Journal Issue

Combatting the “Greenhouse Effect”

International Monetary Fund. External Relations Dept.
Published Date:
January 1992
  • ShareShare
Show Summary Details

As countries worldwide debate how strongly to respond to the threat of global warming, cutting energy subsidies and levying carbon taxes rank high on the list of possible tools. Recent World Bank studies show that getting energy prices right makes good economic as well as environmental sense, and a modest domestic carbon tax would be appropriate for a large number of countries.

We are still no closer to answers on whether the buildup of “greenhouse” gases in the atmosphere could trigger a significant warming of the earth’s surface. Nor do we know with any precision what the social and economic consequences of global warming might be. Yet countries all over the world are increasingly showing a willingness to act now, rather than wait for further scientific evidence:

• At the June 1992 “Earth Summit” in Rio de Janeiro, more than 150 countries signed a treaty aimed at stabilizing greenhouse gases—mainly carbon dioxide (CO2), which is by far the largest contributor to the greenhouse effect.

• Most industrial countries have adopted national targets of stabilizing CO2 emissions at 1990 levels by the year 2000.

• Between January 1990 and January 1991, Finland, the Netherlands, and Sweden imposed “carbon taxes”—that is, taxes on the carbon content of fossil fuels.

• The European Community (EC) is considering levying a Community-wide tax on both carbon and energy in general.

• Many developing countries are beginning to weigh the various options, because although industrial countries account for around 43 percent of total CO2 emissions from fossil fuel consumption, the developing world is fast contributing an increasing proportion.

But given the uncertainty surrounding the likelihood and possible consequences of a global warming, what steps should nations take that they could at least justify on economic or local environmental grounds? This is the question the Bank tried to answer in two recent studies, one on the potential benefits of removing energy subsidies and the other on the desirability of carbon taxes.

Pricing energy properly

Economists have long argued that subsidizing energy distorts consumption and production decisions, hurts government revenues, and harms efforts to protect the environment. Yet such subsidies are pervasive in emerging market economies and developing nations. How much do these subsidies add up to? Our study shows that global annual fossil fuel subsidies exceed $230 billion—roughly 25 percent of the total value of world fossil fuel consumption at world prices—with subsidies to petroleum products accounting for 55 percent of total world subsidies; coal, 23 percent; and natural gas, 22 percent.

The biggest subsidizers in dollar terms are the states of the former U.S.S.R. (accounting for as much as 75 percent of world subsidies), China, and Poland (see Chart 1). Indeed, domestic prices are only 10 percent of world prices in the states of the former U.S.S.R., while domestic prices are much higher than world prices in industrial countries (see Chart 2). Other large subsidizers include a number of energy exporters (Venezuela, Mexico, Indonesia, Saudi Arabia, South Africa, Argentina, and Egypt) and a few energy importers (Czechoslovakia, Brazil, and India.) Some smaller energy exporting nations (Ecuador and Peru) also subsidize most petroleum products. But most smaller Central and South American, as well as a number of Asian nations (Myanmar, the Philippines, Sri Lanka, and Thailand), tend to tax gasoline and price fuel oils close to world prices.

Chart 1How much do we subsidize fossil fuels?

Chart 2How large are energy subsidies?

(Ratio of domestic price to world price)

1 For developing countries, selected petroleum products are mainly fuel oils (gasoline is often taxed and in those cases not included). For industrial countries, however, all petroleum products are included. Petroleum products: ALL-weighted average of all petroleum products; F-fuel oils: K-kerosene; L-liquified petroleum gas; G-gasoline; D-automotive diesel.

2 Difference between domestic and world prices is due to both taxes and other distortions.

If these huge subsidies were removed, would it really make that much difference for greenhouse gas emissions? The answer is yes, although not by as much as many might hope. We found that the removal of fossil fuel subsidies could reduce annual carbon emissions by nearly one third in the states of the former U.S.S.R. in about ten years (see Chart 3). For most developing countries, however, the reduction would be more moderate, around 10-20 percent. In total, global emissions would drop about 9 percent. But this drop in consumption would be expected to reduce world prices of fossil fuels, inducing increased consumption and higher carbon emissions in nonsubsidizing countries by as much as 4 percent of global emissions. This means that net global carbon emission reductions from the removal of subsidies would be about 5 percent.

Chart 3Removing energy subsidies would sharply cut carbon emissions

Source: Larsen and Shah,” World Fossil Fuel Subsidies and Global Carbon Emissions.”

This 5 percent figure may not seem like much, but our calculations show that industrial countries would have to adopt a $60-$70/ton carbon tax to achieve an equivalent global reduction. Such a tax would reduce emissions by about 20 percent in industrial countries, or 9 percent of global emissions, ignoring changes in world energy prices. But if the price changes were included, the emission reduction in industrial countries would be only 13 percent, or 5 percent of global emissions.

As for the welfare implications—that is, the impact of relative price changes on real incomes—both subsidizing and nonsubsidizing countries would be affected. This would occur because of (1) the improved efficiency of production and consumption decisions from subsidy removal, and (2) the expected fall in world prices if subsidy removal induced a sufficiently large reduction in consumption. The first component is expected to be positive for all countries, and the second would be positive for importers but negative for fossil fuel exporters. Thus, if losses in export revenues were larger than the gains stemming from more efficient production and consumption decisions, the net welfare effect would be negative for exporters.

Our study shows that the states of the former U.S.S.R. would see their welfare increase by $22 billion, on top of local and global environmental gains from reduced fossil fuel consumption. The large net fossil fuel importers—Brazil, Czechoslovakia, India, and Poland—would see their welfare increase by more than $2 billion. Other importing developing countries would also experience welfare gains, because of lower import prices. Of the net exporters, welfare effects in Saudi Arabia would be a negative $1.5 billion, and in the range of a negative $150-$350 million in Egypt, Indonesia, Mexico, and Venezuela. But Japan, the United States, and Western Europe would see their welfare increase by $15 billion, because reduced fossil fuel prices imply lower import prices.

In sum, adjusting subsidized energy prices to world prices is the first step in an environmentally responsible and fiscally prudent development strategy. The emerging market economies and developing nations that adopt such policies sooner rather than later will reap rich economic gains, besides helping the global environment.

Global carbon tax

Given that carbon emissions contribute nearly three fourths of the long-run warming potential from all greenhouse gases, a global carbon tax has been advocated in recent years as the single most important way for the international community to combat global climate change. The arguments in its favor are several. When properly implemented, it can be a low-cost insurance against a potentially large environmental threat; it is more desirable than an energy tax, because it better targets carbon emissions; and it would represent a more flexible, lower-cost alternative to regulatory responses—in fact, some studies show that there would be substantial cost savings of a move from regulatory responses to market-based incentives.

But most global carbon tax regimes in practice appear unworkable, in view of the difficulties of getting nations to agree on (1) a common framework for determining the level and base of such a tax; and (2) institutional arrangements for tax collection and administration, as well as revenue distribution among nations.

Low carbon taxes make sense(costs and benefits of a $10/ton carbon tax for selected countries, 1987)
PakistanIndonesiaIndiaUnited StalesJapan
Fossil fuel consumption (million dollars)3,3455,33017,266246,502126,100
Carbon emissions
(millions of tons)13251481,24623
(tons per capita)
Price of carbon (per ton in dollars)253200117198538
Energy taxes (dollars/ton of carbon)6501127105
$10/ton carbon tax revenues (million dollars)1322661,48212,4612,371
Price increase from $10/ton carbon tax (percent)
Petroleum products36231
Natural gas34341
Efficiency costs of a $10/ton carbon tax as percent of carbon tax revenues:
Case A. Revenue neutral change by equal yield reductions in personal income taxes-17.5-1.5-8.7-8.4-11.4
Case B. Revenue neutral change by equal yield reductions in corporate income taxes+9.0+8.7+16.9-6.2+9.0
Case C. Raising additional revenues with no change in existing taxes-17.7-1.5-8.8-10.2-12.3
Case D. Raising additional revenues with no change in existing taxes but accounting for subsidies-17.7+0.40.0-10.2-12.3
Efficiency costs of carbon reductions (dollars/ton)
(efficiency costs in Case C to tons of carbon reductions)39471479
Benefit-cost ratios associated with the impact of carbon taxes on local pollutants:1
Source: Shah and Larsen, “Carbon Taxes, the Greenhouse Effect, and Developing Countries”.

Includes sulfer dioixides, nitrous oxides, and particluate matters. “High” is based on a Norwegian study by Glomsrod et al (1990): “Medium” it based on a US study by Bernow and Marron (1990); “Low” is based on a US study by the US Environment Protection Agency and Energy and Resource Consultants, Inc. (the “low” study does not include chronic effects of nitrous oxides emissions).

Source: Shah and Larsen, “Carbon Taxes, the Greenhouse Effect, and Developing Countries”.

Includes sulfer dioixides, nitrous oxides, and particluate matters. “High” is based on a Norwegian study by Glomsrod et al (1990): “Medium” it based on a US study by Bernow and Marron (1990); “Low” is based on a US study by the US Environment Protection Agency and Energy and Resource Consultants, Inc. (the “low” study does not include chronic effects of nitrous oxides emissions).

Thus, the prognosis for the acceptance of a global carbon tax regime is slim, given the high degree of uncertainty surrounding global warming, the anticipated trade-offs with growth, and the inability of a uniform tax to take into account the differing local pollution problems at the sectoral and national levels. The critical question then becomes whether a strong enough case can be made for the adoption of a national carbon tax on economic and local environmental grounds.

National carbon tax

The idea of imposing a national tax on carbon—the first tax ever to be aimed specifically at global warming—has found increasing favor of late, perhaps because governments would welcome a politically popular way of raising revenue. Such a tax would fall most heavily on the fossil fuel with the greatest carbon intensity per unit of energy. This would mean a relatively higher percentage price increase for coal than for alternate fuels (petroleum and natural gas), because coal not only has the highest carbon content per unit of energy but it also has the lowest price. For example, a $50/ton carbon tax in Western Europe would on average increase end-user prices of coal by 35 percent, natural gas to households by 8 percent, and gasoline by 5 percent. This compares with price increases in the United States of 90 percent for coal, 13 percent for natural gas, and 12 percent for gasoline.

The carbon-cum-energy tax now being considered by the EC would start at $3/barrel of oil equivalent and increase by $1/barrel each year in real terms until it reached $10/barrel (roughly equivalent to a carbon tax of $70/ton) in the year 2000. Since the tax is supposed to be “fiscally neutral,” governments would be expected to lower other taxes by a similar amount. But the chances of the EC adopting this tax anytime soon look poor, as the EC might well wait for others (the United States and Japan) to act, worrying about being at a “competitive disadvantage”—although to date no empirical work has been done to support such a fear.

Revenue potential. The revenue potential of carbon taxes is extremely large. Indeed, a moderate $10/ton carbon tax, if imposed individually by all nations, could raise $55 billion just in the first year. For some countries (China), such revenues could account for about 2 percent of GDP, enough to wipe out the central government’s budgetary deficit. In countries where 1987 per capita GDP was less than $900, such a tax would yield revenues worth an average of more than 1 percent of GDP and 5.7 percent of government revenues. For the OECD countries, the numbers are lower, but still significant—0.21 percent of GDP and 1 percent of government revenues. Moreover, carbon taxes tend to be easier to administer in developing countries than personal and corporate taxes and thus less prone to tax avoidance and evasion.

Distributional implications. The literature on industrialized countries typically portrays carbon taxes as regressive, because outlays on fossil fuel consumption as a proportion of current annual income falls with income. But a recent study (by MIT Professor James Poterba) using US data shows that carbon taxes are considerably less regressive relative to annual consumption expenditures than to annual income, at least for industrial countries. Moreover, there is reason to believe that the same holds true for the rest of the world, although for quite a different reason: in developing countries, institutional factors also play an important role.

Auguring for progressivity, or at least low regressivity in the developing world, are:

• If there is a significant degree of foreign direct investment from countries where investors are allowed foreign tax credits against domestic liabilities, then a significant tax burden could be passed on to foreign treasuries, producers, and consumers;

• If price controls apply, producers often cannot pass the tax on to consumers in terms of higher prices;

• With binding import quotas or rationed foreign exchange, a tax would reduce the excess profits made by the privileged class.

Auguring for regressivity would be factors such as full market power. In this situation, the producer could increase product prices to fully pass on to consumers the carbon tax.

As it turns out, in most developing countries, there is some combination of the above elements, creating a situation where taxes can be only partially shifted to consumers. This means that a carbon tax would either be progressive or much less regressive than most people believe. Further, it is likely to be regressive only for the lowest income groups, which could be protected through direct subsidies or alternate measures. In addition, the overall tax structure could be made even less regressive by using the carbon tax to reduce personal income taxes (the latter are not necessarily progressive because of high tax evasion and urban-rural migration).

Efficiency costs. By their very nature, taxes in general distort decision making, creating losses for both consumers and producers. But can carbon taxes contribute to economic efficiency by replacing other, more distortionary taxes? To answer this question, we examined four scenarios, using a sample of five countries (India, Indonesia, Japan, Pakistan, and the United States). Welfare loss (efficiency costs) was measured in terms of additional income required to make the consumer indifferent to a package of policy change, not counting any benefits of reduced global warming or local pollutant reductions.

Case A: Revenue neutral introduction of a $10/ton carbon tax that displaces a personal income tax. In this case, we found an overall deterioration in welfare, ranging from a low of 1.5 cents per dollar of carbon tax revenues in Indonesia to a high of 17.5 cents in Pakistan (see table). The differences stem mainly from variations in pre-existing fossil fuel taxes, labor income taxes, carbon prices (market value of total fossil fuel consumption divided by carbon emissions), and energy price changes from the carbon tax.

The study indicated that a country with a relatively low level of pre-existing energy taxes (Indonesia) would do much better than a country with a high one (Pakistan). Although energy taxes are even higher in Japan than in Pakistan, welfare loss is lower in Japan—mainly because of lower carbon reductions, which would stem from the very high pre-existing carbon price. The loss for India compares well with that for the United States, even though pre-existing taxes in India are much lower, since India has a high consumption of coal and thus its carbon price is only half that of the United States.

Case B: Revenue neutral introduction of a $10/ton carbon tax that partially replaces a corporate income tax. Under this scenario, with the major exception of the United States, the sampled countries would be better off—lending support to a widely held view that corporate income taxes are far more distortionary than labor income taxes, because higher corporate income taxes reduce investment. The estimated net welfare gain varies from a high of 16.9 percent of carbon tax revenues for India to a low of 8.7 percent for Indonesia. The US welfare loss—about 6.2 percent of carbon tax revenues, or 0.017 percent of GDP—may result from the country’s relatively lower marginal taxation of corporate income.

Case C: Introduction of a $101 ton carbon tax to raise additional revenues, with no change in existing taxes. In this case, we assume that tax revenues are redistributed as a lump sum. Here, again, we find a deterioration in welfare. But the losses—ranging from a low of 1.5 cents per dollar for Indonesia (0.005 percent of GDP) to a high of 17.7 cents per dollar for Pakistan (0.07 percent of GDP)—although significant, represent only a small fraction of carbon tax revenues. The welfare losses for India, Indonesia, and Pakistan are only slightly higher than those obtained in Case A, because income tax evasion in these countries is pervasive, meaning that changes in the income tax would have little bearing on the labor supply. The difference in the two cases is larger for the United States and Japan because of the higher pre-existing labor income taxes and higher tax compliance.

Case D: Introduction of a $10/ton carbon tax to raise additional revenues, with no change in existing taxes but accounting for fossil fuel subsidies. In this scenario, the outcome depends on the level of existing subsidies, underscoring the point that efficiency costs of carbon taxes are overstated if subsidies are ignored. By incorporating existing energy subsidies into the analysis, the tax would not hurt aggregate welfare in India and Indonesia, where fossil fuel price subsidies are high.

Thus, these four scenarios show that a case for carbon taxes on efficiency grounds alone rests upon introducing them in a revenue neutral manner, either as a replacement for corporate income taxes or when fossil fuel subsidies exist. Such taxes do not fare so well as a substitute for personal income taxes, at least in countries with pre-existing energy taxes and no subsidies.

Local environmental benefits. Economic benefits aside, however, can carbon taxes be justified on local environmental grounds, particularly through the reduced emissions of pollutants such as nitrous oxides, carbon monoxides, particulates (soot and smoke), and sulphur dioxides? To answer this, we compared the efficiency costs of carbon taxes to the potential health benefits of local pollutant reduction in the same five countries. We determined the benefits by using estimates from three studies, adjusted by purchasing power parity indexes to afford a rough measure of comparability across countries.

Here, too, we found evidence that developing and developed countries alike stood to gain, especially if they already had low or nonexistent energy taxes. Our calculations suggest that Indonesia, India, and the United States could benefit substantially from a carbon tax on local environmental grounds alone, as benefits exceed costs under all three health benefit estimates (see table). For Pakistan and Japan, however—in view of high pre-existing energy taxes and thus high welfare costs of carbon taxes—the benefit-cost ratios are significantly lower.

Policy options for the 1990s

Faced with the uncertainty of global warming and the economic costs that concrete actions to stem global change would impose on this and future generations, policymakers must find tools that can be justified purely on economic and local environmental considerations. For any nation with energy subsidies, the first priority must be the elimination of these subsidies. This would release $230 billion in revenues for developing and emerging market economies, reduce global carbon emissions by 5 percent, improve the allocation of economic resources, and make scarce public funds available for development projects.

For countries that have eliminated energy subsidies but have little or no taxation of energy, or for countries that cannot overcome political obstacles to removing the subsidies, small carbon taxes would be in order. Such taxes would discourage fossil fuel consumption and reduce local pollution, while raising large amounts of revenues at lower administrative costs than the prevalent taxes entail. They would also impose lower economic costs for the society as a whole than do corporate income taxes, without unduly hurting the poor.

Thus, eliminating energy subsidies and levying small carbon taxes are shaping up to be critical elements of an environmentally responsible development strategy for the 1990s.

This article is based on papers by the authors prepared as background for the Bank’s World Development Report 1992. For a fuller discussion, see “Carbon Taxes, the Greenhouse Effect and Developing Countries,” (World Bank Policy Research Working Paper No. 957, 1992) and “World Fossil Fuel Subsidies and Global Carbon Emissions,” (No. 1002), available from the World Development Report office.

Other Resources Citing This Publication