries, especially those in which generous subsidies have been provided in the past. • Oil companies that are protected by subsidies—such as national oil companies through which all subsidies are channeled—may lose their market power if subsidies are eliminated. Many businesses, and especially energy-intensive industries, benefit from low oil product prices, unless the costs of fuel shortages or low fuel quality outweigh the 41 benefits of low official prices. These firms are most unlikely to be compensated fully for subsidy reform, and hence will oppose price increases. • If oil products are important in electricity generation, reducing subsidies for these feedstocks will immediately affect the financial viability of the power sector. Unless electricity tariffs allow for the immediate pass-through of the oil product price increases (which is not common), power utilities will have to bear financial losses until the next round of tariff adjustments. Many power utilities in developing countries are already bearing the financial costs of underpricing and power tariff reforms are as challenging as oil pricing reforms. Oil pricing reforms add to the challenges in power sector reforms. • The poor and the rich alike benefit from universal price subsidies. Gasoline and LPG price subsidies are particularly popular among the rich, because the consumption of both increases with income (except among those who are connected to natural gas). The rich are politically powerful and their opposition to price increases can be influential. • Those benefitting illegally from subsidies through smuggling and diversion may come from powerful groups, ranging from criminal elements to even high-ranking government officials and oil company executives. The greater the illegal gains, the more fiercely they will oppose attempts to reform subsidies, in the extreme even by resorting to violence. Designing and implementing effective safety nets is imperative. Fuel price increases from price reforms may call for compensating the poor not only for higher fuel prices but even more so for higher food and transport prices, and anything else that is oil-intensive to deliver. All countries fund safety net programs, but not all are well designed and demonstratively effective. Having a functional mechanism in place greatly strengthens the hands of the government in implementing price reforms, and helps exploit economies of scale by addressing a host of related and unrelated problems occurring in parallel, such as food crisis, financial crisis, and high unemployment. Importantly, a good social protection mechanism enables the government to move away from sectoral subsidies, which are inefficient, costly, and suboptimal. While the above steps are being taken in parallel, pricing mechanisms should be made more transparent and automatic based on a formula, accompanied by an inclusive process of consultation and communication. It would be worth devoting resources to developing a communication strategy that uses a variety of means, including Web-based conversations. It is important to highlight the costs of keeping domestic prices artificially low, not just in fiscal terms but in the form of commercial malpractice, acute fuel shortages, flourishing black markets, and deteriorating infrastructure. Comparing spending on fuel subsidies with that on health and education would be less effective if any one of the sectors involved is known to be plagued by corruption—for example in the procurement of textbooks or builders for construction of schools and clinics—which points back to the importance of addressing poor governance in the government. Communication that is opened up to everyone may very well elicit requests for price discounts from a range of stakeholders. Even if businesses can pass through all of price increases to consumers, demand for their goods and services will decline because of the response to higher prices. Requests for such targeted subsidies need to be managed, and it is easier to do so if there are strong institutions to handle erosion of the spending power of the vulnerable. 42 For any significant price reform, it would be good to have at least one benefit or compensatory measure that is immediate and readily noticeable, and publicized widely. Such benefits can signal to the public that the government is making honest efforts to use its budget expenditures to better the welfare of its citizens. An example that does not even add to the government’s budget is rapid disappearance of acute fuel shortages―shortages caused by diversion and smuggling can be eased immediately if the price reform substantially reduces the financial incentives for commercial malpractice. Examples that involve redirection of the government budget include cash given to the poor for LPG purchase through bonagas cards in the Dominican Republic and LPG vouchers through Bolsa Familiar (government social welfare program) in Brazil, both after elimination of LPG price subsidies; a doubling of the monthly allowance in Sri Lanka for kerosene to households without electricity from SL Rs 100 ($0.85) to SL Rs 200 ($1.71) in February 2012 after the kerosene price was increased by 49 percent; and temporary unconditional cash transfers in Indonesia in 2005–2006 and 2008–2009 and unconditional, essentially universal cash transfers in the Islamic Republic of Iran beginning in 2011, in both countries following very large price increases (Kojima 2013). In all examples except the last, a small portion of the savings from the price increases was used to implement the safety nets. The cash transfer schemes in Indonesia and the Islamic Republic of Iran have been widely credited for minimizing incentives for mounting opposition to price reforms and for avoiding reversals of price adjustments. Although there were implementation problems as one would expect, the first of the two transfer schemes in Indonesia was also remarkable in that the whole scheme was designed and deployed in just five months (World Bank 2012). Fuel price increases matter because so much of the economy depends on oil products. Reducing consumption and increasing income, both of which reduces the share of income spent on oil products, make oil less important and the government’s pricing policy less politically sensitive—the low share of total spending on oil due to high income is one of the reasons oil product pricing is not nearly as politically sensitive in high-income OECD countries. Globally, about half of oil consumption in 2010 was for road transport, and transport as a whole (road, air, and sea) accounted for about 70 percent of oil consumption, followed by industry (8 percent), power generation (7 percent) and households (6 percent) (IEA 2012). Significant scope exists for reducing oil consumption in every country. There are barriers to fuel conservation, starting with fuel prices that are kept artificially low. Achieving fuel conservation requires changes, some short term—such as behavioral change—and others long term. Fuel consumption for road transport can reduced immediately through behavioral changes. They include fuel-efficient driving habits and vehicle maintenance practice, taking public transport or walking more, combining trips to reduce kilometers traveled, and minimizing nonessential trips. Drivers and even transport operators are not necessarily aware of fuel-efficient driving habits or fuel-conserving vehicle maintenance practice; public education and information campaigns can help. Government policies that promote fuel conservation include parking policies to discourage private car use, traffic management to smooth traffic flows, and restricting speed limits on highways to about 80 kilometers per hour. Over the long run, urban planning that promotes public transport and minimizes distances traveled to work, schools and shops, and improving the infrastructure for public transport that is convenient, affordable, and attractive can reduce oil consumption (GIZ 2012). Power sector reforms can slash oil use in many markets by providing adequate and reliable electricity, thereby eliminating the need for captive or emergency diesel power generation. Where oil is used in power generation, increasing the efficiency of generation, 43 minimizing technical losses in transmission and distribution, and increasing energy efficiency of power use can all contribute to lower oil consumption. Efficient lighting is a proven way of reducing power consumption markedly; some countries have actively pursued, and even legislated, efficient lighting. This study shows that fuel pricing is interlinked with an array of issues, starting in the oil sector and extending to power sector performance, port infrastructure, the transport sector, social service delivery, safety nets, and governance in budget planning and execution. Price reforms are difficult because there is a self-reinforcing mechanism whereby subsidies reduce the fiscal resources available to strengthen safety nets and deter the development of a competitive, efficient market—by giving no scope for price competition if prices themselves are controlled, channeling subsidies through state-owned companies and inhibiting new entry, and protecting inefficient refineries and oil marketers, to mention a few examples. As with any other reform, no model is universally applicable, and appropriate solutions are highly context-specific, necessitating varying solutions over time even within a given country. Events in the country outside of the oil sector—such as food crisis or a publicized scandal in the government—have significant effects on the public’s willingness to go along with price reform for oil products. It is important to analyze the starting conditions to develop a sequence of steps that are specific to country circumstances, accompanied by an inclusive process of consultation and communication. Table 9 provides a list of issues for consideration in the process of developing a roadmap for price reform. Table 9: Considerations for Drawing a Roadmap for Price reform Area Consideration Specific issues Starting Gap between current price and Is the gap becoming an important fiscal concern? If the gap is conditions in market-based price levels large, what is the time period for bridging the gap that would likely the oil sector find public acceptance, and how large a price increase at a time? If the gap is relatively small, how soon could a formula-based market pricing mechanism be adopted or resumed? How prices are set and who Price levels or price ceilings? At retail or elsewhere? Pan-territorial sets them pricing or geographical variation? Is there a formula for setting prices? Is the formula being followed, or has it been suspended in practice? Is there an agency in charge of setting prices, or is the decision to change prices made by different political groups depending on the state of politics at the time? Who determines under- Is the size of under-recoveries based on self-reporting of costs by recoveries, who pays for oil companies, or some international benchmarks? Is the them, and how government reimbursing oil companies fully or partially, and in a timely manner or often with long delays? Is the downstream oil infrastructure languishing for lack of investment because of price control? Are the reimbursements for under-recoveries channeled through state-owned oil companies? Is the government’s share of subsidies clearly shown in the budget, or are there off-budget transfers of funds, obscuring the magnitude of the subsidies? Are tax expenditures used to cover under-recoveries? Subsidies should be made transparent and easy to track. Competition in the market How concentrated is the market at refining/import, wholesale, and retail? Is there a national oil company that dominates the market? Are inefficient refineries or state-owned oil companies protected by tariffs, subsidy delivery mechanisms, or other means? Is hospitality or third-party access encouraged to facilitate new entry and avoid 44 Area Consideration Specific issues duplication of infrastructure? Subsidies in place Universal or targeted subsidies? Who is targeted and how? Are subsidized fuels rationed, and if so, how? Who uses which fuel and for Is there widespread use of gasoline or diesel for stand-by power what purpose generation? If so, higher fuel prices could threaten access to power. Is gasoline used primarily by the better-off, or is there widespread use of gasoline in motorbikes by small businesses and lower- middle class families so that gasoline is not fuel only of the rich? Is kerosene or LPG widely used for cooking? There is more resistance to raising cooking fuel prices that affect a majority of households. Is there widespread use of kerosene for lighting? If so, pro-poor arguments could be used to argue against raising kerosene prices without compensation. Commercial malpractice Are there flourishing black markets? Are actual prices paid by consumers markedly higher than official prices? If so, raising official prices would have much less adverse impact. Is short- selling routine? If so, enforcing rules against short-selling would lower the effective price increases when official prices are raised. Is there smuggling? Is there diversion of subsidized fuels to consumers who are not eligible? Is there adulteration of higher- priced fuels with subsidized fuels? Perception of the oil sector Is it considered opaque, corrupt, politically well-connected, or a state within a state? Are there scandals to do with large leakages in subsidy delivery? If so, raising prices could be difficult if the public is angry about corruption; on the other hand, it might be possible to persuade the public that higher prices get to the source of the corruption and help stamp it out. If a large oil producer, is the country an EITI member? Social Safety nets Is there an up-to-date database of beneficiaries? Is there an protection administrative system in place to deliver benefits? Does the government have a national identity and smart card system for cash transfer to the needy? Does the government have safety nets that can be scaled up in terms of benefits, coverage, or both to compensate the vulnerable for higher oil prices in a way that would be consistent with medium- and long-term goals for social protection? Or will deployment of sound social safety nets require considerable preparatory work and development? Delivery of essential social What is the state of primary education, primary health, access to services safe water, access to sanitation? Is the track record of delivery such that the public would consider credible the government’s promises of putting the savings from subsidy reduction to better uses? Reform steps Sector structure and regulation Is the market sufficiently large to become competitive over time, making price deregulation a realistic goal? If there is market concentration, what are the main bottlenecks to breaking down the market power of the incumbents—import terminals, refineries, depot terminals, pipelines? Are there laws or regulations on supply that need to be amended? Do regulations and standards reflect current international good practice, or do they need to be updated? Is there monitoring and enforcement, and how can they be strengthened? How is commercial malpractice tackled and is there a plan to reduce it further? Who will set prices and how Will there be an independent regulatory agency in charge of setting prices? Are there laws or regulations on pricing that need to be amended? If there are large price subsidies, what transition steps 45 Area Consideration Specific issues are needed before an automatic formula-based pricing mechanism can be adopted or prices deregulated? Timing Are there events outside the oil sector that could affect timing— national elections, natural disasters, food crisis, large-scale agricultural crop failure, domestic or international financial crisis, soaring unemployment, collapsing prices of other commodities such as coffee or minerals that the economy depends on? Most of these would call for greater social protection measures in response to price reforms. Is there a time when fuel consumption is higher— major national holidays, winter in cold-climate areas, summer travel period—that should be avoided for raising official fuel prices? Analysis of winners and losers Is there a reasonable understanding of effects of price reforms on different segments of society and income groups, including likely effects on inflation and which sectors would be particularly affected? What are the relative effects of higher food prices, higher transport fares, and higher energy prices on the poor? If food prices are more important because of the expenditure patterns of the poor, that would argue even more for moving away from sectoral approaches and combining safety nets for all risks under one umbrella. Does the financial viability of some businesses depend on oil price subsidies? If so, they will lobby to oppose subsidy reforms through industry associations, trade unions, and other groups. Is there a need for managed closure of these businesses and retraining of staff? Would it be possible to make small, regular, incremental price increases that minimize adverse effects? Would it make sense to provide support for fuel switching or fuel efficiency improvement? Does the power sector rely on diesel, fuel oil, or both? Can power utilities pass on oil price increases to consumers, or will they have to bear financial losses until the next round of tariff adjustments? Is underpricing a problem in the power sector? If oil product price increases will significantly increase power tariffs, affect the financial viability of power utilities, or do both, careful consideration needs to be given to coordination between the oil and power sectors, and to the political economy of power tariff reforms. Are there powerful groups benefiting from subsidies, who can exercise their influence to block price reforms? Are some benefitting illegally from subsidies by engaging in smuggling, black marketing, diversion, and fuel adulteration? Do they include high-level government officials and high-level officers in oil companies? If so, building a broad-based coalition of supporters for price reforms would be all the more important. Immediate, tangible benefits Would it be possible to deliver immediate benefits of the price reform? In the oil sector, these could be no more queues, much less fuel adulteration, much lower black market prices, and a crackdown on short-selling. Outside the oil sector, are there existing administrative systems in place that can deliver compensation immediately and is visible to the public? If not, is there something that could be set up quickly, and could the start of any large price adjustments be postponed until that setup is nearly, if not fully, operational? Longer-term assistance Aside from initial compensation to help adjust to higher prices, is 46 Area Consideration Specific issues there a need for longer-term compensation or assistance, such as energy efficiency fund or tax expenditures for acquisition of more efficient equipment and appliances to reduce oil consumption, larger food assistance, and long-term cash compensation to the poor for higher fuel prices? Communication About the current state Is the public aware of the size of under-recoveries, who is benefitting, the distortions caused by keeping prices low, and the opportunity costs of the under-recoveries? Can the public easily find out past and present price gaps? Is there a national dialogue on the pros and cons of the current pricing policy? About future options or plan Are options or a proposal for price reform being communicated effectively and accurately? Or are rumors causing panic buying and hoarding? Is there a mechanism to consult different stakeholders and include them in deliberation and decision-making to the extent possible? Is communication about compensation plans undertaken far in advance of the implementation of the price reform, so that the public is well prepared? Means of communication Are all forms of communication being exploited? Is consideration being given to a Web-based national conversation, giving many people an opportunity to be heard? Are all segments of society being reached, including those without access to the internet or TV? Is electronic communication being complemented by face-to- face stakeholder meetings? Communication about the oil Is there a plan to make price, production, and consumption sector information available regularly and in a timely manner so that consumers and potential investors can take informed decisions? As competition begins to emerge, could the government make price information readily available to further promote price competition? Is there a mechanism for registering complaints? Are companies found in violation of rules named with specific charges outlined? Are all regulations and rules, announcements about pricing policy, calculations of controlled prices, the magnitude of the remaining subsidies and how they are channeled, and any other information related to prices consolidated in one place so that they can be easily found? Is information provided in plain language and comprehensible to many, if not most, people in the country? Source: Author 47 Appendix A This appendix presents the results of converting world prices of gasoline, diesel, and LPG in local currency units and computing changes between 2003 and 2012 in nominal terms and between 2003 and 2011 in real terms. The real prices are computed using CPI in each country. The calculations are based on annual averages data. Kerosene is not shown because world prices of kerosene closely track diesel prices. For gasoline and diesel, U.S. Gulf Coast prices are taken for the Americans, prices in Singapore for East Asia and the Pacific, and Northwest European prices for the rest of the world. For LPG, propane spot prices in Mont Belvieu are taken for the Americas and the Saudi Aramco contract prices for propane and butane, assuming a 50/50 mixture, for the rest of the world. Table A.1 and Table A.2 show the summary statistics for gasoline and LPG, respectively, by income. Table A.3 shows percentage increases in nominal and real prices in local currency units for individual countries. Table A.1: Increase in world price gasoline prices in local currency between 2003 and 2012 (nominal) and 2011 (real) All Low Lower Upper High middle middle Non- All OECD OECD Percent increase in nominal prices in local currency between 2003 and 2012 Minimum 96 220 96 100 120 160 120 Maximum 1,360 1,180 640 1,360 480 260 480 Median 220 300 240 260 220 230 220 No. of countries 172 32 48 47 14 14 31 Percent increase in real prices in local currency between 2003 and 2011 Minimum -51 14 -51 12 53 78 53 Maximum 300 180 170 300 210 190 210 Median 110 110 100 110 140 120 140 No of countries 160 31 44 43 42 11 31 Ratio with price increase in the United States in real terms between 2003 and 2011 Minimum 0.2 0.4 0.2 0.4 0.6 0.7 0.6 Maximum 1.5 1.1 1.0 1.5 1.2 1.1 1.2 Median 0.8 0.8 0.8 0.8 0.9 0.8 0.9 Sources: Author’s calculations. 48 Table A.2: Increase in the world price of LPG between 2003 and 2012 (nominal) and 2011 (real) All Low Lower Upper High middle middle All Non- OECD OECD Percent increase in nominal prices in local currency between 2003 and 2012 Minimum 9 72 19 9 24 74 24 Maximum 1,170 1,020 550 1,170 410 210 410 Median 170 250 180 170 170 170 170 No. of countries 172 32 48 47 45 14 31 Percent increase in real prices in local currency between 2003 and 2011 Minimum -59 -5 -59 -10 29 43 29 Maximum 230 130 120 230 160 140 160 Median 80 80 70 80 100 80 100 No of countries 160 31 44 43 42 11 31 Ratio with price increase in the United States in real terms between 2003 and 2011 Minimum 0.2 0.5 0.2 0.4 0.6 0.7 0.6 Maximum 1.6 1.1 1.0 1.6 1.3 1.2 1.3 Median 0.9 0.9 0.8 0.9 1.0 0.9 1.0 Sources: Saudi Aramco propane and butane contract prices from Reuters, various issues; Mont Belvieu propane spot prices from EIA; consumer price index from WDI; author’s calculations. Table A.3: Percent increases in the world prices of gasoline, diesel, and LPG in local currency units Country Nominal price increase, 2003–2012 Real price increase, 2003–2011 Gasoline Diesel LPG Gasoline Diesel LPG Afghanistan 270 310 230 74 91 45 Albania 220 240 180 130 150 92 Algeria 260 280 210 140 160 100 Angola 360 380 300 16 26 -2 Antigua and Barbuda 220 240 74 160 180 110 Argentina 410 430 170 130 140 86 Armenia 150 160 120 46 58 23 Australia 120 140 96 74 91 45 Austria 220 230 170 140 160 99 Azerbaijan 190 200 150 34 45 13 Bahrain 260 280 210 190 210 140 Bangladesh 400 450 340 140 160 100 Barbados 220 240 74 100 110 63 Belarus 1,360 1,440 1,170 160 180 120 Belgium 220 230 170 130 150 96 Belize 220 240 74 170 180 120 Benin 220 230 170 120 130 82 Bermuda 220 240 74 — — — Bhutan 310 350 260 110 130 73 49 Country Nominal price increase, 2003–2012 Real price increase, 2003–2011 Gasoline Diesel LPG Gasoline Diesel LPG Bolivia 190 210 57 73 83 39 Bosnia and Herzegovina 220 230 170 — — — Botswana 450 480 380 140 160 100 Brazil 110 120 11 12 19 -10 Bulgaria 220 230 170 74 89 47 Burkina Faso 220 230 170 120 140 88 Burundi 380 410 320 83 98 54 Cambodia 260 290 220 110 130 72 Cameroon 220 230 170 130 150 92 Canada 130 140 24 91 100 54 Cape Verde 220 230 180 120 140 89 Central African Republic 220 230 170 120 140 87 Chad 220 230 170 150 170 110 Chile 130 140 23 72 82 38 China 170 200 140 110 130 76 Colombia 100 110 9 40 49 13 Comoros 220 230 170 120 130 83 Congo, Dem. Rep. 710 760 610 110 130 80 Congo, Rep. 220 230 170 100 120 70 Costa Rica 310 330 120 89 100 53 Côte d'Ivoire 220 230 170 120 140 87 Croatia 210 230 170 120 140 84 Czech Republic 150 160 120 76 90 48 Denmark 220 230 170 140 160 100 Djibouti 260 280 210 140 160 100 Dominica 220 240 74 150 170 100 Dominican Republic 310 330 120 68 78 35 Ecuador 220 240 74 130 140 86 Egypt, Arab Rep. 270 290 220 57 70 32 El Salvador 220 240 74 130 140 86 Equatorial Guinea 220 230 170 84 99 55 Eritrea 300 320 250 — — — Estonia 220 230 170 96 110 66 Ethiopia 640 680 540 97 110 66 Fiji 240 270 200 130 150 90 Finland 220 230 170 140 160 110 France 220 230 170 140 160 110 Gabon 220 230 170 130 150 98 Gambia, The 300 320 250 130 150 93 Georgia 180 190 140 52 65 28 Germany 220 230 170 140 160 110 50 Country Nominal price increase, 2003–2012 Real price increase, 2003–2011 Gasoline Diesel LPG Gasoline Diesel LPG Ghana 640 690 550 130 140 90 Greece 220 230 170 120 130 82 Greenland 220 230 170 — — — Grenada 220 240 74 140 150 93 Guatemala 220 240 72 85 96 49 Guinea 1,180 1,250 1,020 Guinea-Bissau 220 230 170 110 130 81 Guyana 240 260 84 110 120 67 Haiti 220 240 72 31 38 5 Honduras 250 270 90 97 110 59 Hungary 260 280 210 110 120 74 Iceland 480 520 410 210 240 160 India 310 350 260 95 110 62 Indonesia 290 320 240 98 120 65 Iran, Islamic Rep. 430 460 360 38 49 16 Iraq 96 110 71 -51 -47 -59 Ireland 220 230 170 140 160 100 Israel 200 220 160 130 150 92 Italy 220 230 170 140 160 100 Jamaica 400 420 170 86 97 50 Japan 150 170 120 140 170 100 Jordan 260 280 210 130 150 94 Kazakhstan 260 280 210 67 81 41 Kenya 300 320 250 59 72 34 Kiribati 120 140 96 — — — Korea, Rep. 240 270 200 150 180 110 Kuwait 240 260 190 120 140 87 Kyrgyz Republic 290 310 240 72 86 45 Lao PDR 170 190 140 63 79 36 Latvia 240 260 200 84 99 55 Lebanon 260 280 210 200 220 150 Lesotho 290 310 240 110 120 73 Liberia 340 370 290 99 120 68 Libya 250 270 210 — — — Lithuania 210 230 170 100 120 69 Luxembourg 220 230 170 130 150 95 Macedonia, FYR 220 230 180 130 150 95 Madagascar 540 570 450 140 160 100 Malawi 820 870 700 160 180 120 Malaysia 190 220 150 130 150 90 Maldives 330 370 280 — — — 51 Country Nominal price increase, 2003–2012 Real price increase, 2003–2011 Gasoline Diesel LPG Gasoline Diesel LPG Mali 220 230 170 130 150 91 Mauritania 300 330 250 110 130 78 Mauritius 290 310 240 120 140 86 Mexico 290 320 110 160 180 110 Micronesia, Fed. Sts. 260 290 210 — — — Moldova 210 230 170 39 51 17 Mongolia 320 360 270 71 88 43 Morocco 220 240 180 150 170 110 Mozambique 330 350 270 100 120 69 Myanmar — — — 14 26 -5 Namibia 290 310 240 110 130 77 Nepal 300 330 250 81 99 51 Netherlands 220 230 170 150 170 110 New Zealand 160 180 120 100 120 69 Nicaragua 400 430 170 130 140 85 Niger 220 230 170 110 130 81 Nigeria 340 360 280 70 84 43 Norway 190 210 160 140 150 99 Oman 260 280 210 140 170 110 Pakistan 480 530 410 120 140 83 Panama 220 240 74 140 150 89 Papua New Guinea 110 130 83 58 74 32 Paraguay 120 130 19 21 28 -2 Peru 140 160 32 100 110 61 Philippines 180 200 140 87 110 56 Poland 200 220 160 110 120 74 Portugal 220 230 170 140 160 99 Qatar 260 280 210 110 130 80 Romania 270 300 230 81 96 53 Russian Federation 260 280 210 49 62 26 Rwanda 310 330 260 92 110 62 Samoa 180 200 140 66 83 39 Saudi Arabia 260 280 210 150 170 110 Senegal 220 230 170 130 150 94 Seychelles 810 860 690 300 330 230 Sierra Leone 560 600 480 130 150 95 Singapore 160 180 120 110 130 71 Slovak Republic 130 140 100 53 66 29 Slovenia 220 240 180 130 150 94 Solomon Islands 250 280 210 95 110 62 South Africa 290 310 240 110 130 81 52 Country Nominal price increase, 2003–2012 Real price increase, 2003–2011 Gasoline Diesel LPG Gasoline Diesel LPG Spain 220 230 170 130 140 91 Sri Lanka 370 410 310 81 99 51 St. Kitts and Nevis 220 240 74 130 140 85 St. Lucia 220 240 74 150 170 100 St. Vincent and the Grenadines 220 240 74 130 140 85 Sudan 390 420 330 — — — Suriname 310 330 120 92 100 55 Swaziland 290 310 240 99 120 68 Sweden 200 220 160 150 170 110 Switzerland 150 160 120 110 130 79 Syrian Arab Republic 390 410 320 110 130 80 Tajikistan 460 490 380 140 160 99 Tanzania 450 480 380 180 200 130 Thailand 170 190 130 99 120 65 Togo 220 230 170 120 130 82 Tonga 190 210 150 68 84 40 Trinidad and Tobago 230 250 78 78 89 43 Tunisia 330 360 280 180 200 140 Turkey 330 350 270 93 110 63 Uganda 360 380 300 120 140 86 Ukraine 440 470 370 97 110 66 United Arab Emirates 260 280 210 — — — United Kingdom 270 290 220 180 210 140 United States 220 240 74 160 170 110 Uruguay 130 140 25 23 31 -1 Uzbekistan 600 640 510 — — — Vanuatu 170 190 140 110 130 72 Venezuela, RB 760 810 370 — — — Vietnam 380 420 320 99 120 66 Yemen, Rep. 320 340 270 64 78 38 Zambia 290 310 240 42 54 20 Sources: Reuters, various issues, for Saudi Aramco contract prices of propane and butane; EIA for gasoline and diesel prices on the U.S. Gulf Coast and for spot propane prices in Mont Belvieu; industry sources for other markets; WDI for CPI. Note: — = not available. 53 Appendix B Vulnerability in Table B.1 is based on the data from the U.S. Energy Information Administration posted in March 2013. The retail prices in Table B.2 are averaged over the 31 days in January 2013 if prices as a function of time are available for the month. Where one grade of fuel is subsidized by the government, the official price for the subsidized fuel is shown. In particular, the LPG prices in Argentina and Panama are those for the poor subsidized by the government. The reference markets and fuel quality to compute pass-through coefficients in Table B.3 are detailed in Table A2.3 in Kojima (2012). For both Tables B.2 and B.3, types of fuel and prices and sources of data are outlined in Table A2.1 of Kojima (2012). For Uganda, the prices for both 2009 and 2013 are now taken from the price surveys in Kampala reported by the Uganda Bureau of Statistics. As mentioned in Kojima (2012, 2013), the pass-through coefficients in Uganda are low in part because the country experienced acute fuel shortages in January 2009, pushing up prices. Table B.1: Vulnerability Levels in 2011 and Change in Vulnerability between 2009 and 2011 (percentage of GDP) Country 2011 ∆09–11 Country 2011 ∆09–11 Country 2011 ∆09–11 Afghanistan 7.1 0.5 Georgia 5.0 1.1 Oman -41 -6.4 Albania 3.6 1.3 Germany 2.4 0.7 Pakistan 6.4 1.8 Algeria -31 -4.8 Ghana -1.3 -5.4 Panama 15 5.9 Angola -64 1.1 Greece 4.4 1.6 Papua New Guinea -3.0 1.3 Antigua and Barbuda 14 5.1 Grenada 9.3 3.5 Paraguay 4.1 -0.4 Argentina -0.7 0.8 Guatemala 5.3 1.5 Peru 0.5 -0.2 Armenia 18 5.8 Guinea 6.7 1.8 Philippines 4.8 1.2 Australia 1.4 0.3 Guinea-Bissau 11 3.3 Poland 4.0 1.4 Austria 2.1 0.7 Guyana 16 4.2 Portugal 4.1 1.4 Azerbaijan -54 -6.8 Haiti 7.2 1.9 Qatar -32 -7.7 Bahamas, The 11 4.5 Honduras 10 3.5 Romania 2.3 0.9 Bangladesh 3.5 1.1 Hungary 3.1 0.8 Russian Federation -15 -1.7 Barbados 8.2 2.9 Iceland 5.6 1.5 Rwanda 3.3 1.0 Belarus 11 3.9 India 5.0 1.1 Samoa 6.5 1.6 Belgium 4.7 1.7 Indonesia 1.5 0.4 Sao Tome and Principe 14 3.6 Belize -1.3 -6.3 Iraq -63 -1.2 Saudi Arabia -56 -11 Benin 19 8.1 Ireland 2.5 0.9 Senegal 10 3.7 Bhutan 4.4 0.8 Israel 4.0 1.4 Serbia 4.9 1.3 Bolivia 1.1 0.4 Italy 2.3 0.7 Seychelles 30 10 Bosnia and Herzegovina 5.9 2.2 Jamaica 15 4.2 Sierra Leone 17 5.5 Botswana 4.4 0.8 Japan 2.8 0.8 Singapore 20 4.6 Brazil -0.1 0.0 Jordan 15 5.6 Slovak Republic 2.9 0.9 Brunei Darussalam -32 -2.4 Kazakhstan -29 -2.8 Slovenia 4.2 1.6 Bulgaria 7.2 2.5 Kenya 9.2 3.6 Solomon Islands 6.8 1.8 Burkina Faso 4.5 1.3 Kiribati 9.1 1.2 South Africa 3.8 0.9 Burundi 2.3 0.5 Korea, Rep. 7.4 1.5 Spain 3.5 1.2 Cambodia 9.8 2.8 Kuwait -49 -3.6 Sri Lanka 5.8 1.2 Cameroon -4.1 0.4 Kyrgyz Republic 21 5.3 St. Kitts and Nevis 9.8 4.0 Canada -2.9 -1.0 Lao PDR 1.4 0.2 St. Lucia 9.0 3.5 Cape Verde 5.4 1.7 Latvia 4.6 1.7 St. Vincent and the Grenadines 8.3 3.3 Central African Republic 4.0 1.3 Lebanon 13 4.3 Sudan -19 -3.8 Chad -45 -9.3 Lesotho 4.7 2.6 Swaziland 4.6 0.9 54 Country 2011 ∆09–11 Country 2011 ∆09–11 Country 2011 ∆09–11 Chile 4.8 0.1 Liberia 9.8 2.5 Sweden 2.1 0.4 China 2.8 0.7 Lithuania 4.7 1.3 Switzerland 1.5 0.3 Colombia -7.5 -3.4 Luxembourg 3.9 1.4 Tajikistan 8.0 3.4 Comoros 6.2 2.0 Macedonia, FYR 7.1 2.5 Tanzania 5.6 1.9 Congo, Dem. Rep. -2.0 0.1 Madagascar 4.6 1.4 Thailand 6.7 1.7 Congo, Rep. -75 -13.3 Malawi 3.9 1.2 Timor-Leste -300 -25 Costa Rica 4.4 0.8 Malaysia -0.2 0.9 Togo 10 3.9 Cote d'Ivoire -2.0 1.0 Maldives 13 4.9 Tonga 10 2.0 Croatia 4.3 1.8 Mali 1.8 0.6 Trinidad and Tobago -16 -3.6 Cyprus 8.9 3.4 Malta 19 8.1 Tunisia 1.9 2.1 Czech Republic 3.3 1.0 Mauritania 4.0 2.8 Turkey 3.2 0.8 Denmark -0.8 -0.1 Mauritius 7.4 2.2 Turkmenistan -15 -2.7 Dominica 7.8 3.6 Mexico -2.7 -0.4 Uganda 5.2 2.0 Dominican Republic 8.2 2.8 Moldova 9.2 3.2 Ukraine 5.0 1.0 Ecuador -16 -4.2 Mongolia 5.2 -0.4 United Arab Emirates -26 -7.4 Egypt, Arab Rep. 1.1 1.2 Montenegro 3.4 1.3 United Kingdom 0.7 0.5 El Salvador 7.5 2.8 Morocco 8.9 3.2 United States 2.2 0.5 Equatorial Guinea -56 7.0 Mozambique 5.0 1.5 Uruguay 4.4 0.0 Eritrea 4.4 0.3 Namibia 6.8 1.2 Uzbekistan -0.6 0.2 Estonia 2.6 0.6 Nepal 4.0 0.5 Vanuatu 5.0 1.8 Ethiopia 5.9 2.3 Netherlands 4.3 1.6 Venezuela, RB -21 -8.4 Fiji 8.9 1.5 New Zealand 2.3 0.5 Vietnam 1.4 2.4 Finland 2.8 0.8 Nicaragua 12 4.4 Yemen, Rep. -3.8 10 France 2.3 0.8 Niger -1.1 -3.3 Zambia 3.3 1.0 Gabon -51 -4.4 Nigeria -36 -9.7 Zimbabwe 5.5 0.9 Gambia, The 13 4.9 Norway -14 -0.9 Sources: Author’s calculations based on data from EIA (2013a) and WDI (2012). Table B.2 Retail prices in January 2013 in U.S. dollars and the ratio of kerosene prices to diesel prices Country Gasoline Diesel Kerosene LPG Kerosene/diesel Angola $0.63 $0.42 $0.27 $0.39 0.65 Argentina $1.30 $1.40 $1.20 $0.32 0.86 Bangladesh $1.25 $0.86 $0.86 $1.72 1.00 Bolivia $0.69 $0.54 $0.39 $0.33 0.73 Brazil $1.40 $1.09 $1.57 Cambodia $1.34 $1.29 $1.47 Cameroon $1.15 $1.05 $0.71 $0.97 0.67 Chile $1.58 $1.27 $1.32 1.04 China $1.25 $1.25 $1.24 Colombia $1.53 $1.19 Costa Rica $1.28 $1.26 $1.11 $1.38 0.88 Cote d'Ivoire $1.61 $1.25 $1.26 $0.68 1.00 Dominican Republic $1.42 $1.38 $1.25 $1.16 0.91 Egypt, Arab Rep. $0.26 $0.16 $0.16 $0.06 1.00 El Salvador $1.05 $1.10 $1.01 Ethiopia $1.02 $0.92 $0.75 0.82 Gabon $1.08 $0.95 $0.56 $0.88 0.59 55 Country Gasoline Diesel Kerosene LPG Kerosene/diesel Ghana $0.90 $0.90 $0.48 $0.68 0.53 Guatemala $1.09 $1.07 $1.01 Honduras $1.14 $1.11 $0.94 $1.02 0.85 India $1.25 $0.88 $0.28 $0.54 0.32 Indonesia $0.46 $0.46 $0.26 $0.44 0.56 Iran, Islamic Rep. $0.33 $0.29 Iraq $0.39 $0.30 $0.29 $0.32 0.97 Jordan $1.10 $0.94 $0.94 $1.13 1.00 Kazakhstan $0.73 $0.77 Kenya $1.28 $1.20 $0.97 $2.82 0.81 Lao PDR $1.32 $1.16 $1.17 Liberia $1.15 $1.21 $1.21 1.00 Madagascar $1.51 $1.23 $0.95 $2.82 0.77 Malawi $1.66 $1.63 $0.47 $4.38 0.29 Malaysia $0.61 $0.58 $0.61 Mexico $0.85 $0.87 $0.94 Mongolia $1.20 $1.29 Morocco $1.46 $0.97 $0.40 Mozambique $1.56 $1.21 $0.94 $1.83 0.78 Namibia $1.16 $1.24 Nepal $1.43 $1.15 $1.15 $1.20 1.00 Nicaragua $1.22 $1.17 $1.16 $0.92 0.99 Niger $1.09 $1.09 $1.00 $0.61 0.92 Nigeria $0.62 $0.98 $0.32 $2.03 0.32 Pakistan $1.04 $1.12 $1.01 $1.44 0.90 Panama $1.01 $1.01 $0.39 Peru $1.27 $1.30 $1.25 Philippines $1.25 $1.01 $1.27 $1.59 1.26 Russian Federation $0.94 $1.07 Rwanda $1.58 $1.58 Senegal $1.80 $1.60 $1.28 $0.62 0.80 South Africa $1.31 $1.25 $0.96 $2.51 0.76 Sri Lanka $1.25 $0.91 $0.84 $1.51 0.92 Syrian Arab Republic $0.75 $0.42 $0.61 Tajikistan $1.38 $1.39 Tanzania $1.24 $1.22 $1.21 0.99 Thailand $1.48 $1.00 $1.19 $0.61 1.19 Togo $1.20 $1.27 $0.99 $0.89 0.78 Tunisia $0.94 $0.83 $0.52 $0.37 0.62 Turkey $2.43 $2.20 $1.76 $3.16 0.80 Uganda $1.41 $1.30 $1.06 $2.75 0.82 Uruguay $1.95 $1.86 $1.41 $1.53 0.76 Venezuela, RB $0.01 $0.01 $0.07 Vietnam $1.11 $1.03 $1.04 $1.69 1.00 Yemen, Rep. $0.58 $0.47 $0.56 Zambia $1.55 $1.43 $0.98 0.68 Canada $1.19 $1.25 56 Country Gasoline Diesel Kerosene LPG Kerosene/diesel France $2.07 $1.53 Germany $2.14 $1.64 Italy $2.34 $1.87 Japan $1.62 $1.40 $1.03 $3.88 0.74 Spain $1.89 $1.51 United Kingdom $2.05 $1.81 United States $1.19 $1.03 $1.27 Source: Author’s calculations using data from the sources cited in Table A1.2 in Kojima 2012. Note: The prices in Uruguay are ceilings on wholesale prices set by the government. They appear to be close to retail prices, and hence are shown as retail. Table B.3: Pass-through Coefficients, January 2009− January 2012 and January 2009− January 2013 Gasoline Diesel Kerosene LPG Country Jan 2012 Jan 2013 Jan 2012 Jan 2013 Jan 2012 Jan 2013 Jan 2012 Jan 2013 Angola 22 20 9 8 -18 -18 -22 -16 Argentina 91 116 108 117 84 109 -18 -46 Bangladesh 2 28 23 55 23 54 113 136 Bolivia 3 3 2 2 2 1 1 2 Brazil 104 77 54 42 ― ― 110 154 Cambodia 126 128 126 140 128 ― 99 92 Cameroon -7 -2 -7 -3 -4 -2 -5 -1 Chile 199 222 132 132 90 100 153 China 94 98 107 121 ― ― 78 79 Colombia 88 127 103 125 142 145 240 301 Costa Rica 103 118 89 94 67 80 112 142 Côte d'Ivoire 39 66 0 21 72 86 -3 14 Dominican Republic 138 151 139 140 132 134 114 112 Egypt, Arab Rep. -6 -12 -4 -9 -4 -8 -1 -1 El Salvador 122 117 129 118 ― ― 182 220 Ethiopia 102 76 94 70 74 61 145 ― Gabon -9 -5 4 8 -4 -1 -10 -6 Ghana 100 65 97 64 0 -8 1 31 Guatemala 96 102 99 104 66 ― 95 -32 Guinea-Bissau 65 ― 83 ― ― ― ― ― Honduras 118 122 115 113 113 113 124 188 India 78 70 32 55 26 23 24 16 Indonesia 20 17 23 21 13 11 19 12 Iran, Islamic Rep. 56 47 72 66 ― ― ― ― Iraq 9 5 -10 -15 40 38 8 3 Jamaica 128 130 130 127 144 138 133 273 Jordan 83 126 55 108 56 107 6 67 Kazakhstan 34 57 28 74 ― ― ― ― Kenya 93 68 89 75 ― ― ― ― Lao PDR 125 125 127 126 ― ― 155 49 Liberia 95 106 81 86 80 81 ― ― Madagascar 8 36 2 26 -2 28 128 150 57 Gasoline Diesel Kerosene LPG Country Jan 2012 Jan 2013 Jan 2012 Jan 2013 Jan 2012 Jan 2013 Jan 2012 Jan 2013 Malawi 115 -28 129 -9 -26 ― ― ― Malaysia 28 29 25 28 ― ― 26 21 Mexico 45 85 57 92 ― ― 33 123 Mongolia 99 87 71 86 ― ― ― ― Morocco -35 32 -96 -50 -2 2 Mozambique 179 136 64 37 67 49 95 43 Namibia 126 122 112 125 ― ― ― ― Nepal 65 84 52 100 51 97 34 27 Nicaragua 105 115 116 111 84 83 86 71 Niger 28 14 6 -9 33 36 Nigeria 70 39 122 117 -7 -5 202 150 Pakistan 58 65 91 99 83 88 99 92 Panama 104 120 103 112 ― ― 0 0 Peru 99 104 111 116 ― ― 48 93 Philippines 120 123 98 96 99 117 142 114 Russian Federation 52 85 65 122 ― ― ― ― Rwanda 52 51 58 60 ― ― ― ― Senegal 129 161 148 145 116 113 79 -34 South Africa 161 155 157 151 120 111 132 94 Sri Lanka 32 42 30 72 46 96 96 84 Syrian Arab Republic 8 -22 -65 -30 ― ― 1 28 Tajikistan 158 136 154 167 ― ― ― ― Tanzania 60 63 38 38 139 133 ― ― Thailand 122 176 108 119 22 34 12 15 Togo 37 40 60 66 ― ― ― ― Tunisia -4 12 -2 17 5 6 -6 -4 Turkey 127 184 125 215 95 134 146 157 Uganda 24 -4 64 17 40 -8 62 99 Uruguay 166 204 169 183 128 145 91 172 Venezuela, RB -4 -5 -3 -3 -17 -35 Vietnam 77 99 81 99 65 84 50 64 Yemen, Rep. 114 58 14 71 15 87 56 Zambia 97 82 97 87 60 51 ― ― Canada 130 133 133 124 ― ― ― ― France 125 143 111 123 ― ― ― ― Germany 124 146 109 130 ― ― ― ― Italy 165 192 160 187 ― ― ― ― Japan 148 98 122 75 99 69 151 18 Spain 143 169 124 144 ― ― ― ― United Kingdom 186 169 154 145 ― ― ― ― United States 100 176 102 101 ― ― 71 74 Source: Author’s calculations based on data cited in Table A2.1 in Kojima 2012. Note: Three figures are retained for coefficients exceeding 99% for formatting reasons only and are not intended to signify the number of significant figures. For kerosene and LPG in Colombia, for all fuels in Jamaica and Uruguay, and for LPG in South Africa, wholesale or ex-refinery prices are used for pass-through computation because retail 58 prices were not available for at least one of the two years. The pass-through coefficient for LPG in Egypt is based on the official depot price. — = not available. 59 References Andriamihaja, Noro and Giovanni Vecchi. 2008. “An Evaluation of the Welfare Impact of Higher Energy Prices in Madagascar.” Africa Region Working Paper Series No. 106. Washington DC: World Bank. www.worldbank.org/afr/wps/wp106.pdf. Al Jazeera. 2013. “Mali speaks.” www.aljazeera.com/indepth/interactive/2013/01/201312113451635182.html. All Africa. 2011. “Marketer takes delivery under questionable circumstance.” July 11. ———. 2012. “Mozambique: Government Restructures Imopetro.” November 14. AP (Associated Press). 2007. “Report: Gas costs up to 10 cents more without price cap.” December 27. Arze del Granado, Francisco Javier, David Coady, and Robert Gillingham. 2012. “The Unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries.” World Development 40(11): 2234–2248. www.sciencedirect.com/science/article/pii/S0305750X12001155#. Asia Pulse. 2005. “Kerosene sales surge raises suspicions in Vietnam.” May 10. Bacon, Robert. 2001. “Petroleum Taxes.” Viewpoint No. 240. Bacon, Robert, Soma Bhattacharya, and Masami Kojima. 2010. “Expenditures of low-income households on energy: Evidence from Africa and Asia.” Extractive Industries for Development Series #16. Washington DC: World Bank. http://siteresources.worldbank.org/EXTOGMC/Resources/336929- 1266963339030/eifd16_expenditure.pdf. Bacon, Robert, and Masami Kojima. 2006. “Coping with Higher Oil Prices.” ESMAP Report 323/06. Washington DC: World Bank. http://www.esmap.org/esmap/sites/esmap.org/files/FR323- 06_GBL_Coping_with_Higher_Oil_Prices.pdf. ———. 2008a. “Coping with Oil Price Volatility.” ESMAP Special Report 005/08. Washington DC: World Bank. http://esmap.org/esmap/sites/esmap.org/files/8142008101202_coping_oil_price.pdf. ———. 2008b. “Vulnerability to Oil Price Increases: A Decomposition Analysis of 161 Countries.” Extractive Industries for Development Series #1. Washington DC: World Bank. http://siteresources.worldbank.org/INTOGMC/Resources/eid1_oil_price_vulnerability.pdf. ———. 2010a. “Expenditures of low-income households on energy: Evidence from Africa and Asia.” Extractive Industries for Development Series #16. Washington DC: World Bank. http://siteresources.worldbank.org/EXTOGMC/Resources/336929-1266963339030/ eifd16_expenditure.pdf. ———. 2010b. “Rockets and Feathers: Asymmetric Petroleum Product Pricing in Developing Countries.” Extractive Industries for Development Series #18. Washington DC: World Bank. http://siteresources.worldbank.org/EXTOGMC/Resources/336929-1266963339030/ eifd18_rockets_feathers.pdf. BMI (Business Monitor International). 2011. “Fuel supply crisis forces further reforms at Mombasa refinery.” May 11. Borneo Post. 2012. “New rule for local fishermen to reduce catch leakages.” August 8. 60 Bradley, Jr., Robert L. 1995. Oil, Gas & Government: The U.S. Experience. Lanham, Maryland: Rowman and Littlefield Publishers. CNE (Comisión Nacional de Energía). 2013. “Sistema de Información en Línea de Precios de Combustibles en Estaciones de Servicio.” www.bencinaenlinea.cl/web2/. Coady, David, Robert Gillingham, Rolando Ossowski, John Piotrowski, Shamsuddin Tareq, and Justin Tyson. 2010. “Petroleum Product Subsidies: Costly, Inequitable, and Rising.” IMF Staff Position Note 10/05. www.imf.org/external/pubs/ft/spn/2010/spn1005.pdf. Dahl, Carol A. 2012. “Measuring global gasoline and diesel price and income elasticities.” Energy Policy 41: 2–13. www.lantm.lth.se/fileadmin/fastighetsvetenskap/Nationalekonomi/Elasticities.pdf. Deacon, Robert T., and Jon Sonstelie. 1989. “The welfare costs of rationing by waiting.” Economic Inquiry 27(2): 179–196. http://onlinelibrary.wiley.com/doi/10.1111/j.1465- 7295.1989.tb00777.x/pdf. EIA (Energy Information Administration). 2013a. “International Energy Statistics.” www.eia.gov/cfapps/ipdbproject/IEDIndex3.cfm#. ———. 2013b. “Petroleum and other liquids – Data – Prices.” www.eia.gov/petroleum/data.cfm#prices. ———. 2013c. “Prices, Sales Volumes & Stocks by State.” www.eia.gov/dnav/pet/pet_sum_mkt_dcu_SHI_m.htm EITI (Extractive Industries Transparency Initiative). http://eiti.org/. Energypedia. “International Fuel Prices.” https://energypedia.info/index.php/International_Fuel_Prices. EPPO (Energy Policy and Planning Office). 2013. “Price Structure of Petroleum Products in Bangkok.” www.eppo.go.th/petro/price/index.html. Fiszbein, Ariel, Norbert Schady, Francisco Ferreira, Margaret Grosh, Nial Kelleher, Pedro Olinto, and Emmanuel Skoufias. 2009. Conditional Cash Transfers: Reducing Present and Future Poverty. Washington DC: World Bank. http://siteresources.worldbank.org/INTCCT/Resources/5757608- 1234228266004/PRR-CCT_web_noembargo.pdf. FTC (Federal Trade Commission). 2003. “Competition and the effects of price controls in Hawaii’s gasoline market.” www.ftc.gov/be/v030005.shtm. GIZ (Deutsche Gesellschaft für Internationale Zusammenarbeit). 2012. “Urban Transport and Energy Efficiency. Moduel 5h. Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities.” Bonn. www2.gtz.de/wbf/4tDx9kw63gma/SUT_module5h.pdf. Grosh, Margaret, Carlo del Ninno, Emil Tesliuc, and Azedine Ouerghi. 2008. For Protection & Promotion: The Design and implementation of Effective Safety Nets. Washington DC: World Bank. http://siteresources.worldbank.org/SPLP/Resources/461653- 1207162275268/For_Protection_and_Promotion908.pdf. GSI (Global Subsidies Initiative). “Fossil-fuel subsidies.” www.iisd.org/gsi/fossil-fuel-subsidies. Guardian. 2011. “Nigeria to set up special body to manage funds accruing after oil subsidy removal.” October 26. 61 Guatemala MEM. (Ministry of Energy and Mining). 2013. “Rutas de Precios Bajos – Estaciones de Servicio Área Metropolitana.” www.mem.gob.gt/viceministerio-de-mineria-e-hidrocarburos- 2/direccion-general-de-hidrocarburos/precios/rutas-de-precios-bajos-estaciones-de-servicio-area- metropolitana/. Harvey, Scott and Calvin T. Roush, Jr. 1981. “Petroleum product price regulations: output, efficiency, and competitive effects.” Staff report of the Bureau of Economics to the Federal Trade Commission. Washington DC. Hindu. 2011. “Brazen criminality.” January 31. IANGV (International Association for Natural Gas Vehicles). 2012. “Current natural gas vehicle statistics.” July 2012. www.iangv.org/current-ngv-stats/. IEA (International Energy Agency). 2004. Energy Policies of OECD Countries: Portugal 2004 Review. Paris. ———. 2012. “World energy statistics.” IEA World Energy Statistics and Balances (database). doi: 10.1787/data-00510-en. IMF (International Monetary Fund). 2012. “Ghana: Fifth Review Under the Three-Year Arrangement Under the Extended Credit Facility and Request for Modification of Performance Criteria—Staff Report; Staff Supplements; Press Release on the Executive Board Discussion; and Statement by the Executive Director for Ghana.” IMF Country Report No. 12/36. www.imf.org/external/pubs/ft/scr/2012/cr1236.pdf. ———. 2013a. “Energy Subsidy Reform: Lessons and Implications.” Washington DC. www.imf.org/external/np/pp/eng/2013/012813.pdf. ———. 2013b. “Case Studies on Energy Subsidy Reform: Lessons and Implications.” Washington DC. www.imf.org/external/np/pp/eng/2013/012813a.pdf. India. Ministry of Consumer Affairs, Food, and Public Distribution. 2013. “Smart cards for PDS.” February 27. http://pib.nic.in/newsite/erelease.aspx?relid=92612. INE (Instituto Nicaragüense de Energía). 2013. “Monitoreo de precios de los combustibles 2013.” www.ine.gob.ni/dgh.html. Keane, Michael P., and Eswar S. Prasad. 1996. “The employment and wage effects of oil price changes: A sectoral analysis.” The Review of Economics and Statistics, 78(3): 389–400. Keynes, John Maynard. 1942. “The international regulation of primary products.” In Collected Writings of John Maynard Keynes, ed. D. Moggridge. London: Macmillan and Cambridge University Press, 1980. Kojima, Masami. 2009a. “Changes in End-User Petroleum Product Prices: A Comparison of 48 Countries.” Extractive Industries and Development Series #2. Washington DC: World Bank. http://siteresources.worldbank.org/INTOGMC/Resources/ei_for_development_2.pdf. ———. 2009b. “Government Response to Oil Price Volatility: Experience of 49 Developing Countries.” Extractive Industries and Development Series #10. Washington DC: World Bank. http://siteresources.worldbank.org/INTOGMC/Resources/10-govt_response-hyperlinked.pdf. 62 ———. 2011. “The Role of Liquefied Petroleum Gas in Reducing Energy Poverty.” Extractive Industries and Development Series #25. Washington DC: World Bank. http://siteresources.worldbank.org/INTOGMC/Resources/LPGReportWeb-Masami1.pdf. ———. 2012. “Oil price risks and pump price adjustments.” World Bank Policy Research Working Paper No. 6227. http://go.worldbank.org/SIM8QA2VI0. ———. 2013. “Petroleum product pricing and complementary policies: Experience of 65 developing countries since 2009.” World Bank Policy Research Working Paper No. 6396. http://go.worldbank.org/6A2XVLPTY0. Medlock, Kenneth B., and Amy Meyers Jaffe. 2009. “Who is in the oil futures market and how has it changed?” James A. Baker III Institute for Public Policy, Rice University. www.bakerinstitute.org/publications/EF-pub-MedlockJaffeOilFuturesMarket-082609.pdf. Meyler, Aidan. 2009. “The pass through of oil prices into euro area consumer liquid fuel prices in an environment of high and volatile oil prices.” Energy Economics 31: 867–881. doi:0.1016/j.eneco.2009.07.002. Modern Ghana. “Premix shortage again as fuel mafias hijack supply process.” August 28. Newbery, David M.G., and Joseph E. Stiglitz. 1981. The Theory of Commodity Price Stabilization: A Study in the Economics of Risk. Oxford: Oxford University Press. OECD (Organisation for Economic Co-Operation and Development). 2013. “Inventory of Estimated Budgetary Support and Tax Expenditures for Fossil Fuels 2013.” www.oecd.org/site/tadffss/. OFT (Office of Fair Trading). 2011. “OFT takes action against heating oil companies and price comparison websites.” Press release 96/11. www.oft.gov.uk/news-and-updates/press/2011/96-11/. ———. 2012. “OFT boosts transparency on heating oil prices.” Press release 64/12. http://www.oft.gov.uk/news-and-updates/press/2012/64-12. Panama. Authority for Consumer Protection and Competition. 2013. “Combustibles.” http://www.autoridaddelconsumidor.gob.pa/estadisticas.asp?area=3. PUC (Public Utilities Commission). 2006. “Annual Report: Fiscal Year 2006–2006.” http://lrbhawaii.info/reports/legrpts/bf/puc/269_5hrs_07.pdf. ———. 2010. “Petroleum Industry Monitoring, Analysis and Reporting (PIMAR).” http://puc.hawaii.gov/industries/petroleum. Reuters. 2013. “UPDATE 1-Egypt sees 3.8 pct growth in fiscal year from July.” April 23. Shell. 2013. “Target One Million: Learn how to save fuel and become one in a million.” www.shell.com/global/products-services/on-the-road/fuels/fuel-save-challenge.html. South Africa. Department of Minerals and Energy. 2007. “Biofuels Industrial Strategy of the Republic of South Africa.” www.info.gov.za/view/DownloadFileAction?id=77830. Stevens, Paul. 2010. “The ‘shale gas revolution’: Hype and Reality.” Chatham House Report. www.chathamhouse.org/sites/default/files/public/Research/Energy,%20Environment%20and%20 Development/r_0910stevens.pdf. Tanzania Daily News. 2012. “Oil sector hails Muhongo over BPS reports.” August 22. 63 Trade Arabia. 2011. “Total Jordan launches move to cut fuel use.” June 26. USDA (U.S. Department of Agriculture). 2013. 2013. “Oil Crop Outlook.” www.ers.usda.gov/publications/ocs-oil-crops-outlook.aspx. Vagliasindi, Maria. 2012. “Implementing Energy Subsidy Reforms: An Overview of the Key Issues.” Policy Research Working Paper 6122. Washington DC: World Bank. http://elibrary.worldbank.org/content/workingpaper/10.1596/1813-9450-6122 Victor, David. 2009. “The Politics of Fossil Fuel Subsidies.” A paper in the series, “Untold billions: fossil-fuel subsidies, their impacts and the path to reform,” by the Global Subsidies Initiative. www.iisd.org/publications/pub.aspx?id=1254. WDI (World Development Indicators) Online. 2012. World Bank, Washington, DC. http://databank.worldbank.org/data/home.aspx. World Bank. 2008. “Guidance for Responses from the Human Development Sector to Rising Food and Fuel.” Priceshttp://siteresources.worldbank.org/SAFETYNETSANDTRANSFERS/Resources/HDNFoo dandFuel_Final.pdf. ———. 2012. “BLT Temporary Unconditional Cash Transfer: Social Assistance Program and Public Expenditure Review 2.” Jakarta. http://documents.worldbank.org/curated/en/2012/02/15893823/bantuan-langsung-tunai-blt- temporary-unconditional-cash-transfer. ———. 2013. “Commodity market.” Price data series at http://go.worldbank.org/QNYZRM5540.