Fossil_Fuel_Subsidies_in_Developing_CountriesSalvador_Ortigueira

Fossil Fuel Subsidies in Developing Countries

Salvador Ortigueira (University of Miami)

Fuel Subsidies

1. Fuel subsidies can be on-budget (explicit) or off-budget (implicit).

2. On-budget subsidies are created, for example, when budgetary resources are used to make direct cash transfers to a producer or a consumer, or when publicly owned refineries and oil marketing companies are mandated to sell below the cost of production and their losses are covered by budgetary funds.

3. Funding a supply of low-priced energy from the budget entails a reduction in public expenditure in other areas, higher taxes, or public borrowing.

Salvador Ortigueira (University of Miami)

Fuel Subsidies

1. Off-budget subsidies can can take the form of tax exemptions or non-reimbursement of losses incurred by state-owned energy companies.

2. The standard procedure to estimate the size of the subsidies is to calculate the difference between an international benchmark price and the domestic retail price

3. This price-gap measure captures the aggregate effect of many subsidies at once

Salvador Ortigueira (University of Miami)

Why Developing Countries Subsidize Fuel Consumption?

1. As an anti-poverty policy: To make fuel consumption accessible to low-income households

2. Oil-producing countries subsidize gasoline to share the wealth from natural resources across the population

3. Fuel subsidies are used to maintain employment, especially in periods of economic transition

4. To fight inflation

Salvador Ortigueira (University of Miami)

Several Countries

Linking Fossil Fuel and Electricity Policies: Subregional Experience 7

Figure 3: Global Subsidies for Direct Use and Power Generation Inputs, 2011

Subsidies for direct use

350

300

250

200

150

100

Coal

$ bi

lli on

Gas Oil

Subsidies for power generation inputs

Source: IEA (2013b).

Subsidies on primary energy resources such as coal, oil, and gas affect their use in the energy mix, investment in the power sector, and, ultimately, prices and the supply of electricity to wider groups of consumers. Coal is cheap and abundant in several South and Southeast Asian countries, especially India and Indonesia, and will remain a large share of the region’s energy mix. Coal is used to produce electricity for manufacturing and households; petroleum, particularly subsidized diesel, is used mainly for transportation. Natural gas exploration and production is rapidly changing the energy landscape in both regions. In Thailand and Malaysia, gas-fired power plants will be the predominant source of electricity generation for some time.

Many developing countries are trying to increase energy prices and reduce subsidies, but governments still heavily subsidize and regulate fossil fuels. Globally, subsidies for the direct use of oil products and gas (that is, when these fuels are used in transport and industry) are by far the largest component of subsidies. Subsidies for the direct use of oil products totaled $280 billion in 2011. By contrast, subsidies for oil as an input to electricity generation contributed only $28 billion (Figure 3). Subsidies for power generation inputs are also significant, at $65 billion for gas, $32 billion for coal, and $28 billion for oil.

In most countries, the state-owned institutions that provide fossil fuels and electricity have historically been supported by government subsidies and have not been required to earn commercial rates of return. In such cases, one immediate impact of fossil fuel subsidies on power sectors has been a shortage of capacity, because financially starved state-owned energy companies cannot invest and maintain infrastructure such as refineries or new generation capacity. Another impact has been to bias

Salvador Ortigueira (University of Miami)

Several Countries

3 Measuring Subsidies Subsidies on fossil fuel consumption are prevalent in developing countries and particularly high in oil- exporting countries (Table 3). Kemp (2014) notes that three-fourths of worldwide fuel consumption subsidies in 2012 stemmed from energy-exporting countries, and Organization of the Petroleum Exporting Countries (OPEC) members accounted for over half of the total (IEA 2014a). Today, such subsidies are nonexistent or small in most OECD countries, but are important in non- OECD ones—and production subsidies that seek to expand domestic supply are important in both (OECD, OECD-IEA Fossil Fuel Subsidies and Other Support).

Table 3: Share of Consumption Subsidies in the Full Cost of Supply, 2013 (%)

Country Average

subsidization rate Country Average

subsidization rate Venezuela 92.7 Ukraine 28.9 Algeria 77.5 Nigeria 28.8 Saudi Arabia 77.3 Pakistan 23.0 Iran 77.1 El Salvador 20.9 Libyan Arab Jamahiriya 76.7 Russian Federation 20.5 Turkmenistan 65.7 India 19.9 Egypt 61.2 Malaysia 15.6 Uzbekistan 58.7 Mexico 11.9 Iraq 53.3 Gabon 8.7 Ecuador 51.2 Ghana 8.5 Bolivia 44.1 Thailand 6.7 Angola 35.9 Viet Nam 4.3 Bangladesh 33.6 China, People’s Rep. of 2.6 Kazakhstan 32.8 Korea, Rep. of 0.2 Indonesia 31.3 Colombia 0.0 Argentina 29.6

Source: International Energy Agency online database. http://www.iea.org/subsidy/index.html

Fuel subsidies can be on-budget (explicit) or off-budget (implicit). On-budget subsidies are created, for example, when budgetary resources are used to make direct cash transfers to a producer or a consumer, or when publicly owned refineries and oil marketing companies are mandated to sell below the cost of production and their losses are covered by budgetary funds. Funding a supply of low-priced energy from the budget entails a reduction in public expenditure in other areas, higher taxes, or public borrowing. In contrast, off-budget subsidies are often “hidden” and difficult to calculate. Such subsidies

Salvador Ortigueira (University of Miami)

Several Countries © OECD/IEA 2016 Fossil Fuel Subsidy Reform in Indonesia and Mexico

Page | 17

Figure 4 • Value of fossil fuel subsidies by fuel in the top 25 countries, 2014

Notes: GDP = gross domestic product; MER = market exchange rates; UAE = United Arab Emirates.

Source: IEA (2015a), World Energy Outlook 2015.

The average rate of subsidisation, i.e. the ratio of the subsidy to the international reference price, also varies significantly from country to country. The total subsidisation rate among the countries identified as subsidising fossil fuel consumption is 21%, with the maximum being in Venezuela at 93%.

Fossil fuel subsidies in ten countries account for USD 364 billion or around three-quarters of the world total. Of the 25 countries with the largest subsidies, 10 are in the Middle East or North Africa – and almost all of them are oil or gas exporters. In fact IEA estimates reveal that fossil fuel subsidies are becoming increasingly concentrated in the major oil and gas-exporting countries. For example, the share of Middle East oil exporters in the world total has risen from 35% to 40% over the last four years. (IEA, 2015a)

5% 10% 15% 20% 25%

20 40 60 80 100

Nigeria Mexico

Malaysia Kazakhstan

Ecuador Qatar

Ukraine Pakistan

Libya Oman

Turkmenistan Kuwait

Uzbekistan Iraq

Argentina China

UAE Algeria

Egypt Indonesia

Venezuela India

Russia Saudi Arabia

Iran

USD Billion

Oil

Gas

Coal

Electricity

Total subsidies as share of GDP (MER) (top axis)

Salvador Ortigueira (University of Miami)

Mexico

1. Mexico is an oil-producing country but currently imports about half of the gasoline demanded in the country

2. Gasoline pricing policy: A pre-determined crawling gas price fixed by the government at the beginning of each year

3. A gas subsidy emerges when the international gasoline price is above the pre-determined price level

4. The subsidy, which may produce a final price below its market price, is frequently justified on economic and political grounds

Salvador Ortigueira (University of Miami)

Mexico

Author's personal copy

motor vehicle fleet than advanced countries, which increases pollution (Harrington and McConnell, 2003). All of these features call for the calculation of an appropriate gasoline tax for LDCs.

The objective of this paper is to estimate the optimal gasoline tax for a representative middle-income country. For that purpose, we apply the methodology of Parry and Small (2005) to Mexico, a prominent oil-producing LDC that heavily subsidizes gasoline consumption. The advantage of this method is the decomposition of the second-best optimal fuel tax into several components, including those related to congestion, accidents, and air pollution. As previously mentioned, these negative externalities may in fact be more severe for the economies of LDCs than for developed economies.

Our results suggest an optimal gasoline tax of $1.90/gallon at 2011 prices. The (adjusted) Pigouvian tax is the largest portion of the tax, amounting to $1.62/gallon. The accident component explains approximately one-third of the Pigouvian tax, followed by distance-related pollution damage and congestion externalities. The Ramsey tax component, arising from a relatively inelastic fuel demand, contributes another $0.28/gallon.

The optimal gas tax in Mexico is larger than the estimate reported by PS (2005) for the US (even after updating their results at 2011 prices), and that of Lin and Prince (2009) for California, but lower than that estimated by Parry and Strand (2012) for Chile. In particular, PS (2005) report an optimal tax rate of $1.01/gallon for the US at 2000 prices. This estimate increases to $1.43/gallon at 2011 prices. Lin and Prince (2009) obtain a rate of $1.37/gallon for California at 2006 prices. Finally, Parry and Strand (2012) calculate a corrective fuel tax for Chile of $2.35 per gallon at 2006 prices.

To understand what accounts for the differences with respect to PS (2005), we change each one of the parameters that is different in Mexico than in the US, one at a time. We find that distance-related pollution damage and accident costs explain the majority of the differences. The presence of fuel subsidies (typical of oil-producing countries) explains approximately 20% of the differences. Perhaps surprisingly, the lower fuel efficiency attrib- uted to an older vehicle fleet does not explain a significant proportion of the differences in tax estimates.

Using the optimal gas tax estimate, we address the effects of such a tax across income deciles in Mexico. Contrary to the conventional wisdom, we find that the fuel tax is progressive. The intuition is simple: only 9% of the poorest households demand fuel because the majority of these households (87%) do not own a car. Conversely, 85% of the wealthiest households demand fuel because 91% own at least one car.4

This paper is structured as follows. Section 2 briefly describes the fuel pricing policy in Mexico. Section 3 outlines the model, and Section 4 presents the results. Section 5 compares the results to

those obtained for advanced economies and includes a sensitivity analysis. Section 6 provides concluding remarks.

2. How is the gasoline price set in Mexico?

Mexico is an oil-producing country but currently imports about half of the gasoline demanded in the country, because of capacity constraints. Because of an inappropriate gasoline pricing policy (a pre-determined crawling gas price fixed by the government at the beginning of each year), a subsidy emerges when the international petroleum price is above its pre-determined level, as has been the case most of the time since 2006 (see Fig. 1).5 This subvention, which may produce a final price below its market price, is frequently justified on political, i.e., populist, grounds.

On average, Mexico registers the second-lowest excise tax rate, after the USA, among a sample of representative OECD countries for the period 2001–2011 (see Fig. 2).6 For the period 2007–2011, the tax is in fact negative (a subsidy). This subsidy has cost the government an average of 1.2% of GDP over the period 2007–2011, an amount equivalent to the expenditures on poverty alleviation and public health care programs in the country. These large fiscal subsidies merit careful re-evaluation, especially in this present time of fiscal

Ja n/2

00 0

Ja n/2

00 1

Ja n/2

00 2

Ja n/2

00 3

Ja n/2

00 4

Ja n/2

00 5

Ja n/2

00 6

Ja n/2

00 7

Ja n/2

00 8

Ja n/2

00 9

Ja n/2

01 0

Ja n/2

01 1

Ja n/2

01 2

Ja n/2

01 3

Mexico US

Fig. 1. Gasoline price, 2000–2013 (Mexican pesos per gallon). Sources: US Energy Information Administration and Mexico's Energy Information System. The US price is the weekly Gulf Coast regular conventional retail price. The Mexican price is the regular conventional retail price.

Fig. 2. Excise tax rate (pre-tax fuel price), average 2001–2011. Source: Author calculations from OECD-IEA, Energy Price and Taxes, Quarterly Statistics (2012).

(footnote continued) whereas the rate in high-income countries is the lowest (8.7 per 100,000). In addition, 80% of road traffic deaths occur in middle-income countries, which have 72% of the world population but only 52% of the world's registered vehicles. In fact, nearly 70% of road deaths occur in 13 countries, 12 of which are developing countries (including Mexico).

4 The source is Mexico's Household Income and Expenditure Survey (2010). Unfortunately, the survey does not provide further information that would help to explain the gap between fuel consumption and car ownership across households. The survey only registers whether the household owns a vehicle and monthly total expenditures on fuel; it does not provide information on vehicle use or driving patterns. This gap might be due to a combination of statistical errors and to people either deciding not to use a car (car pooling) or not being able to use a car because it is not working. It is common, especially for people in lower deciles, to own old broken-down cars with the expectation of putting them back into service in the future.

5 Since 2010, the Mexican government has increased the fuel price by $0.016/ gallon every month, which corresponds to an increase of $0.19/gallon per year.

6 US data are not included in Fig. 2 because the OECD-IEA, Energy Price and Taxes, Quarterly Statistics (2012) study does not report excise taxes for this country. Only total gasoline taxes (i.e., excise plus sales taxes) in the US of $0.11/l (the average for the period 2001–2011) are reported.

A. Antón-Sarabia, F. Hernández-Trillo / Energy Policy 67 (2014) 564–571 565

Salvador Ortigueira (University of Miami)

Mexico

Southwest Economy • Federal Reserve Bank of Dallas • Third Quarter 201312

}If the goal of the subsidies is to protect the lowest income brackets, the cost to do so is heavy.

the subsidy is applied per unit of fuel, the data show that 97 percent of the assistance went to the top 80 percent of income earners in Mexico.

This is in line with other countries. The International Energy Agency (IEA) estimated that out of the $409 billion spent globally in 2010 on all subsidies covering consumption of fossil fuels (oil, natural gas and coal as well as the electricity they produced), only $35 bil- lion, or 8 percent, reached the bottom 20 percent of the income distribution.

If the goal of the subsidies is to protect the lowest income brackets, the cost to do so is heavy. Of the $15.9 bil- lion Mexico spent on subsidies in 2011, roughly $15.4 billion went to higher income groups—in other words, it cost $15.4 billion to provide about $500 mil- lion in aid to Mexico’s poorest.

Costly Subsidies Mexico’s subsidies have been

expensive on an absolute dollar basis (Chart 3, left axis).2 While Mexico’s fuel subsidies in recent years were below the peak levels in 2008 (when subsidies exceeded $20 billion), they still exceeded $15 billion in 2011. Preliminary Mexican government data suggest that the 2012 total will be close to the cost in 2011.

Sometimes it’s useful to consider the size of a subsidy relative to gross domestic product (GDP). This method

helps illustrate how big a burden the subsidy might impose on the economy, taking into account the country’s ability to pay. As a share of GDP, Mexican fuel subsidies were at least 1 percent of GDP in four of the last five years for which data are available (Chart 3, right axis). By comparison, expenditures on education amounted to 3.5 percent of GDP in 2010; health spending, 2.8 percent; and pen- sions, 1.2 percent.

Relative to other countries, Mexico typically ranks high in terms of the subsidies’ dollar value. Mexico ranked seventh in such spending in 2011, ac- cording to IEA data. Only Saudi Arabia ($46.12 billion), Iran ($41.39 billion), India ($30.86 billion), Venezuela ($21.97 billion), Iraq ($20.35 billion) and China ($18.45 billion) spent more. However, on the basis of subsidies as a percentage of GDP, Mexico ranked relatively low—19th out of 33 countries in 2011 (Chart 4).

Other Negative Impacts Subsidies work by artificially reduc-

ing prices for fuel, making it relatively cheaper than other goods. Households and firms respond, changing their economic decisions. This introduces distortions in the economy that can hinder performance. For example, households may choose to consume an outsized amount of fuel and to consume less of other goods because of pricing,

Chart

3 Mexico’s Spending on Oil Subsidies Rises Again Billions of U.S. dollars Percent of GDP

1.5

2.5

20112010200920082007 0

Percent of GDP

Subsidy cost

NOTE: The IEA includes subsidies on gasoline, diesel and other oil products, such as LPG (propane). SOURCES: International Energy Agency; International Monetary Fund; authors’ calculations.

Salvador Ortigueira (University of Miami)

Brazil

11 Lessons Learned from Brazil’s Experience with Fossil-Fuel Subsidies and their Reform

Figure 3 Estimated petroleum subsidies or revenue in Brazil and revenue from the Contribuição de Intervenção no Domínio (CIDE) levy on imported petroleum products

Notes: Positive subsidies indicate that consumers were being subsidized (Petrobas was selling products below market values). Negative subsidy values (between 2005 and 2007) show net taxes for consumers.

Source: Calculations by the author.

2.6 Future subsidy levels A group of energy bills were sent by the government to the Congress in September 2009. The main objectives of these bills are to grant the government greater control of the oil sector, increase government oil revenues and enhance the government’s ability to use this revenue for public-policy purposes.

The bills propose to replace the current concession regime by production-sharing agreements. Under these arrangements, the government would receive a share of the oil produced and will use or trade it based on government objectives. A new oil company, 100 per cent state owned, would be created to govern the large oil reserves recently identified offshore. Petrobras will become the single operator for all future oil fields discovered.

The government also intends to create a special fund with its oil revenues. This fund would be used to finance several types of programs (social, environmental, regional, educational and technological, including the support of renewable energy technologies), as well as to assist the development of domestic suppliers of equipment and services for the oil industry.

So far, there is no indication of how the oil fund or the government’s share of the oil supply will be used. Lobbyists in Congress are actively pushing for the creation of new subsidies with the revenues from the oil fund, and the government offered clear signs that it intends to use oil as a major source of revenues for achieving its economic and social objectives.9

9 Ministério de Minas e Energia (Ministry of Mines and Energy): www.mme.gov.br

Subsidies and CIDE Levy

-6

-4

-2

2002 2003 2004 2005 2006 2007 2008

US$ billion

100

120

So u rce: Elab o rated u sin g AN P, EIA, IPEA

US$/bbl

Subsidies CIDE Oil (WTI)

Salvador Ortigueira (University of Miami)

South East Asia

A Guidebook to Fossil-Fuel Subsidy Reform for Policy-Makers in Southeast Asia p.17

FOSSIL-FUEL SUBSIDIES FOR ENERGY CONSUMERS IN SOUTHEAST ASIA

1.2 The Scale of Fossil-Fuel Subsidies for Consumers in Southeast Asia Governments in Southeast Asia subsidize different fuels to varying extents. As shown in Figure 1, according to the International Energy Agency (IEA), Indonesia subsidizes mostly petroleum products and electricity. Malaysia subsidizes all fuel types except for coal. The Philippines have largely removed all energy subsidies, but have preferential taxation provisions for some petroleum products, such as diesel. Thailand subsidizes all energy types, while the bulk of energy subsidies in Vietnam are in the electricity sector.

Figure 1 also shows that Southeast Asia’s subsidy costs fluctuate significantly year by year, regardless of the absolute volume of subsidization or the fuels being subsidized. This is because many subsidy mechanisms do not let domestic consumer prices fluctuate fully in response to international changes; consequentially, when the world price rises, the cost of the subsidy rises too. Figure 1 illustrates this by plotting the average international oil price. The cost of subsidies for oil, gas and coal tends to follow this indicator because world oil prices are used as an index for many gas prices in Asia, and gas prices are, in turn, linked to coal, though with coal prices being the least responsive of the three. Since fossil fuels are the main input for electricity generation in most countries, price changes affect electricity subsidies too.

100

120

2007 2008 2009 2010 2011

Av era

ge pr

ice of

Br en

t c rud

e, US

$ p er

ba rre

Su bs

idi es

, U S$

bi llio

INDONESIA

100

120

2007 2008 2009 2010 2011

Av era

ge pr

ice of

Br en

t c rud

e, US

$ p er

ba rre

Su bs

idi es

, U S$

bi llio

MALAYSIA

100

120

2007 2008 2009 2010 2011

Av era

ge pr

ice of

Br en

t c rud

e, US

$ p er

ba rre

Su bs

idi es

, U S$

bi llio

THE PHILIPPINES

100

120

2007 2008 2009 2010 2011

Av era

ge pr

ice of

Br en

t c rud

e, US

$ p er

ba rre

Su bs

idi es

, U S$

bi llio

n THAILAND

100

120

2007 2008 2009 2010 2011

Av era

ge pr

ice of

Br en

t c rud

e, US

$ p er

ba rre

Su bs

idi es

, U S$

bi llio

VIETNAM

Electricity subsidies

Coal subsidies

Gas subsidies

Oil subsidies

Average price of Brent crude

FIGURE 1 | STRUCTURE OF ENERGY SUBSIDIES IN INDONESIA, MALAYSIA, THE PHILIPPINES, THAILAND AND VIETNAM, 2007–2011

Source: IISD-GSI graphic interpretation of IEA (2012) subsidy estimates, derived with the price-gap method, and BP (2012) oil price data.

Salvador Ortigueira (University of Miami)

South East Asia

A Guidebook to Fossil-Fuel Subsidy Reform for Policy-Makers in Southeast Asiap.18

FOSSIL-FUEL SUBSIDIES FOR ENERGY CONSUMERS IN SOUTHEAST ASIA

1.3 Fiscal Burden and Opportunity Cost As illustrated in Figure 2, the estimated value of fossil-fuel subsidies has been above two per cent of GDP for most of Southeast Asia’s biggest subsidizing countries over the past five years. This can represent a significant fiscal burden for net energy-importing countries that set a fixed price of fuel. In other cases—such as in net exporting countries, where domestic reserves are sold in reference to production cost, or when power sector subsidies are paid for through lack of investment in infrastructure—no fiscal cost is recorded. These off-budget subsidies still represent an opportunity cost, however, that is as real as their on-budget counterparts.

The opportunity cost of fossil-fuel subsidies is the money that is not spent on other priorities such as public transport and infrastructure, or improving health care and education systems. As an illustration, Figure 2 shows fossil-fuel subsidies compared to the value of budgetary deficits and surpluses. The subsidies were larger than the budgetary deficits of Indonesia in 2007–2011; Thailand in 2008, 2010 and 2011; and Vietnam in 2007 and 2011. In all other cases, the subsidies were equal to a considerable share of budgetary deficits.

FIGURE 2 | ENERGY SUBSIDIES AND BUDGETARY DEFICIT OR SURPLUS AS A PERCENTAGE OF GDP IN INDONESIA, MALAYSIA, THE PHILIPPINES, THAILAND AND VIETNAM IN 2007–2010.

Source: IISD-GSI calculations based on IEA (2012) subsidy estimates, derived using the price-gap method, and ADB (2012) data on GDP and budgetary deficits and surpluses.

Fossil-fuel subsidies Budgetary deficit Budgetary surplus

2007 2008 2009 2010 2011

pe rce

nta ge

of GD

INDONESIA

2007 2008 2009 2010 2011

pe rce

nta ge

of GD

MALAYSIA

2007 2008 2009 2010 2011

pe rce

nta ge

of GD

THE PHILIPPINES

2007 2008 2009 2010 2011

pe rce

nta ge

of GD

P THAILAND

2007 2008 2009 2010 2011

pe rce

nta ge

of GD

VIETNAM

Salvador Ortigueira (University of Miami)

South East Asia

Fossil Fuel Subsidies in Asia: Trends, Impacts, and Reforms16

Producer subsidies may have an important impact on investment decisions, and many of the subsidies aimed at producers provided a quasi-consumer subsidy. For example, state-owned electricity suppliers in Indonesia operate at a loss due to controlled consumer prices. The government provides discounted credit and loan guarantees to help finance these losses and infrastructure investment. However, credit subsidies would be unnecessary if electricity providers were permitted to sell electricity at the long-term cost plus profit. A similar situation occurs in upstream sectors in India (coal) and Thailand (natural gas for vehicles) where producers are required to sell product at below-market prices, but are partially compensated through measures that reduce their costs of production or supply.

The detailed inventories of fossil fuel subsidies provide the most comprehensive estimates of subsidies to date in India, Indonesia, and Thailand. The information from these inventories improves transparency on the true level of government finances being used to support fossil fuel consumption and production. Nonetheless, the complex nature of fuel subsidies and their questionable merit warrant an analysis of the potential impacts of subsidy reform.

Figure 7: Breakdown of Total Consumer Subsidies, 2012

Electricity Natural gas Coal Petroleum

27%

15%

57%

49 billion

36 billion

7 billion

30%

3% 10% 87%

68%

India Indonesia Thailand

20$ bi

lli on

Note: Totals may not add up to 100 due to rounding. Source: Authors’ estimates.

Salvador Ortigueira (University of Miami)

South East Asia

Fossil Fuel Subsidies in Asia: Trends, Impacts, and Reforms32

received only 18% of direct subsidy benefits and 19% of indirect benefits. By contrast, the top 20% received 48% of direct benefits and 42% of indirect benefits (Del Granado, Coady, and Gillingham 2012). However, such assessments tend to assume that the benefits of fuel subsidies are transferred to households, whereas in reality, it is often the case that low prices do not reach the intended beneficiaries because of diversion, leakage, and smuggling.

a The study reviewed data from 2005 household expenditure surveys from Bangladesh, Cambodia, India, Indonesia, Pakistan, Thailand, and Viet Nam. The precise proportions of energy products being consumed are likely to have changed since this time, given increasing world oil prices and efforts from a number of countries to encourage the use of LPG, but there is no reason to suppose that the broad proportions have changed significantly.

b A “universal” fossil fuel subsidy is one that is available to the entire population or a large majority of the population, without any attempt to target it to users defined as being “in need." The large majority of fossil fuel subsidies in Asia are universal in nature.

Box 2 continued

Figure 9: Magnitude of Fossil Fuel Subsidies Compared with Social Assistance, 2012

60 49

37 36

3 7 7

India

$ bi

lli on

Indonesia Thailand

Fossil fuel subsidies Social assistance spending

Source: Authors’ estimates.

Salvador Ortigueira (University of Miami)

Are Gasoline Subsidies the Right Policy?Fossil Fuel Subsidy Reform in Indonesia and Mexico © OECD/IEA 2016

Page | 12

Figure 1 • Potential unintended effects of fossil fuel subsidies

Note: CO2 = carbon dioxide.

Source: IEA (2010), World Energy Outlook 2010, www.worldenergyoutlook.org/media/weo2010.pdf.

• Increased price volatility: Fossil fuel subsidies also exacerbate energy price volatility on global markets, by dampening normal demand responses to changes in international prices.

• Black marketeering, smuggling and adulteration: Fossil fuel subsidies may also encourage black marketeering, smuggling and fuel adulteration in the case of oil products, which are easy to transport and store. Fuel shortages and flourishing black markets with high prices are common in countries with where low official prices constrain supply. In certain countries, subsidised kerosene intended for household cooking and lighting is diverted as a diesel substitute due to wide price differentials. Smuggling can also arise, since an incentive is created to sell subsidised products in neighbouring countries where prices are unsubsidised and, therefore, higher. This has been an issue for years in many parts of the world, particularly in Southeast Asia, Africa and the Middle East. The effect in subsidising countries is a substantial financial transfer to smugglers, while recipient countries experience losses from uncollected taxes and excise duties, due to reduced sales in the legitimate market. Removing subsidies would eliminate incentives both to adulterate fuels and to smuggle them across borders. In exporting countries, subsidies reduce the availability of fuels for export by driving up domestic demand. In all countries, fossil fuel subsidies ultimately undermine economic competitiveness and growth.

• Environmental effects: Fossil fuel subsidies can have varying environmental effects. In some instances, for example where subsidies enable poor communities to switch from the traditional use of biomass to modern fuels, they can have positive implications for the local environment by minimising deforestation and household air pollution. In the vast majority of cases, however, fossil fuel subsidies are counterproductive in reaching local and global environmental goals. Subsidised energy prices dampen incentives for consumers to use energy more efficiently, resulting in higher consumption and GHG emissions than would otherwise occur.

• Barriers for clean energy investments: Fossil fuel subsidies undermine the development and commercialisation of renewable energy and other technologies that could become more economically attractive. Even marginal shifts from fossil fuels to renewable energy could help to accelerate the learning effect for renewables and cause unit production costs to decline.

Encourage wasteful consumption

Hasten the decline of exports

Encourage fuel adulteration and smuggling

Discourage investment in energy infrastructure

Disproportionally benefit the middle class and rich

Increase CO emissions and exacerbate local pollution

Dampen global-demand responsiveness to

high prices

Distort markets and create barriers

to clean energy investment

Drain state budgets for importers

Threaten energy security by increasing imports

Typical rationale

Alleviate poverty

Promote economic development

Salvador Ortigueira (University of Miami)

Fuel Subsidies Are Regressive

targeted subsidy. Figure 6 presents the shares of the total benefits that each income group

would receive from subsidized fuel prices, separately for the total, direct, and indirect welfare

impact as well as for the direct benefit for gasoline, LPG, and kerosene. On average, the top

income quintile receives more than six times more in total subsidies than the bottom quintile.

The concentration of subsidy benefits in the hands of the top income groups is even more

pronounced for gasoline and LPG, where the top income quintile receives 27 and 12 times that of

the bottom quintile, respectively. Although the poorest households receive a much higher share

of kerosene subsidies than for other fuel subsidies, there is still substantial leakage of kerosene

subsidies to higher-income groups. Annex Table 1.6 presents the shares of the total benefits that

each income group would receive from subsidized fuel prices, disaggregated by region.

Figure 6. Distribution of Subsidy Benefits by Income Group (Percent of total subsidy benefit)

Source: Authors’ calculations based on results from reviewed studies.

Note: LPG = liquefied petroleum gas. The indirect impact is the welfare impact of higher

prices of goods and services due to an increase in the price of diesel.

The substantial leakage of subsidy benefits to the top income groups means that universal fuel

subsidies are an extremely costly approach to protecting the welfare of poor households. For

example, if we take the poorest 40 percent of households to be the target ‘‘poor” group, the cost

to the budget of transferring one dollar to this group via gasoline subsidies is about 14 dollars.

This occurs because nearly 93 out of every 100 dollars of gasoline subsidy ‘‘leaks” to the top

three quintiles. These leakages are higher in Africa and in Asia and Pacific, where poor

households’ use of gasoline and LPG is comparatively lower than in other regions (Figure 7,

Annex Table 1.6). Even for kerosene, this cost-benefit ratio is about 3 dollars.

Indirect impactTotal impact Direct impact

Gasoline KeroseneLPG

Bottom 2 3 4 Top Quintiles

Salvador Ortigueira (University of Miami)

World Bank Country Classification

Salvador Ortigueira (University of Miami)

January 15, 2017

Salvador Ortigueira (University of Miami)

FY17 Country Classification

1. As of 1 July 2016, the World Bank updated its classification of countries using their GNI per capita in 2015

1.1 Low-income economies are defined as those with a GNI per capita of $1,025 or less

1.2 Lower middle-income economies are those with a GNI per capita between $1,026 and $4,035

1.3 Upper middle-income economies are those with a GNI per capita between $4,036 and $12,475

1.4 High-income economies are those with a GNI per capita of $12,476 or more.

Salvador Ortigueira (University of Miami)

Changes with respect to previous year classification

Economy Old group New group

Cambodia Low Lower middle Equatorial Guinea High Upper middle Georgia Lower middle Upper middle Guyana Lower middle Upper middle Mongolia Upper middle Lower middle Russian Federation High Upper middle Senegal Lower middle Low Tonga Upper middle Lower middle Tunisia Upper middle Lower middle Venezuela, RB High Upper middle

Salvador Ortigueira (University of Miami)

Is GDP a Good Measure of Well-being?

Salvador Ortigueira (University of Miami)

January 12, 2017

Salvador Ortigueira (University of Miami)

Introduction

I Gross Domestic Product (GDP) is the market value of all final goods and services produced within a country in a given time period

I Typically, GDP is used as a measure of economic well-being in cross-country and across periods comparisons

I Many economists have criticized the use of GDP as a measure of economic well-being as it abstracts from

I Leisure

I Mortality

I Inequality, etc.

I So the question is: How good is GDP as a measure of economic well-being?

Salvador Ortigueira (University of Miami)

Two attempts at measuring well-being

I Objective: Construct a welfare measure that combines a more comprehensive list of variables associated with well-being

I Two attempts are: I The United Nations Human Development Index (HDI): This

index combines income, life expectancy and literacy. It first construct sub-indexes each variable and then averages them

I A consumption-equivalent measure of well-being. It combines data on consumption, leisure, inequality and mortality

Salvador Ortigueira (University of Miami)

The consumption-equivalent measure of well-being

I Example: Let us compare the US and France

I The consumption-equivalent measure of well-being in France is equal to the proportion of consumption in the US, given the US values of leisure, mortality and inequality, that would deliver the same expected utility as in France.

Salvador Ortigueira (University of Miami)

The consumption-equivalent measure of well-being

Let us start by looking at summary statistics for the three components of this measure

I Consumption

I Leisure

I Mortality rates

Salvador Ortigueira (University of Miami)

Within-country consumption inequality 2436 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 2016

Consumption.—Figure 1 reports the standard deviation of log consumption across people in our household survey countries. We divide household expenditures equally across people in each household, and add per capita government consumption in the same year from the Penn World Tables 8.0 (Feenstra, Inklaar, and Timmer 2015). We use sampling weights and discount using US survival rates by age, analogous to the way we construct the mean of log consumption in equation (17). The resulting inequality is highest in South Africa, Brazil, and Mexico. Inequality is lower in France, Italy, and the United Kingdom than in the United States.

Leisure.—Figure 2 summarizes annual hours worked per person in our house- hold surveys. Figure 3 reports the standard deviation across people of annual hours worked.12 Hours worked are substantially lower in France, Italy, Spain, and the United Kingdom than in the United States, as has been widely noted. More novel, inequality of hours worked is lower in these same countries than in the United States.

Mortality Rates.—Figure 4 presents life expectancy in years from the World Health Organization for our baseline household survey years. It ranges from 50 in Malawi, the poorest country, to above 75 in the richest countries.

B. Calibration

To implement our welfare calculations, we need to specify the baseline utility function. (In Section IV we will explore a range of robustness checks to our choices

12 Parente, Rogerson, and Wright (2000) argue that barriers to capital accumulation explain some of the varia- tion in market hours worked. Like us, they emphasize that the gain in home production can partially offset the loss in market output. Prescott (2004) attributes some of the OECD differences in hours worked to differences in tax rates, as do Ohanian, Raffo, and Rogerson (2008).

Figure 1. Within-Country Inequality

Notes: The standard deviation of log consumption within each economy is measured from the household surveys listed in Table 1. We use survey-specific sampling weights and US survival rates across ages using an analog of equation (17), with no discounting or growth.

1/64 1/32 1/16 1/8 1/4 1/2 1 0.4

0.5

0.6

0.7

0.8

0.9

Brazil

China Spain

France UK Indonesia India

Italy

Mexico

Malawi

Russia

South Africa

GDP per person

S ta

n d

a rd

d e

vi a

tio n

o f

lo g

c o

n su

m p

tio n

Salvador Ortigueira (University of Miami)

Hours worked 2437JONES AND KLENOW: WELFARE ACROSS COUNTRIES AND TIMEVOL. 106 NO. 9

here.) Following the macro literature, we assume utility from leisure takes a form that implies a constant Frisch elasticity of labor supply (that is, holding the mar- ginal utility of consumption fixed, the elasticity of labor supply with respect to the wage is constant). Since labor supply in our setting is 1 − ℓ , in terms of the utility function in equation (4) this gives v(ℓ) = − θϵ _

1 + ϵ (1 − ℓ) 1+ϵ _ ϵ , where ϵ denotes the

Frisch elasticity. This leaves five parameters to be calibrated: the growth rate g , the

Figure 2. Annual Hours Worked across Countries

Notes: The measure shown here of annual hours worked per capita is computed from the household surveys noted in Table 1, using survey-specific sampling weights and US survival rates across ages as in equation (16), with no time discounting.

1/64 1/32 1/16 1/8 1/4 1/2 1 500

550

600

650

700

750

800

850

900

950

Brazil

China

Spain

France

Indonesia

India

Italy

Mexico

Malawi

Russia

South Africa

GDP per person

A n n u a l h

o u rs

w o rk

e d p

e r

p e rs

o n

1/64 1/32 1/16 1/8 1/4 1/2 1 700

800

900

1,000

1,100

1,200

Brazil

China

Spain

France

Indonesia

India

Italy

Mexico

Malawi Russia

US South Africa

GDP per person

S ta

n d

a rd

d e

vi a

tio n

o f

a n

n u

a l h

o u

rs w

o rk

e d

Figure 3. Inequality in Annual Hours Worked

Note: See notes to Figure 2.

Salvador Ortigueira (University of Miami)

Inequality in hours worked

2437JONES AND KLENOW: WELFARE ACROSS COUNTRIES AND TIMEVOL. 106 NO. 9

1 + ϵ (1 − ℓ) 1+ϵ _ ϵ , where ϵ denotes the

Frisch elasticity. This leaves five parameters to be calibrated: the growth rate g , the

Figure 2. Annual Hours Worked across Countries

1/64 1/32 1/16 1/8 1/4 1/2 1 500

550

600

650

700

750

800

850

900

950

Brazil

China

Spain

France

Indonesia

India

Italy

Mexico

Malawi

Russia

South Africa

GDP per person

A n

n u

a l h

o u

rs w

o rk

e d

p e

r p

e rs

o n

1/64 1/32 1/16 1/8 1/4 1/2 1 700

800

900

1,000

1,100

1,200

Brazil

China

Spain

France

Indonesia

India

Italy

Mexico

Malawi Russia

US South Africa

GDP per person

S ta

n d

a rd

d e

v ia

ti o

n o

f a

n n

u a

l h

o u

rs w

o rk

e d

Figure 3. Inequality in Annual Hours Worked

Note: See notes to Figure 2. Salvador Ortigueira (University of Miami)

Life expectancy 2438 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 2016

discount factor β , the Frisch elasticity ϵ , the utility weight on leisure or home pro- duction θ , and the intercept in flow utility u – .

We choose a common growth rate of 2 percent per year. An alternative would be to try to forecast future growth rates for each country, but such forecasts would have very large standard errors, particularly since we would need forecasts for every year over the next century. With an annual real interest rate of 4 percent in mind, we set the discount factor to β = 0.99 . Recall that there is already additional discounting inherent in the expected utility calculation because of mortality. A 4 percent real interest rate is consistent with the standard Euler equation with log preferences, 2 percent consumption growth, roughly 1 percent discounting for mortality, and 1 percent from the discount factor.

Surveying evidence such as Pistaferri (2003), Hall (2009a, b) suggests a bench- mark value for the Frisch elasticity of 0.7 for the intensive (hours) margin and 1.9 for the extensive and intensive margins combined. Chetty (2012) reconciles micro and macro estimates of the Frisch elasticity and recommends a value of 0.5 or 0.6 for the intensive margin. We consider a Frisch elasticity of 1.0 for our benchmark calibration, which implies that the disutility from working rises with the square of the number of hours worked. As we discuss in the robustness section, the results are not sensitive to this choice.

To get the weight on the disutility from working, θ , recall that the first-order condition for the labor-leisure decision is u ℓ / u c = w(1 − τ) , where w is the real wage and τ is the marginal tax rate on labor income. Our functional forms then imply θ = w (1 − τ) (1 − ℓ) −1/ϵ /c . For our benchmark calibration, we assume this first-order condition holds for the average prime-age worker (25–55 years old) in the US Consumer Expenditure Survey (CE) in 2006. We take the marginal tax rate in the United States from Barro and Redlick (2011), who report a value of 0.353 for 2006. Taking into account the ratio of earnings to consumption and average leisure

1/64 1/32 1/16 1/8 1/4 1/2 1 GDP per person

Brazil China

Spain France

Indonesia

India

Italy

Mexico

Malawi

Russia

South Africa L if e

e x p

e c ta

n c y

Figure 4. Life Expectancy

Note: Life expectancy at birth in each country is measured as the sum over all ages of the probability of surviving to each age, using life tables from the World Health Organization.

Salvador Ortigueira (University of Miami)

Welfare across countries and over time 2440 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 2016

welfare measures 22 percent higher than their incomes. The remaining countries, in contrast, have welfare levels that are typically 25 to 50 percent below their incomes. The way to reconcile these large deviations with the high correlation between wel- fare and income is that the “scales” are so different. Incomes vary by more than a factor of 64 in our sample, i.e., 6,300 percent, whereas the deviations are on the order of 25 to 50 percent.

KEY POINT 2: Average Western European living standards appear much closer to those in the United States when we take into account Europe’s longer life expec- tancy, additional leisure time, and lower levels of inequality.

Figure 5. Welfare and Income across Countries

1/128

1/64

1/32

1/16

1/8

1/4

1/2

Brazil

China

Spain

France UK

Indonesia

India

Italy

Mexico

Malawi

Russia

South Africa

GDP per person (US = 1)

W elf

are , λ

Panel A. Welfare and income are highly correlated at 0.98

1/64 1/32 1/16 1/8 1/4 1/2 1

1/641/128 1/32 1/16 1/8 1/4 1/2 1

0.5

0.75

1.25

1.5

Brazil China

Spain

France

Indonesia

India

Italy

Mexico Malawi

Russia

South Africa

Th e r

ati o o

f w elf

are to

in co

Panel B. But this masks substantial variation in the ratio of λ to GDP per capita

Salvador Ortigueira (University of Miami)

Welfare across countries 2442 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 2016

penalized less.15 As we will show, this is a key place where the equivalent variation differs from the compensating variation. The compensating variation weights differ- ences in mortality by US flow utility. In the robustness section, we’ll see this leads to much larger welfare differences.

A second reason that welfare is lower than income in several countries is that average consumption—as a share of income—is low relative to the United States. Utility depends on consumption, not income. Of course, an offsetting effect is that the low consumption share may raise consumption in the future. To the extent that countries are close to their steady states, this force is already incorporated in our calculation. However, in countries with upward trends in their investment rates, our calculation will understate steady-state welfare. China is an obvious candidate for this qualification, though correcting for this has a modest effect.16

15 Table A3 in the online Appendix reports the implied value of life in each of our 13 countries. 16 See Table 8 of Jones and Klenow (2010).

Table 2—Welfare across Countries

Decomposition

Welfare λ Income log ratio Life exp. C/Y Leisure Cons. ineq.

Leis. ineq.

US 100.0 100.0 0.000 0.000 0.000 0.000 0.000 0.000 77.4 0.897 877 0.538 1,091

UK 96.6 75.2 0.250 0.086 −0.143 0.073 0.136 0.097 78.7 0.823 579 0.445 826

France 91.8 67.2 0.312 0.155 −0.152 0.083 0.102 0.124 80.1 0.790 535 0.422 747

Italy 80.2 66.1 0.193 0.182 −0.228 0.078 0.086 0.075 80.7 0.720 578 0.421 905

Spain 73.3 61.1 0.182 0.133 −0.111 0.070 0.017 0.073 79.1 0.786 619 0.541 904

Mexico 21.9 28.6 −0.268 −0.156 −0.021 −0.010 −0.076 −0.005 74.2 0.879 906 0.634 1,100

Russia 20.7 37.0 −0.583 −0.501 −0.248 0.035 0.098 0.032 67.1 0.733 753 0.489 1,027

Brazil 11.1 17.2 −0.436 −0.242 0.004 0.005 −0.209 0.006 71.2 0.872 831 0.724 1,046

S. Africa 7.4 16.0 −0.771 −0.555 0.018 0.054 −0.283 −0.006 60.9 0.887 650 0.864 1,093

China 6.3 10.1 −0.468 −0.174 −0.311 −0.016 0.048 −0.014 71.7 0.658 888 0.508 1,093

Indonesia 5.0 7.8 −0.445 −0.340 −0.178 −0.001 0.114 −0.041 67.2 0.779 883 0.445 1,178

India 3.2 5.6 −0.559 −0.440 −0.158 −0.019 0.085 −0.028 62.8 0.785 918 0.438 1,143

Malawi 0.9 1.3 −0.310 −0.389 0.012 −0.020 0.058 0.028 50.4 0.923 934 0.533 997

Notes: The table shows the consumption-equivalent welfare calculation based on equation (19). See Table 1 for sources and years. The second line for each country shows life expectancy, the ratio of consumption to income, annual hours worked per capita, the standard deviation of log consumption, and the standard deviation of annual hours worked, all computed from the cross-sectional micro data, with no discounting or growth.