The Kyoto Protocol, CAFE Standards, and Gasoline Taxes
Agras, Jean, Chapman, Duane, Contemporary Economic Policy
The Kyoto Protocol requires the U.S. to reduce the rate of emissions of greenhouse gases to 93% of their 1990 levels by the period 2008-2012. The protocol includes six greenhouse gases, but of these, carbon dioxide (C[O.sub.2]) is the most important. In 1995, energy consumption in the U.S. was 91 quadrillion Btu per year. Transportation accounted for 27% of energy consumption (EIA, 1997a, pp. 101-102) and 35% of C[O.sub.2] emissions in 1995 (Davis, 1997, p. 7-2). Of the energy used in the transportation sector, light-duty vehicles (passenger cars and light trucks) accounted for 59% of the total energy consumed (EIA, 1997a, p. 111). C[O.sub.2] reductions necessary to meet Kyoto Protocol objectives can come equally from all sectors, or in differential amounts depending on marginal costs of abatement. This paper analyzes policies that will allow light-duty vehicles to reduce C[O.sub.2] emissions to 93% of their 1990 levels by 2010. It evaluates two policies: Corporate Average Fuel Economy (CAFE) standards and gasoline taxes.
II. TRANSPORTATION POLICIES TO REDUCE EMISSIONS
A. CAFE Standard
The CAFE standard requires automotive manufacturers to meet sales-weighted minimum fuel efficiency standards for each model year on light-duty vehicles sold in the U.S. The standard for passenger cars was introduced in 1978 with a sales-weighted fuel efficiency minimum of 18 miles per gallon (mpg). In that same year, domestic manufacturers produced new cars with an average fuel efficiency of 18.7 mpg. Imported cars were a more impressive 27.3 mpg. The standard slowly increased over the next 10 years, reaching 27.5 mpg in 1990 (it had previously reached 27.5 mpg in 1985, but was reduced to 26.0 mpg the following year). The CAFE standard for passenger cars has remained at 27.5 mpg to the present (AAMA, 1997, p. 80).
The standard for light trucks was introduced in 1979 at 17.2 mpg for two-wheel drive vehicles and 15.8 mpg for four-wheel drive vehicles. Two-wheel and four-wheel drive vehicles were combined to one standard in 1992 at 20.2 mpg, and the standard has increased to 20.7 mpg today (AAMA, 1997, p. 81). When the standard for light trucks commenced in 1979, it only applied to light trucks with a gross vehicle weight (GVW) less than 6,000 lbs. In 1980, the standard expanded to include all light trucks up to 8,500 lbs GVW. Typically, light trucks refer to Class I and II trucks up to 10,000 lbs, but the heavier trucks are not regulated for fuel efficiency or emission standards. Light trucks greater than 8,500 GVW include some of the larger Dodge Rams, Ford Econolines, Ford F and C/K series pickups, GMC Sierra pickups, GMC Suburbans, and GMC Vanduras (Ward's Communications, 1996, pp. 245-252). Larger light trucks with GVW from 8,500 to 10,000 lbs remain unregulated by CAFE and emission standards.
Fines for violating CAFE standards are $55/mpg per vehicle sold. Manufacturers can bank and borrow fuel economy surpluses and deficits for up to three years. For example, Ford Motor Company's sales-weighted fuel economy in 1994 for their domestically produced light trucks was 21 mpg, 0.5 mpg above the average. In the next two years their fuel economy averages exceeded the standard by 0.2 mpg and 0.1 mpg. However, in 1997, Ford's sales weighted fuel economy average dropped to 19.9 mpg, 0.8 mpg below the standard. By carrying forward their surpluses from the previous three years, 0.5, 0.2, and 0.1, Ford is still in compliance in 1997. However, if they do not meet the standard in 1998, they will be fined $5.50 per 0.1 mpg that they fall short of the standard unless they file a carryback plan to demonstrate that they anticipate earning credits in future model years to offset current deficits (NHTSA, 1998). With a domestic light truck sales fleet of approximately 2 million vehicles, this is an $11 million fine for each 0.1 mpg they fall short of 20.7 mpg.
B. Gasoline Taxes
Consumers pay local, state, and federal gasoline taxes at the pump. Federal gasoline taxes were 4[cents] per gallon from 1960 to 1982 (in current prices). Taxes rose to 9[cents] per gallon by 1984 and stayed at that level until 1991, when they increased again to 14.1[cents] per gallon. In 1994, federal taxes increased to 18.4[cents] per gallon and remain at that level in 1999. State gasoline taxes increased more gradually, from an average of 6.1[cents] per gallon in 1960 to 40.8[cents] per gallon in 1997 (API, 1998). In 1995, state taxes ranged from a low of 7.5[cents] in Georgia to a high of 25.4[cents] in Nebraska (AAMA, 1997). In real terms, total state and federal taxes steadily decreased until 1982, then gradually increased to the present [ILLUSTRATION FOR FIGURE 1 OMITTED]. In 1997, the total tax component of gasoline price (40.8[cents]/gallon) was lower in real dollars than in 1960 (55[cents]/gallon).
III. ELASTICITY ESTIMATES FROM PREVIOUS WORK
Several studies have addressed the policy issues arising from CAFE and tax proposals. [TABULAR DATA FOR TABLE 1 OMITTED] Typically, the mathematical structure includes three main variables: fuel efficiency (MPG), vehicle miles traveled (VMT), and fuel consumption ([Q.sub.f]). These variables are linked through the identity [Q.sub.f] = VMT/MPG. While these equations take many forms, the final result is typically an estimation of various elasticities. The model used in this paper builds on work done by Haughton and Sarkar (1996), Jones (1993), Greene (1990a, 1992), Mayo and Mathis (1988), Blair et al. (1984), and Nivola and Crandall (1995). All of these papers estimate fuel efficiency and/or vehicle miles traveled and then typically calculate fuel consumption.
The basic equation for fuel efficiency includes the price of gasoline, income, and time. However, researchers typically include additional variables, especially when testing specific policy changes. Table 1 presents the variables used in fuel efficiency models in previous studies. All models include the price of gasoline. Most of the studies include a variable for income, measured as real GDP, per capita GDP, or disposable income. Haughton and Sarkar (1996), Espey (1996), Jones (1993), and Gately (1990) use a dynamic model, including a lag of MPG on the right-hand side. This allows estimates of both long- and short-run price and income elasticities for fuel efficiency. Mayo and Mathis (1988) added a variable for average highway speed to test the effect of a change in average highway speeds on demand for fuel efficiency. The basic equation for vehicle miles traveled (VMT) is a function of fuel efficiency, price of gasoline, income, and a stock variable. Alternatively, through the identity CPM = price/MPG, cost per mile (CPM) replaces the variables for price and fuel efficiency. The stock variable can be represented by the number of drivers, the number of vehicles in use, population, or the number of registered vehicles. The number of drivers is the variable most often used, followed in frequency by vehicles in use. Greene (1992) estimated equations for vehicle miles traveled testing the use of vehicle stock versus licensed drivers and found that the results did not change with the choice of stock variable.
Table 2 presents the elasticities calculated in prior work and the ranges of elasticities surveyed by Dahl (1986). However, since each model uses different parameters, the interpretation [TABULAR DATA FOR TABLE 2 OMITTED] of elasticities must be done with caution. For studies that used lagged dependent variables in MPG and VMT, both short-run and long-run elasticities are reported. For studies that did not use this dynamic framework, the direct elasticity estimate is reported. Nivola and Crandall (1995) estimated separate equations for passenger cars and light trucks and found differing elasticities. For the most part, the elasticity estimates in the individual papers concur with the survey by Dahl (1986).
IV. THE MODEL
Using meta-analysis for parameterization, this paper projects fuel efficiency, total vehicle miles traveled, fuel consumption, and C[O.sub.2] emissions in the light-duty vehicle sector. Using data from 1982-1995, equations for fuel efficiency and vehicle miles traveled are calibrated and applied to Kyoto goals for 2010. This is done separately for passenger cars and light trucks. The identity [Q.sub.f]. = VMT/MPG calculates fuel consumption, and then C[O.sub.2] emissions are estimated.(1) Once a base case with no new policies has been established (i.e., CAFE standards remain at 27.5 for passenger cars and 20.7 for light trucks, and gasoline taxes remain at their current level), then higher CAFE standards and increasing gasoline taxes are introduced individually to force fuel consumption in 2010 to equal 93% of actual 1990 levels. Finally, this model estimates what levels of these policies would be necessary when used in combination to meet Kyoto objectives.
As seen from the previous studies, fuel efficiency is typically estimated as a function of the price of gasoline, income, and time. Since the objective is to meet Kyoto Protocol objectives using gasoline taxes and CAFE standards, the fuel efficiency equation includes a variable for CAFE and adds an additional gasoline tax variable, included within the retail price variable. Equation (1) is the function used to define fuel efficiency:
(1) ln MP[G.sub.i,t] = A + [[Alpha].sub.0] ln MP[G.sub.i,t-1]
+ [[Alpha].sub.1] ln [(P + tax).sub.i,t] + [[Alpha].sub.2] ln GD[P.sub.i,t]
+ [[Alpha].sub.3] ln [CAFE.sub.i,t] + [[Alpha].sub.4] Time,
where i = passenger cars and light trucks and t = 1982-2010; MPG is the average fuel economy for all passenger cars and light trucks on the road; P + tax is the pretax price of gasoline plus additional taxes; GDP is real per capita income; CAFE is the standard in year t; and Time is a trend capturing technological change (see appendix for data sources). Equation (1) includes the lag of fuel efficiency to capture long- and short-run changes in fuel efficiency. As the CAFE standard only affects new car fuel efficiency, Equation (1) simulates the process of stock turnover, taking into account new cars on the road replacing older, less fuel efficient cars.
The equation for vehicle miles traveled takes the form
(2) ln VM[T.sub.i,t] = B + [[Beta].sub.0] ln VM[T.sub.i,t-1]
+ [[[Beta].sub.1] ln [(P + tax).sub.i,t] + [[Beta].sub.2] ln GD[P.sub.i,t]
+ [[Beta].sub.3] ln VE[H.sub.i,t] + [[Beta].sub.4] ln MP[G.sub.i,t].
VMT is total vehicle miles traveled by passenger cars and light trucks (in millions), VEH is the number of vehicles in use, and the rest of the variables are as defined above. Again, the VMT equation includes the lag of the dependent variable to capture long-run changes in vehicle miles of travel. Alternatively, the number of drivers could be used as the stock variable; however, as the results show, the vehicles in use variable more accurately reflects trends in the light truck market.
The first equation predicts MPG to the year 2010 and then this estimate is used to predict VMT to 2010. From these estimates, the identity In [Q.sub.f] = In VMT - ln MPG calculates total fuel consumption. Using the parameters for the MPG and VMT equations, the elasticities for fuel consumption can be calculated and compared to elasticities for fuel demand reported by Dahl and Sterner (1991).
Our objective is to select elasticities that reflect those reported in Table 2 and are calibrated so that the projections are similar to predictions for fuel consumption and vehicle miles traveled estimated by other sources. EIA (1997a) projects future energy use to the year 2020. For the transportation sector, they predict [TABULAR DATA FOR TABLE 3 OMITTED] a VMT annual growth rate of 1.5% and a 1.3% increase in energy use by light-duty vehicles (p. 111). Greene's (1990b) base case scenario shows both fuel consumption and vehicle miles traveled increasing at 1.5% annually from 1990 to 2015.
Since the EIA (1997a) model uses a variable for drivers in their calculations, initially the model was calibrated using number of drivers instead of vehicles in use. Equations (1) and (2) were estimated with the coefficients listed in Table 3. These values were chosen to be consistent with the elasticities reported in Tables 2 and 3, while emphasizing results of more recent analyses. For example, 0.12 is the short-run average price elasticity for fuel efficiency ([[Alpha].sub.1]), which is comparable to the work done by Haughton and Sarkar (1996). Using the coefficient for MP[G.sub.-1], the long-run price elasticity for fuel efficiency is [Mathematical Expression Omitted], which in the base case equals 0.6.
The coefficient for CAFE reflects the standard's impact on new cars on the road. Cars under 1 year old represent 12% of all cars on the road and cars under 2 years old represent 19% of all cars on the road (Davis, 1998). A short-run elasticity of 17.5% adequately reflects CAFE's impact on the stock of vehicles on the road. This also represents a long-run effect of 87.5%, which is a realistic estimate, especially as higher CAFE standards are imposed and possibly not met.
The coefficients used in the VMT equation have similar representations. We use a short-run price elasticity of -0.15, which results in a long-run elasticity of -0.32. The short-run income elasticity of 0.26 calculates into a long-run elasticity of 0.55. These numbers correspond to the elasticities reported in Table 2.
Overall, the results for the driver model are similar to other predictions. Vehicle miles traveled increases on average 1.5% annually and fuel consumption increases 1.6% annually on average. This is consistent with predictions by EIA (1997a) and Greene (1990b). Passenger cars achieve a fuel efficiency of 23 mpg by 2010 and light trucks on the road get on average 14 mpg.
However, using the number of drivers variable rather than vehicles in use may underestimate fuel consumption in the light truck sector. In the driver model, the number of drivers increases by 1% annually for both cars and light trucks, whereas cars in use have been increasing at less than 1% in the 1990s, while light trucks in use are increasing at 2-3%. This problem is reflected in the estimates for average vehicle miles traveled by individual vehicles. In 1995, both passenger cars and light trucks, on average, drove 11,000 miles per year. Under the base case scenario, the average annual miles for passenger cars increases to 14,400 miles while the light truck average decreases to 9,600 miles per year. While this mileage differential is possible, a "vehicles in use" model is utilized to better reflect anticipated growth in the light truck sector.
Using the same parameters as in the base case scenario, vehicles in use is substituted for the number of drivers in the VMT equation. Unlike gasoline prices, income, and population, predictions of vehicles in use are not widely available. Furthermore, predictions for cars and light trucks separately are not found in the literature. We developed a relationship to predict vehicles in use with equations capturing both the shift between consumption of passenger cars and light trucks, and the overall demand for vehicles as functions of the price of gasoline (including taxes) and income (see Appendix for estimation results).(2) These equations capture the shift toward light trucks with low gasoline prices (i.e., low taxes) and high incomes, and project a shift away from light trucks with higher gasoline taxes.
Using this model, the growth in total vehicle miles traveled remains the same at 1.5%, but the growth in fuel consumption is much higher (2.0% instead of 1.6%). This is consistent with estimates by the EPA when taking into consideration the growth in the light truck market. As well, the projected annual average of vehicle miles traveled by cars and trucks is more representative of current growth rates: 13,000 and 12,000 miles, respectively, traveled annually by 2010. The increase in fuel consumption is more rapid with the vehicle model, exceeding the driver model by 9% in 2010. Fuel consumption predictions with the vehicle model also better capture recent increased demand for light trucks.
V. INTRODUCTION OF POLICIES TO MEET OBJECTIVE
Now that a base case has been established, policies are introduced to force fuel consumption and C[O.sub.2] emissions to reach 93% of their 1990 levels by 2010. This paper analyzes the effect of the use of CAFE standards and gasoline taxes together. While most of the previous research has focused on determining which of these policies is superior, we make a case for their combined use. The past few years have seen steadily decreasing gasoline prices, accompanied by consumer demand for larger, less fuel-efficient vehicles. Forcing producers to make more fuel-efficient vehicles while customers want more powerful cars would be ineffective. Alternatively, raising gasoline prices alone would lead to a longer adjustment to the production of cars with higher fuel economies. By introducing these policies together, consumers will be faced with higher utilization costs and producers would be investing in research and development to lower fuel economies.
Stricter CAFE standards and increased gasoline taxes are initially introduced individually to determine what levels of each are necessary to reach Kyoto objectives. All policies are introduced in 1999 and a certain level is added each year to 2010. For example, using only the CAFE standard to reach the objective, 3.2 mpg needs to be added each year from 1999 to 2010 to reach the objective. This increases the new [TABULAR DATA FOR TABLE 4 OMITTED] passenger car CAFE standard from 27.5 mpg in 1998 to 66 mpg in 2010 and the light truck standard increases to 59 mpg by 2010.
Using only a gasoline tax, 10.5[cents]/gallon needs to be added each year to adequately decrease fuel consumption. This means that the price of gasoline would be $2.42/gallon in 2010 (in 1996$). Table 4 presents the annual increases in the CAFE standard and gasoline tax necessary to meet the objective and the resulting levels in 2010.
Once the increases are estimated individually, both CAFE and gasoline taxes are allowed to adjust, and we estimate the necessary levels using the policies jointly. When the policies are used jointly, less than 50% of each is necessary to achieve the goal. Only 41% of initially established CAFE and gasoline taxes would be necessary to meet the objective. Figure 2 presents the locus of points of varying percentages that reduce fuel consumption and C[O.sub.2] emissions to 93% of 1990 levels. From this isoquant, levels of each policy can be chosen that will fulfill Kyoto obligations. If, for example, policy makers favored using both policies equally, only 41% of the initial levels of each is needed (this is represented by the solid lines in Figure 2). However, if policy makers feel that their constituents are more favorable toward the CAFE standard, higher CAFE standard could be used with a slightly lower gasoline tax (a 56% CAFE and a 28% tax represent a 2-to-1 preference for CAFE standards - see dashed lines in Figure 2).
The shape of this isoquant depends mainly on the price elasticities in each equation ([[Alpha].sub.1] and [[Beta].sub.1]), the CAFE elasticity ([[Alpha].sub.3]), and the fuel efficiency elasticity of vehicle miles traveled ([[Beta].sub.4]). The coefficient on CAFE will always be negative as long as [[Beta].sub.4] [less than] 1, and the degree of curvature of the isoquant will vary depending on the magnitudes of the other elasticities, but will remain convex.
To test the model's robustness, the upper and lower ranges of elasticities listed in Table 3 were substituted for each of the parameter estimates in the model. For comparison, the CAFE and gasoline tax model was then estimated to project the CAFE standard and gasoline tax increases necessary to reach the objective. Figure 3 illustrates these results. With the exception of using a CAFE elasticity of 0.05 in the MPG equation,(3) none of the elasticity extremes change the results significantly. The magnitude of the CAFE and gasoline tax increases does change, but the result that using the policies together is more effective than using either policy individually still holds.
VI. C[O.sub.2] EMISSIONS
As mentioned above, the estimate for C[O.sub.2] emissions (in metric tons) can be directly derived from estimates of fuel consumption through the identity C[O.sub.2] = 2424.88[Q.sub.f], where 2424.88 metric tons carbon per million gallons of gasoline are emitted. Figure 4 presents C[O.sub.2] emissions from the base case results for the vehicle model. This figure separates emissions from light trucks and passenger cars. In the driver model, C[O.sub.2] emissions from light trucks never reach the level of passenger cars. However, this is inconsistent with EPA predictions. The EPA has estimated that C[O.sub.2] emissions from light trucks will surpass that of passenger cars before the year 2000 (Bradsher, 1997). The vehicle model is a better predictor of this result. Figure 4 shows that around 2006, C[O.sub.2] emissions from light trucks will be equal to passenger car emissions at 180 million metric tons, and they exceed passenger car emissions by 13% in 2010. From the vehicle model, emissions from light trucks and passenger cars are predicted to be 211 and 187 million metric tons, respectively by 2010. The EPA predicts levels of 250 and 150 million metric tons in 2010 for light trucks and passenger cars, respectively (Bradsher, 1997).
VII. DISCUSSION OF THE MODEL
A CAFE standard that increased by 3 mpg per year is both technologically and politically infeasible. Long-run estimates of light vehicle fuel economy find passenger cars able to reach 34-37 mpg and light trucks to reach 26-28 mpg by 2006 (NRC, 1992). This is an increase of 1 mpg per year from 1999 to 2006. From this, one can see that using the combination of taxes and CAFE standards proposed in this model is technologically feasible and could result in significant decreases in gasoline consumption and C[O.sub.2] emissions.
However, with steadily increasing CAFE standards, producers may choose to pay the fines and continue selling vehicles with lesser fuel efficiencies. This would mean that fuel consumption and emissions exceed the model. On the other hand, if manufacturers strive to achieve these higher fuel economy standards, prices of vehicles could rise. With higher vehicle prices, consumers would purchase fewer vehicles, implying that fuel consumption and emissions would be less than the model projects. This model could be strengthened by deriving demand for vehicles as a function of the distribution of vehicle prices, which is itself a function of the CAFE standard. In general, raising CAFE (a) may lower average price because the average vehicle is lighter in weight or (b) may increase the average price if more costly materials are used to attain lighter vehicles. Almost certainly, CAFE leads producers to lower small car prices and increase large car prices.
The final result of this model is that there is a range of policy options available that would allow the U.S. to reduce its demand for gasoline. We have discussed the choice in a political economy framework. While this may be appropriate for implementation of policies, it is certainly not the only choice variable. Ideally, the optimal combination of policies should be the one that maximizes social welfare. This means that the policies should be equitable (i.e., the benefits and costs are shared by all) and efficient. Several papers have addressed the issue of the costs and benefits of increased CAFE standards and gasoline taxes (see Crandall, 1992; Crandall et al., 1986).
Table 5 presents some of the economic and environmental consequences of each policy. Both CAFE and gasoline taxes reduce consumer surplus through increased prices and reduce fuel consumption and emissions. However, CAFE may have additional costs in the form of market distortions. With an increasing demand for larger vehicles (attributable to low gasoline prices and aggressive advertising by U.S. car producers), manufacturers may find they have to sell an increased number of smaller, more fuel-efficient vehicles to counterbalance their increased sales of larger, less fuel-efficient vehicles. This could create a distribution of vehicles on the road that leads to more traffic fatalities. Ultimately, the optimal policy choice should be one that reflects the costs and benefits listed in Table 5 and chooses the best combination of CAFE and gasoline taxes to promote the achievement of goals derived from the Kyoto Protocol.
This paper has shown one way that the U.S. could achieve Kyoto Protocol objectives in the light-vehicle sector. An implementation of increased CAFE standards and higher gasoline taxes would significantly reduce C[O.sub.2] emissions in the light-vehicle sector. The U.S. initiated the CAFE standards twenty years ago but has maintained the same standard for passenger cars for the last ten years and has only increased the standard for trucks by 0.2 mpg in that same time. The policy is in place, yet it requires some increase to effectively reduce fuel consumption.
Some may argue that since the Kyoto Protocol applies to overall C[O.sub.2] and greenhouse gas emissions, reductions can more efficiently be made in other sectors. However, ultimately gasoline consumption and emission from the transportation sector need to be addressed. Alternatively, these savings could come in the form of reformulated gasoline and other fuel equivalents. The near-term technologies being studied result in decreases in C[O.sub.2] emissions from conventional gasoline vehicles fueled with conventional gasoline on the order of 8-50% (Davis, 1998).
Additionally, any reductions in fuel consumption to curb C[O.sub.2] emissions will have co-benefits for local air pollution problems. Since 1970, the U.S. has been very successful in reducing vehicle emissions of carbon monoxide, nitrogen oxides, hydrocarbons, and lead. However, in those same years, the number of miles driven has more than doubled, offsetting some of the overall emissions reductions (EPA, 1994). While technologies are still being developed to offset more of these pollutants, reductions in fuel consumption will have an immediate effect in improving local air quality.
Using CAFE standards exclusively will not give consumers sufficient economic incentive to use less fuel. A combination of increased CAFE standards and higher gasoline taxes will help the U.S. meet Kyoto Protocol objectives and improve air quality everywhere.
TABLE 5 Costs and Benefits of CAFE Standards and Gasoline Taxes CAFE Gas Tax Costs Increased vehicle prices Increased gasoline prices More older cars on the road Politically unpopular Market distortions (increased Possibly taxing poor more than number of big vehicles rich (regressive) supplemented by an increased number of smaller cars) Increased vehicle miles traveled Benefits Decreased fuel consumption Decreased fuel consumption Decreased emissions Decreased emissions Lower fuel economy car prices Increased government revenue could subsidize higher fuel Increased demand for fuel economy car prices (progressive) efficient cars
Vehicle Miles of Travel (VMT): American Automobile Manufacturers Association (AAMA), Motor Vehicle Facts & Figures, various years.
Fuel Consumption ([Q.sub.f]): American Automobile Manufacturers Association (AAMA), Motor Vehicle Facts & Figures, various years.
Fuel Efficiency (MPG): Calculated from VMT and [Q.sub.f].
Number of Drivers (DRIV): American Automobile Manufacturers Association (AAMA), Motor Vehicle Facts & Figures, various years. The number of drivers was increased by 1% per annum 2010.
Price of Gasoline (price): Energy Information Administration (EIA), Monthly Energy Review, May 1998. The price data for 1982-1997 came from the series for All Types from Table 9.4 (Motor Gasoline Retail Prices, U.S. City Average, deflated to 1996 [cents]/gallon by CPI-U; (EIA, 1998, p. 114). These prices include federal and state taxes. The data for 1998-2010 are from Table 12 (Petroleum Product Prices). They are the sales-weighted average price for all grades, including predicted federal and state taxes, but excluding county and local taxes.
Gasoline Taxes (tax): American Petroleum Institute (API), How Much We Pay for Gasoline, 1997 Annual Review, Policy Analysis and Strategic Planning Department, April 1998.
Corporate Average Fuel Economy Standards (CAFE): American Automobile Manufacturers Association (AAMA), Motor Vehicle Facts & Figures, various years.
Per Capita Income (GDP): Economic Report of the President, 1997. GDP is then forecast to increase by 1.3% from 1996-2000 and by 1.5% from 2001-2010.
Vehicles in Use (VEH): AAMA, various years. The 1982-1995 data for vehicles in use comes from the AAMA's Motor Vehicle Facts & Figures. This data was then forecast using equations to capture the shift between passenger cars and light trucks and total demand for vehicles with changing gasoline prices and incomes. We estimated the following equations:
(1A) [Mathematical Expression Omitted]
(2A) [Mathematical Expression Omitted]
where i = passenger cars and light trucks, t = 1978-1995, SHARE is the percentage of passenger cars in use, P + tax is the average retail price of gasoline (all types) plus federal and state gasoline taxes, GDP is per capita income, and VEH is the number of vehicles in use. Equation (1A) uses a logit transformation of SHARE and both equations are estimate by ordinary least squares. The estimated coefficients for equations (1A) and (2A) are listed in Appendix Table 1.
This is a revision of a paper presented at the Western Economic Association International 73rd annual conference, Lake Tahoe, June 29, 1998, in a session organized by Jane Hall, California State University, Fullerton. The authors are grateful for the comments and suggestions made by Tim Mount, Neha Khanna, and anonymous referees.
1. One million gallons of gasoline emits 2424.88 metric tons of C[O.sub.2] (EIA, 1997b, p. 100).
2. We thank Timothy Mount for his suggestion of this estimating method.
3. With a CAFE elasticity of 0.05, the CAFE standard would have to increase 12.7 mpg per year and the gasoline tax would have to be increased 3.2[cents]/gallon per year.
AAMA, (American Automobile Manufacturers Association), Motor Vehicle Facts & Figures, Washington, D.C., various years.
API (American Petroleum Institute), How Much We Pay for Gasoline, the Cost of Motor Gasoline to Consumers: 1997 Annual Review, Policy Analysis and Strategic Planning Department, Washington, D.C., April 1998.
Blair, R. D., D. L. Kaserman, and R. C. Tepel, "The Impact of Improved Mileage on Gasoline Consumption," Economic Inquiry, 22, 1984, 209-217. Bradsher, K., "Trucks, Darlings of Drivers, Are Favored by the Law, Too," New York Times, November 30, 1997, 1, 42.
Crandall, R. W., "Policy Watch: Corporate Average Fuel Economy Standards," Journal of Economic Perspectives, 6:2, 1992, 171-180.
Crandall, R. W., H. K. Gruenspecht, T. E. Keeler, and L. B. Lave, Regulating the Automobile, Brookings Institute, Washington, D.C., 1986.
Dahl, C., "Gasoline Demand Survey," The Energy Journal, 7:1, 1986, 67-82.
Dahl, C., and T. Sterner, "Analysing Gasoline Demand Elasticities: A Survey," Energy Economics, July 1991, 203-210.
Davis, S.C., Transportation Energy Databook, 17th ed., Center for Transportation Analysis, Oak Ridge National Laboratory, Oak Ridge, Tenn., August 1997.
-----, Transportation Energy Databook, 18th ed., Center for Transportation Analysis, Oak Ridge National Laboratory, Oak Ridge, Tenn., August 1998.
EIA (Energy Information Administration), Annual Energy Outlook 1998, with Projections to 2020, Office of Integrated Analysis and Forecasting, U.S. Department of Energy, Washington, D.C., December 1997a, http://www.eia.doe.gov/forecasting_index.html.
-----, Emissions of Greenhouse Gases in the United States, 1996, U.S. Department of Energy, Washington, D.C., October 1997b.
-----, Monthly Energy Review, U.S. Department of Energy, Washington, D.C., May 1998.
EPA (Environmental Protection Agency), "Automobile Emissions: An Overview," EPA 400-F-92-007, Fact Sheet OMS-5, August 1994, Washington, D.C., http://www.epa.gov/OMSWWW/.
Espey, M., "Watching the Fuel Gauge: An International Model of Automobile Fuel Efficiency," Energy Economics, 18, 1996, 93-106.
Gately, D., "The U.S. Demand for Highway Travel and Motor Fuel," The Energy Journal, 11:3, 1990, 59-73.
Greene, D. L., "CAFE or PRICE? An Analysis of the Effects of Federal Fuel Economy Regulations and Gasoline Price on New Car MPG, 1978-89," The Energy Journal, 11:3, 1990a, 37-57.
-----, "Short Options for Controlling C[O.sub.2] Emissions of Light Duty Vehicles," SAE Technical Paper Series 901111, Oak Ridge National Laboratory, Oak Ridge, Tenn., May 1990b.
-----, "Vehicle Use and Fuel Economy: How Big Is the 'Rebound' Effect?" The Energy Journal, 13:1, 1992, 117-143.
Haughton, J., and S. Sarkar, "Gasoline Tax as a Corrective Tax: Estimates for the United States, 1970-1991," The Energy Journal, 17:2, 1996, 103-126. Jones, C., "Another Look at U.S. Passenger Vehicle Use and the 'Rebound' Effect from Improved Fuel Efficiency," The Energy Journal, 14:4, 1993, 99-110.
Mayo, J. W., and J. E. Mathis, "The Effectiveness of Mandatory Fuel Efficiency Standards in Reducing the Demand for Gasoline," Applied Economics, 20, 1988, 211-219.
NHTSA (National Highway Traffic Safety Administration), Twenty-second Annual Report to Congress, Calendar Year 1997, Department of Transportation, Washington, D.C., 1998, http://www.nhtsa.dot.gov.
NRC (National Research Council), Automotive Fuel Economy, How Far Should We Go? National Academy Press, Washington, D.C., 1992.
Nivola, P.S., and R. W. Crandall, The Extra Mile.' Rethinking Energy Policy for Automotive Transportation, Brookings Institute, Washington, D.C., 1995.
Ward's Communications, Ward's Automotive Yearbook, D. Ferris (ed.), Detroit, Mich., 1996.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: The Kyoto Protocol, CAFE Standards, and Gasoline Taxes. Contributors: Agras, Jean - Author, Chapman, Duane - Author. Journal title: Contemporary Economic Policy. Volume: 17. Issue: 3 Publication date: July 1999. Page number: 296. © 2003 Western Economic Association International. COPYRIGHT 1999 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.