Climate Change Policies, the Distribution of Income, and U.S. Living Standards
This study examines the impact on U.S. consumers of taxes designed to stabilize carbon dioxide (CO2) emissions. A major conclusion is that a tax of as much as $260 per ton could be needed to stabilize emissions at 1990 levels by 2010 if mandatory emission reductions begin in 1997 or 1998. In fact, stabilizing emissions by 2010 would slow the growth of GDP about 1 percent annually, reduce income and real wages by 5 to 10 percent per year, reduce fuel oil and coal consumption by 25 to 40 percent, cut electricity consumption by 20 to 32 percent, reduce new car and truck purchases by 3 to 5 percent, and worsen the distribution of income even if the carbon tax revenues were recycled back to consumers. Such a program would cause consumers to “feel like” they were living through the oil price shocks of the 1970s and 1980s all over again.
A number of measures to reduce greenhouse gas emissions have been proposed during the international debate over possible climate change and how governments should respond to the threat of global warming. Each program, whether based on voluntary measures, standards, regulation, or taxes, has direct implications for economic activity in general and for individual consumers in particular. The importance of understanding the impact of policies designed to mitigate climate change was amplified in July 1996 when Tim Wirth, the administration’s undersecretary for global affairs, announced at the Second Conference of the Parties of the Framework Convention on Climate Change that the United States now supports negotiations designed to set “a realistic, verifiable, and binding medium-term emissions target” for greenhouse gases. With the Third Conference of the Parties looming in 1997 and with 1990 emissions levels the most popular benchmark against which preliminary targets have been specified, it is critical that the United States address the fundamental issue of the potential cost of dramatically reducing CO2 emissions in the near term.
Emissions Reductions and Economic Growth
This study relies in large measure on simulation models exercised to explore specific policy options in the twelfth Energy Modeling Forum (EMF-12). The Forum conducted its work from 1990 through 1993, but results from models developed in subsequent years are also used. The models developed for EMF-12 facilitate the analysis of a carbon tax to implement climate change policies. While most of the researchers who participated in EMF-12 have moved on to construct full-blown, integrated assessment models of climate change, they still use the cost sides of their earlier (EMF-12) efforts to anchor more recent estimates.
The models show that a climate change mitigation policy which includes the imposition of new or increased taxes could profoundly affect economic growth due to the impact on the return to investment. For example, a carbon tax could easily interrupt streams of return to capital investment to the point of causing the premature obsolescence of some existing capital. This would cause an accelerated erosion of the existing capital stock and significantly affect the growth of potential GDP and real income.
Specifically, this study shows that the imposition of carbon taxes sufficient to stabilize emissions at 1990 levels by 2010 would reduce the growth of U.S. per capita income (a reasonable proxy for real wage growth) by about five percent per year. If emissions were reduced to 20 percent below 1990 levels, per capita income growth would fall by about ten percent per year (see Figure 1). U.S. income and real wage growth would slow as emissions were reduced because of the lost output stemming directly from higher prices for carbon-using goods-goods that must be produced using less carbon and/or by more expensive processes. Output also falls because of diminished net capital accumulation, reflecting premature obsolescence, and in some cases, altered rates of technological change.
[Figure 1: Decline in Growth Rate of U.S. Per Capita Income When CO2 Emissions are Reduced to 1990 Levels and 20 Percent Below 1990 Levels]
Note: Results for all bars except the far right bar are for various models which measure the impact of CO2 emission reductions through 2010. The baseline assumes a growth rate of 1.3 percent per year. Models are identified in “References to Models” at the end of this Special Report.
The EMF-12 simulations used to make the estimates in the study (as well as the forthcoming Scientific Assessment of the Intergovernmental Panel on Climate Change) all predate 1993 and all contemplate the costs of policies that were assumed to have been initiated in 1990. It is, however, already nearly 1997. No policy will be imposed until actual targets and timetables have been negotiated; thus no policy will be implemented before 1998. The cost of stabilizing emissions at 1990 levels by 2010 must therefore be higher than shown in Figure 1, since the transition time will be no larger than 60 percent of the twenty-year horizon considered in those models. Why? Because it always costs more to do things in a hurry. A case can be made that costs shown in Figure 1’s 20-percent reduction case represent an upper bound for the cost of achieving stabilization at 1990 levels with carbon taxes that are not imposed until 1997 or 1998. Starting late in an effort to achieve an emissions target indexed to 1990 levels could double the cost. The low-cost tax in 2010 might be $160 per ton with GDP loss estimated at 0.8 percent; the high-cost estimates suggest a tax of $260 per ton with GDP loss at 1.3 percent annually. (“Low cost” refers to models which assume that the economy can change its fuel use and product mix easily in response to a carbon tax; “high-cost” models assume that the adjustments are more costly and difficult.)
Emissions Reductions and Household Consumption
This study also provides an indication of how consumers would alter their lifestyles and rearrange their spending when carbon taxes are imposed. Household consumption patterns change significantly when energy taxes are put in place, according to estimates based on EMF-12 simulations.
For example, a $160 per ton carbon tax causes consumers to reduce fuel oil and coal consumption by 25 percent, and a $260 per ton tax causes consumers to reduce fuel oil and coal consumption by 40 percent (see Figure 2). Gasoline purchases drop by 12 to 20 percent; new truck and auto purchases decline by 3 to 5 percent. Expenditures for housing also decline.
[Figure 2: Negative Impacts by 2010 on U.S. Household Consumption Due to Stabilizing CO2 Emissions at 1990 Levels]
Note: These estimates are based on EMF-12 simulations; however, they assume that emission reductions do not begin until 1997 or 1998.
Emission Reductions and Income Distribution
Policies to curb emissions not only reduce income growth and curtail household consumption, they also worsen the distribution of income in the United States. Based on a standard measure of the degree of income inequality among a country’s population called the GINI coefficient, analysis shows that carbon taxes, even when recycled through personal income tax reductions, cause relatively large losses in the poorest quintile (lowest one-fifth of the population). These losses, added to modest losses in the middle quintiles, underwrite gains for the richest fifth of the population (see Figure 3). Revenue-neutral reductions in personal income taxes exacerbate these distributional effects, presumably because the personal income tax code is still progressive.
[Figure 3: The Impact of Energy Taxes on Income Received by U.S. Households by Quintile]
*A $260 per ton tax would be needed to stabilize emissions at 1990 levels by 2010 if emission reductions begin in 1997 or 1998.
Comparison with three successive but eventful periods during recent U.S. economic history offers some insight into the size of these income distribution changes. Figure 4 compares changes in income inequality in the recent past, as measured by the GINI coefficients, with those predicted if the United States imposes carbon taxes to stabilize emissions.
[Figure 4: The Impact of Stabilizing CO2 Emissions at 1990 Levels by 2010 on Income Inequality in the United States]
Note: The bars show changes in GINI coefficients. The higher the bar, the greater the increase in the inequality in the distribution of income. For example, the high unemployment rates and recession in the late 1970s and early 1980s caused those in the lowest quartiles to receive a smaller share of income; thus, the GINI coefficient rose (see black bar). The bars showing the changes in GINI coefficients using either lump sum revenue recycling or personal income tax reductions are based on various econometric models from EMF-12. Reflecting the fact that some of the models assume lower adjustment costs than others, two estimates are shown; one requires a tax of $160 per ton, the other $260 per ton to stabilize emissions at 1990 levels.
The black bars in Figure 4 represent: (1) 1968-73, the period prior to the oil shocks of the 1970s; (2) 1973-78, the period of the most dramatic increase in oil prices; and (3) 1978-83, the subsequent period of sharply rising oil prices and dramatic recession.
The tallest black bar reflects the growth of income inequality in the United States from1978 through 1983. This is as it should be, of course, because the distributional effects of the 1980s recession were noticed by nearly everyone and documented in the professional and popular press. Poverty rates climbed. Unemployment hit highs that had not been seen since the Great Depression. For example, in 1983 the unemployment rate reached 9.6 percent. Plants and factories closed. People moved in search of jobs and/or improved public assistance.
It is equally significant that the second largest distributional effect depicted in Figure 4 reflects the cost of a carbon tax designed to achieve stabilization in emissions–a tax-cum-recycling scheme which could have a distributional effect more than half as large as the effect of the 1980s recession. Put another way, contrasting the more “normal experience” of 1968 through 1978 with the effects of a carbon tax, it is easy to see that other policies designed to stem even the long-term trend toward less equitable distributions of income might have to work more than twice as hard just to hold the line if they were forced to work in an overall policy environment that included substantial taxes on carbon emissions.
The strong qualitative conclusion to be drawn from this study is that the economic consequences of policies designed to restrict carbon emissions severely over the relatively near term are not to be taken lightly.
Imposing taxes designed to stabilize emissions at 1990 levels could therefore “feel” like living through the oil-price shock of the early 1970s and 1980s all over again. This “feel” would be closer to actual experience if the taxes were not administered in ways that would guarantee maximal intertemporal efficiency in the economic response to their imposition. Even if the economy adapted quickly and efficiently, the average annual pace of growth in real wages is expected to fall by between five and ten percent against the baseline.
Timing is a critical issue. Enacting policies in 1997 or 1998 that would target emissions at 1990 levels or less by 2010 would impose a 12- or 13-year time constraint on adaptation that has not yet been modeled and analyzed. Rough calculations based on the estimated cost of meeting those targets in 20 years suggests that contracting the adjustment period in this way could double all costs.
Policy structure is equally critical. Markets are powerful tools and their careful creation and maintenance holds enormous potential for improved efficiency. Embedding a U.S. policy designed to achieve either a stable or reduced emissions reduction target within a consistent global commitment built around market-based mechanisms could reduce all costs by more than fifty percent.
References to Models
- CRTM: Rutherford, T. F. 1992. The Welfare Effects of Fossil Carbon Restrictions: Results from a Recursively Dynamic Trade Model. Working Paper. Paris: OECD.
- DGEM: Jorgenson, D. and P. Wilcoxen. 1991. Reducing U.S. Carbon Emissions: The Cost of Different Goals. In Moroney, J. (ed.), Energy, Growth and the Environment, Greenwich: JAI Press.
- ERM: Edmonds, J. and J. Reilly. 1985. Global Energy: Assessing the Future. New York: Oxford University Press; and
- Edmonds, J. and D. Barns. 1991. Factors Affecting the Long-Term Cost of Global Fuel CO2 Emissions Reductions. Washington, D.C.: Pacific Northwest Laboratory.
- Fossil 2: AES Corporation. July, 1990. An Overview of the Fossil 2 Model. Prepared for the U.S. Department of Energy, Office of Policy and Evaluation.
- Global 2100: Manne, A. and R. Richels. 1992. Buying Greenhouse Insurance: The Economic Costs of CO2 Emission Limits. Cambridge: MIT Press.
- Goulder: Goulder, L. 1993. Effects of Carbon Taxes in an Economy with Prior Tax Distortions. Working Paper. Stanford University.
- GREEN: Burniaux, J., J. Martin, G. Nicoletti, and J. Oliveira-Martins. June 1991. GREEN-A Multiregion Dynamic General Equilibrium Model for Quantifying the Costs of Curbing CO2 Emissions: A Technical Report. Report No. 104, Paris: OECD Department of Economics and Statistics, Resource Allocation Division.
- MWC: Mintzer, I. and V. Schaper. 1992. The Model for Warming Commitment. Working Paper. Stockholm Environment Institute.
Gary W. Yohe is professor of economics, Wesleyan University. This paper was prepared for a September 11, 1996, policy conference sponsored by the ACCF Center for Policy Research, and will be published in the ACCF’s forthcoming book, Climate Change Policy, Risk Prioritization, and U.S. Economic Growth.