Page images
PDF
EPUB

contents. Raising the prices of fuels and energy-intensive products would discourage all fossil fuel uses in proportion to their carbon contents and encourage development of less carbon-intensive alternatives. A recent statement by leading economists points out that a tax mechanism would be much more efficient than a regulatory approach (Economists' Statement on Climate Change, 1997).

An alternative proposal under serious consideration in the United States is a tradable permits program, in which permits would be required in order to sell or use fossil fuels. By limiting the total number of permits, the regulatory authority could control carbon emissions. If the government allowed permits to be bought and sold, the program would create efficient incentives like those of a carbon tax, because the permit price in the marketplace would signal how much firms should reasonably spend on abatement measures. If the government initially distributed the permits through an auction, it could mitigate adverse economic impacts by using the revenues to reduce other taxes without increasing fiscal deficits. In this sense, auctioned-off tradable permits to sell or use fossil fuels have economic implications similar to those of a carbon tax. (In this report, statements about the effects of a carbon tax apply equally to the impacts of tradable carbon permits that are auctioned off.)

Though one argument for tradable permits is the perceived political

difficulty of proposing a change in the tax structure, the tradable permits approach also faces potential difficulties. It would be less efficient than a revenue-neutral tax and would encounter political opposition if valuable permits were given away to energy companies and utilities. Moreover, it would be difficult to include small fuel users in a tradable permits program, though their aggregate energy use is important, without creating administrative burdens much greater than those implied by raising energy taxes. Furthermore, if new scientific information necessitated further emissions reduction, canceling carbon permits that had been purchased in an auction or market transaction would be more difficult than raising a carbon tax.

Many interest groups claim that a carbon tax or any other efficient policy to reduce carbon emissions, such as a tradable permits policy, would impose high economic costs and reduce economic growth. For support, they point to simulations with economic models, some of which have suggested that stabilizing CO2 emissions at 1990 levels could require a tax of up to $430 per ton of carbon by 2030 and could impose total costs of up to 2.5 percent of annual gross domestic product (GDP) (Charles River Associates, 1997). Of course, other economic models predict that similar emissions reductions could be achieved with far smaller energy taxes and negligible, or even favorable, overall impacts on the economy (Gaskins and Weyant, 1993).

[merged small][ocr errors][ocr errors]

Which predictions should we believeif any? Interested groups on different sides of the issues have their own preferred models (and modelers), and these tend to produce simulations supporting the policy positions of their respective sponsors. The underlying economic models may contain dozens of complicated equations that are nearly impenetrable to all but trained econometricians. How and why such models reach the predictions that they do is hard to comprehend. Yet, it matters greatly what the economic impacts of policies to reduce the long-term risks of global warming will be.

This report provides a guide for the perplexed—an explanation in simple terms of the key assumptions in the models being used to simulate the

economic effects of carbon taxes or similar policies to control carbon dioxide emissions. The report also provides a quantitative analysis of 16 widely used models, demonstrating how key assumptions affect the predicted economic impacts of reaching CO2 abatement targets. It turns out that despite the complexity of the models, only a handful of easily understandable assumptions are important in determining the simulation results. By showing the effect of these assumptions on the predicted economic costs, not just in one particular model but in all of them, this report can help readers to apply their own judgments about which models are more realistic and to reach their own conclusions about which economic predictions are more credible.

MODELS, ASSUMPTIONS, AND CONCLUSIONS

a

An economic model is no more than a coherent set of assumptions about the structure and functioning of the economy. A model is used to predict the consequences of some change, often a policy change like the imposition of carbon tax. Naturally, the prediction depends entirely on the assumptions imbedded in the model-how could it be otherwise? Many of the assumptions of an economic model are simplifications, adopted to make the model easier to analyze or to compute. Modelers hope that in making these simplifying assumptions the baby is not disappearing along with the bathwater, but, alas, that is not always so. Many are based on empirical studies, often quite sophisticated, of particular relationships in the economy, and the modeler hopes not only that the relationship has been described accurately but also that it will continue in the future as it was in the past.

Many people are critical of the assumptions economists make but none more so than economists themselves. Typically, economic modeling of important issues is subjected to widespread and intense scrutiny within the economics profession, and unrealistic assumptions tend to be identified,

improved, or discarded. Just as climate scientists and modelers over the past decade have criticized and improved the atmospheric models linking greenhouse gas emissions to changes in climate, so have economists improved the modeling of the economic impacts of a carbon tax. There has been prolonged economic debate and significant intellectual progress in making the models used for economic simulation more realistic. This report reflects some of that intellectual history.

Two kinds of assumptions in the models are critical: those that largely determine the predicted economic costs of abating carbon emissions and those relating to the economic benefits from forestalling environmental impacts from fossil fuel emissions. With respect to the costs of limiting carbon emissions, the key assumptions are

1. the extent to which substitution among energy sources, energy technologies, products, and production methods is possible;

2. the extent to which market and policy distortions create opportunities for low-cost (or no-cost) improvements in energy efficiency;

Just as climate scientists and modelers over the past
decade have criticized and improved the atmospheric
models linking greenhouse gas emissions to changes
in climate, so have economists improved the modeling
of the economic impacts of a carbon tax.

1

[blocks in formation]

Predictions of the economic impacts of climate protection policies have been made on the basis of two main kinds of economic analyses, commonly referred to as 'top-down' and 'bottom-up' models. Top-down models are aggregate models of the whole economy that represent the sale of goods and services by producers to households and the reciprocal flow of labor and investment funds from households to industries. Models used for policy simulations also describe the role of government in imposing taxes, transferring income, and purchasing goods and services. Computable general equilibrium (CGE) models depict the formation of market-clearing prices in the process of matching the demand for goods and services from users to their supply by producers. Demand and supply conditions in such models are based on assumptions that consumers and producers allocate their resources to maximize their welfare or profits, respectively. However, demand and supply conditions are typically based on statistically estimated relationships observed in the past. Optimizing models derive their dictions by explicitly maximizing some assumed mathematical formula representing household welfare as a function of present and future consumption. Such general equilibrium models assume that households and industries will eventually respond efficiently to any policy change, though some models describe irreversible investment decisions and imperfect foresight regarding future prices that serve to delay the adjustment process.

pre

In contrast, macroeconomic models predict economic behavior from statistically estimated relationships among economic variables in the past. Although such relationships are developed from accepted economic theory, macro models do not derive the predicted response to a policy shift from an explicit assumption that firms and households respond efficiently or with accurate foresight. Because they are estimated from actual macroeconomic behavior, they can reflect the short-term adjustment costs in response to an unexpected economic policy change, including business cycles, inflation, and unemployment. Macro and CGE models can be complementary in predicting short-run and longrun responses to a policy change. Moreover, modelers have learned to combine features of both (Hourcade and Robinson, 1996; Shackleton et al., 1992).

Top-down models used to analyze climate policies emphasize interactions between the energy sector and the rest of the economy. A tax-induced change in the price

of carbon fuels directly affects demand and supply for energy. and indirectly affects other markets for commodities, labor, and capital. Therefore, consumer prices, incomes, savings, and labor supply are also affected, resulting in new levels of GDP, investment, and future growth. These can all be compared to baseline projections. More detailed analyses offer insights into distributional incidence on particular industries and income groups. When key assumptions are standardized among models, the range of their predictions narrows (Gaskins and Weyant, 1993).

Bottom-up analyses examine the technological options for energy savings and fuel-switching that are available in individual sectors of the economy, such as housing, transportation, and industry. Information on the costs of these options in individual sectors is then aggregated to calculate the overall cost of achieving a reduction in CO2 emissions. In contrast to top-down models, in which the scope for technological substitution is extrapolated from past experience, bottom-up analyses estimate possibilities by considering explicitly the actual technologies that firms could profitably adopt at various energy price levels. Bottom-up analyses tend to be more optimistic about the scope for cost-effective energy savings.

To some extent, this optimism comes from overlooking important barriers to implementation, such as management and retraining time, risk-aversion toward unproven technologies, capital constraints, household preferences, or lack of information (Boero et al., 1991). Top-down models based on past rates of substitution and technological change implicitly incorporate such effects. Moreover, bottom-up analyses do not deal as adequately with overall macroeconomic constraints on capital availability or market demand as top-down models do. Despite these limitations, bottomup analyses have highlighted energy inefficiencies and technological opportunities. Some top-down climate models have adopted features of bottom-up analyses by incorporating detailed descriptions of technological options for energy supply, conversion, and use.

Assumptions about the use of revenues generated from carbon taxes or from auctioning off carbon permits are crucial.

3. the likely rate of technological innovation and the responsiveness of such change to price signals;

4. the availability and likely future cost of non-fossil, backstop energy

sources;

5. the potential for international 'joint implementation' of emissions reductions; and

6. the possibility that carbon tax revenues would be recycled through the reduction of economically burdensome tax rates.

Top-down models that assume limited substitution, slow technological change that does not respond to price signals, limited, expensive, or no availability of non-fossil energy sources, and the absence of international cooperation in achieving emissions reductions at least cost unfailingly predict that the economic costs of achieving any given carbon abatement target will be high. At the other extreme, bottom-up analyses that embody optimistic assumptions about the potential availability and rapidity of cost-effective, energy-saving technological improvements, and that neglect capital and other resource constraints can be counted on to predict low abatement costs.

In addition, assumptions about the use of revenues generated from carbon taxes or from auctioning off carbon permits are crucial. These revenues can be used to offset reductions in revenues if rates are cut on other taxes that are economically burdensome,

without raising fiscal deficits. Many existing taxes on incomes, profits, and payrolls discourage savings, work, or investment by lowering after-tax returns to those activities. Economic studies suggest that lowering marginal tax rates for such existing forms of taxation and making up the revenue through a carbon tax would lessen the economic impacts of achieving a carbon abatement target. However, many early economic modeling simulations assumed that revenues from a carbon tax would not be used in this way but somehow returned in arbitrary "lumpsum" distributions to households, with no effect on incentives to work, save, or invest.

The final set of key assumptions concerns the environmental damages a carbon tax would avoid. Though averting these potential damages is the rationale for a carbon tax, most economic models are not constructed in ways that can take these damages into account. However, some models have factored in two types of savings:

1. avoiding the economic damages from climate change (the 'climate benefits'); and

2. reducing other air pollution dam

ages associated with the burning of coal and other fossil fuels (the 'nonclimate benefits').

The impact of climate change on the U.S. economy is a matter of great uncertainty, with predictions ranging from potential disasters-floods, hurricanes, droughts, and pestilence—-to

« PreviousContinue »