Transportation Energy Management
The transportation sector accounts for 28 percent of the fossil fuel and 63 percent of the petroleum consumed in the United States, resulting in the emission of a number of greenhouse gases, including 30 percent of U.S. CO2 emissions (International Energy Agency, 1984; European Economic Community, 1988; Lyman, 1990). Petroleum-fueled personal passenger vehicles (both automobiles and light trucks) account for 58 percent of all transportation energy use (Ross, 1989) and thus the majority of transportation-related CO2 emissions.
In reviewing methods of reducing greenhouse gas emissions from the transportation sector, the panel focused on three areas: vehicle efficiency, alternative transportation fuels, and transportation system management.
In the short run, transportation energy consumption can change rapidly as consumers adjust their demands concerning when and how to travel. On a slightly longer time scale, higher vehicle and fuel prices, along with shifts in vehicle and transportation demand, will lead to changes in the types of vehicles in use. On a significantly longer time scale, investments can be made in alternative transportation fuels, construction of new mass transit facilities and high-occupancy-vehicle (HOV) lanes, land use planning and jobs/housing balance, research and development, and tooling for technological improvements.
The oil embargoes of the 1970s heightened concern for the efficient use of energy in the transportation sector. In the 1980s, however, declining oil costs led to reduced vehicle operating costs, and the concern for energy efficiency diminished. More recently, oil prices have begun to rise again,
and that, combined with a growing concern about greenhouse warming and the role of CO2 as a greenhouse gas, has once again focused attention on the efficient use of transportation fuel. Vehicle efficiency involves technological improvements in fuel economyimproving the miles per gallon of vehicles. The more efficient a vehicle is, the less fuel it burns to travel a given distance. The less fuel it burns, the lower are the amounts of CO2 emitted. These methods of emission reduction and their cost-effectiveness are evaluated in the following sections for light-duty vehicles, heavy-duty trucks, and domestic air carriers. The emphasis is on light-duty vehicles because they represent the largest and most thoroughly studied sector. Table 23.1 shows the amount of fuel used by each type of vehicle for different modes of operation. The information presented here indicates that light-duty vehicles consume the largest quantity of transportation fuel, with heavy-duty trucks second and aircraft third.
The recent trend in fuel economy from 1975 to 1989 for the new U.S. passenger car fleet is presented in Figure 23.1a (Amann, 1989). Figure 23.1b shows the fuel economy index (FEI) for the period from 1930 to 1990. The FEI is an index of powertrain efficiency including weight and performance. Studies by Leone and Parkinson (1990) and by Greene (1989) indicate that the trend in the period from 1975 to 1982 was a response to increased fuel prices and fuel economy regulations. As discussed below in the ''Barriers to Implementation" section, there is some disagreement on the relative impact of fuel prices and regulations on the supply of fuel-efficient vehicles. The vehicles manufactured during that period were, on the average, 450 kg (1000 pounds) lighter within each market segment, were degraded in performance and other attributes, and incorporated various fuel-efficient technologies. Figure 23.2 indicates how consumer preferences for vehicles changed from 1972 to 1986. As shown, some consumers accepted the smaller vehicles offered to improve energy efficiency, while others resisted the change in performance and either did not buy cars or shifted to light-duty trucks and vans.
Because market conditions and fuel prices cause consumer preferences for fuel-efficient vehicles to change over time, one should distinguish between the trends in overall vehicle fuel economy and powertrain efficiency. Therefore it is important to look not only at miles per gallon but also at the FEI. The FEI is used to control for other vehicle changes, as shown in Figure 23.1b for the period from 1930 to 1990. This parameter, used to judge passenger cars for many decades, provides a better indicator of powertrain efficiency than does fuel economy alone by controlling for both weight and performance.
TABLE 23.1 Transportation Energy Use by Mode, 1987
(continued on page 289)
(Table 23.1 continued from page 288)
An envelope of currently "acceptable trades" in the United States is shown in Figure 23.3 in a schematic calculated by Amann (1989). The envelope represents measured performance and fuel economy data for 50 different models of cars produced in 1985 and 1986, all equipped with automatic transmissions. The y-axis in Figure 23.3 represents FEI, the product of vehicle mass and fuel economy. The weight class averages are from 1979, 1984, and 1989. The weight is defined as the test-weight class and the fuel economy is measured by the Federal Test Procedure, both as
defined by the Environmental Protection Agency (EPA). The performance index on the x-axis is the time (in seconds) required to accelerate from 0 to 60 mph. Though subject to abuse, performance is an attribute desired for freeway merging, highway passing, hill climbing, and towing. The negative slope of this parallelogram quantifies the inherent trade-off between fuel economy and performance for a car of given mass and level of powertrain technology.
Emission Control Methods
The efficiency of transportation vehicles can be increased in a number of ways, including improving the engine and transmission efficiency, reducing
weight, improving vehicle aerodynamics, and reducing the tire-rolling resistance.Each of these methods and their cost are discussed below.
Technological Improvements in Light-Duty and Heavy-Duty Vehicles
Efforts to improve conventional gasoline engines focus on obtaining the optimal engine operation possible through methods such as improving part-load operations (power required at less than peak efficiency) and combustion characteristics, reducing engine warm-up time, increasing compression ratios, reducing frictional losses, and extending the operating regime of engines (Bleviss, 1988).
Another method of improving vehicle efficiency is to increase transmission efficiency. The focus here is on reducing the losses associated with either automatic or manual transmissions through design changes. Automatic transmissions provide shifting at the optimal moment (unlike manual transmissions, which depend on the driver) so that the vehicle can remain at full load, but they have pumping and frictional losses. Manual transmissions do not have these losses. However, drivers do not generally operate manual transmissions at maximum efficiency by shifting at the optimal moment. A number of methods for increasing transmission efficiency have
been suggested, including designs that include the continuously variable transmission (CVT).
Reducing a vehicle's weight can yield major improvements in fuel economy. By reducing the weight of a vehicle 10 percent, city fuel economy can be improved 7 to 8 percent, and highway fuel economy 4 to 5 percent. The weight of the vehicle is reduced by substituting advanced materials for the steel and cast iron that were used in the past. New materials that can reduce vehicle weight include high-strength low-alloy (HSLA) steel, aluminum, magnesium, and plastics (Bleviss, 1988).
Improving the aerodynamics of a vehicle involves changing the exterior
structure through changes in areas such as wheel openings, windshield wiper location, glass location, and underbody. A 10 percent decrease in aerodynamic drag will result in a 5 to 6 percent (highway) and 2 to 3 percent (city) reduction in fuel consumption (Bleviss, 1988).
Another method of improving fuel economy is to reduce tire-rolling resistance. In the 1970s the introduction of radial and P-metric tires improved average fuel economy by 5 to 9 percent relative to the previously used bias-ply tires. Currently, a 10 percent improvement in tire-rolling resistance improves vehicle fuel economy by 3 to 4 percent (Bleviss, 1988).
Additional fuel economy improvements can be made by improving the efficiency of vehicle accessories such as the air conditioner, alternator, power steering, water pump and fan, oil pump, and lights (Bleviss, 1988).
Approximately 20 percent of the fuel expended on the U.S. highway system is utilized by heavy trucks and other heavy-duty vehicles. The opportunities to conserve fuel and minimize emissions parallel those for light-duty vehicles and light-duty trucks.
Technological Improvements in Aircraft
The third most significant component of transportation energy use is that of domestic airlines. The most advanced new jet aircraft are far more efficient than older aircraft still in service. Carlsmith et al. (1990) estimate that on a 1000-mile trip, aircraft produced in the 1960s are capable of between 40 and 50 seat-mpg, while the new Boeing 757 and 767 now in service have a fuel efficiency of 70 seat-mpg. Improvements now being introduced arise from a combination of higher-bypass-ratio engines, increased compressor and turbine efficiencies, and more energy-efficient flight planning and operations. Like highway vehicles, aircraft can also benefit from weight reduction and better aerodynamics. Efficiencies approaching 130 to 150 seat-mpg are estimated for planned vehicles utilizing improved fanjets along with other new technologies. A 20 percent gain for some existing vehicles in the fleet can be achieved via improved fanjet technology alone.
Cost Estimates for Vehicles and Air Carriers
Costs for achieving higher fuel economy for light- and heavy-duty vehicles have been derived from data produced in three studies. Data from the Mellon Institute study (Shackson and Leach, 1980) and from two Department of Energy (DOE) studies (Ledbetter and Ross, 1989; Difiglio et al., 1990) were assembled (see Appendix E) to calculate the three different curves in Figures 23.4 (6 percent discount rate) and 23.5 (30 percent discount rate). These curves show the cost in dollars per gallon relative to the improvement in on-road fuel economy. Values from the average new car
fleet efficiencies along with the average fuel prices for Japan, Sweden, United Kingdom, United States, and West Germany have been plotted as squares on the same figures. Cost-effectiveness values are highly uncertain and vary significantly from technology to technology. Although some cost-effective technologies create a net savings to the consumer, less efficient technologies generate costs exceeding $10/t CO2 equivalent (at 6 percent discount rate) for those not involving life-style adjustments and more than $700/t CO2 (at 6 percent discount rate) for those that involve changes in life-style.1 The costs are much higher when a discount rate of 30 percent is usedthe rate of return consumers generally desire for investing in vehicle efficiency. Overall cost estimates depend on a variety of assumptions involving the costs, effectiveness, and interactions of technology combinations as well as consumer preferences and market behavior. Further complications arise from variations in the cost structures of different manufacturers, variations across market segments, cost differences due to exchange rates, and international differences in driving behavior and vehicle use patterns.
The discontinuity in the slope of the Difiglio curve at approximately 31 mpg (on-road fuel economy) is the arbitrary point at which sales shifts in the vehicle mix are required to gain higher average fuel economy levels. Difiglio has labeled this point the "maximum technology" point for the year 2000.2 In the region beyond this arbitrarily defined point, downsizing of the fleet or development of expensive technology will be required to achieve future efficiency improvements. Efficiency gains beyond this point are achieved by downsizing (which reduces passenger and carrying capacity) and by compromising other attributes such as performance, comfort, and safety.
The average cost-effectiveness values, as distinct from marginal cost-effectiveness values, which increase substantially as one moves past 25 mpg, are summarized in Table 23.2 as a function of discount rate for all three cost curves (see Appendix E). This table shows the net implementation cost and potential emission reduction for light-duty vehicles, heavy-duty trucks, and aircraft. Both the more accurate full-cycle emission accounting and an accounting including only the emissions generated during consumption are shown. Cost and emission reductions for light-duty vehicles are shown for the three analyses available for that sector. The net implementation costwhich subtracts the fuel saved from the implementation costis presented for four different interest rates (3, 6, 10, and 30 percent). Beyond the "maximum technology" point, the amount of CO2 reduction and the corresponding cost-effectiveness values are summarized in Table 23.3, which repeats the analysis in Table 23.2 for improvements that involve life-style adjustmentthe "maximum technology" point on the Difiglio supply curve shown in Figures 23.4 and 23.5. The net implementation cost and potential emission reduction are shown in Table 23.3 for light-duty vehicles, both with the more accurate full-cycle emission accounting
TABLE 23.2 Implementation Cost of Vehicle Efficiency Improvements with No Change in Fleet Mix
TABLE 23.3 Implementation Cost of Vehicle Efficiency Improvements with Change in Fleet Mix
(i.e., the emissions generated from production to consumption of the fuel), and the consumption emission accounting. The full-cycle costs are the most appropriate for determination of the cost of vehicle efficiency. Even so, the costs and emission reductions in the consumption accounting section have to be used for comparison with the other options in this reportwhich, unfortunately, are not full cyclebecause information on the full-cycle emissions is not currently available. As illustrated here, the difference between full-cycle emission accounting and consumption emission accounting is quite large and could make a difference in comparing different emission reduction options. Note also that the CO2 emissions determined in these analyses are multiplied by 1.55 to obtain CO2 equivalents and thus account for non-CO2 greenhouse gas emissions equivalence for both the cost-effectiveness and the emission reduction numbers. Because the technologies save fuel, the cost-effectiveness values were credited with $1.00/gallon for the social discount rates of 3, 6, and 10 percent and $1.25/gallon for the 30 percent discount rate.
The cost estimate for heavy-duty vehicle improvements is more difficult than that for light vehicles because of the meager data base. However, on the basis of this limited information, it is possible to determine some cost data, as summarized in Table 23.2.
Cost calculations for domestic air carriers are also shown in Table 23.2. This estimate is based on retrofitting half of the 4000 transportation vehicles in the fleet with the improvement in fanjet technology mentioned above.
Barriers to Implementation
Until now, the industry has offered each market segment an amount of fuel-saving technology commensurate with the perceived value of fuel savings to the consumer. Of course, as the price of fuel varies, the offerings to the consumer must vary. In principle, the consumer has no preference in a choice between buying a technology and buying fuel when they are perceived as equal. The equality is dependent on the baseline fuel economy, the gain from the technology, the technology price, the fuel price, and discount rates, as well as expectations involving amount of driving, vehicle life, vehicle retention, and future fuel prices. Because of variations in these parameters, many analyses using the same data can differ dramatically. Therefore there is disagreement among social scientists as to the relative impact of fuel prices and mandated fuel economy levels on the supply of fuel-efficient vehicles. For example, comparisons of the results of econometric analyses by Godek (1990), Greene (1989), and Leone and Parkinson (1990) reveal differences on the relative importance of command-and-control mechanisms.
If it is assumed that market mechanisms are effective incentives for introducing new technologies when a sufficient number of competitors are present, a look at the real-world data in Figures 23.4 and 23.5 can guide the discussion. As fuel prices increase, many technologies become cost-effective, and manufacturers supplying these technologies gain a competitive advantage.
As indicated in Figures 23.4 and 23.5, fuel prices in many nations are sufficiently high to justify the introduction of more-efficient technologies. Inappropriate assumptions, however, involving driving habits, vehicle use patterns, scrappage rates, price expectations, industry cost structure, or cost of capital limit the applicability of this information to the United States. It is also possible that inappropriate exchange rates distort these international comparisons.
Although, as shown in Figure 23.4, a number of efficiency measures are cost-effective at a social discount rate of 6 percent, studies indicate that the average consumer often evaluates efficiency investments (as discussed in Chapter 21 and shown in Figure 21.11) at a discount rate of 30 percent. The perspective in Figure 23.5 is that of a consumer expecting a 10-year benefit stream from a 30 percent discount rate on future benefits. If consumers in these five nations had no preference in a choice between purchasing fuel economy technology and purchasing gasoline, they would choose a set of technologies on one of the three curves. If, on the other hand, consumers valued other attributes in their vehicles that are sacrified by fuel economy technologies, they would choose a level of fuel economy lower (i.e., to the left) of the curves in Figure 23.5.
There are a number of explanations of why the actual data points generally show consumers buying less fuel economy than they might in a perfect market. First, the cost curves assume perfect markets, but the real-world data points may reflect a variety of market imperfections. Second, because Figure 23.5 is based on a consumer's personal discount rate of 30 percent, the Mellon curve may represent the real supply curve; in this case, the figures demonstrate that consumers are using a high discount rate. Third, consumers may value attributes in their vehicles that are degraded as fuel economy increases; in this case, the reason for the disparity is that the value of these lost attributes is not captured by the supply curves. Finally, since government regulation that alters free-market consumer choice can increase consumer costs, the cost curves may be too low because they do not reflect the costs of regulatory measures used by different countries to promote increased fuel economy.
Difiglio et al. (1990) calculate that fuel economy achieved by significant mix shifting has a potential consumer loss equivalent to $24 billion per year$3500 per car in their scenario, wherein only 8 of the original 40 car lines remain. The simulation model of Kleit (1988) indicates that mix
shifting to achieve a 2.5-mpg increase beyond the point at which technologies are cost-effective will impose an average consumer cost of over $800 per car sold. These costs are equivalent to $8.4 billion per year in the United States or $4.61/gallon above the cost of gasoline (Kleit, 1988).
Manufacturers produce vehicles with the attributes that they believe consumers will purchase. Based on past experience and current market conditions, producers apparently do not believe there is a market for a high-attribute, high-fuel-economy vehicle. Although some consumers have expressed a desire for high-attribute vehicles, with gas mileage of 50+ mpg (as is available in small, low-attribute vehicles), this does not currently constitute a market in the eyes of both large- and small-volume producers. Therefore a consumer cannot purchase a high-attribute vehicle with gas mileage at the 50+ mpg range. Until these producers believe that a market for such a vehicle is available, they are unlikely to produce oneespecially as it would involve a potentially risky investment in light of their available information.
The information provided here has been designed to emphasize the importance of both fuel economy and other consumer-driven attributes at reasonable cost in all sectors of the automotive market. As Atkinson and Halvorsen (1984, 1990) have made clear, consumers recognize the tradeoffs between these and other attributes in exercising their choices in the market. It is also clear that some levels of fuel economy require technologies that are not cost-effective when compared to other societal options (Bleviss, 1988). Furthermore, with regard to specific technologies, there are significant differences between the efficiency gains estimated by DOE and those estimated by engineers in the industry.
To summarize, there are a number of barriers to improving vehicle efficiency:
• Uncertainty in technology availability: Whether a technology will ever be available that can produce desired efficiency gains at a cost lower than the lifetime fuel savings from the technology is uncertain. There is some chance that this barrier can be overcome by increased spending on research and development, particularly the development and testing of prototypes.
• Consumer resistance: Some consumers resist buying efficient vehicles at the expense of other desired attributes (e.g., perceived safety and performance).
• Consumer perception: Some consumers often do not perceive the financial benefit from buying a more fuel-efficient vehicle because they look only at the capital cost of the vehicle they are paying at that time instead of the financial savings due to lower fuel consumption over the life of the vehicle.
These barriers imply that (at current fuel prices) changes in current consumer perceptions and behavior would be needed for significant reductions in transportation fuel consumptionand the resulting greenhouse gas emissions. These consumer perceptions and behavior affect the actions taken by producers of vehicles.
One problem with achieving vehicle efficiency improvements is that it will take years for their full effect to be felt. Long lead times are required to meet and reach the emission reductions necessary. Cars must be designed, manufacturing equipment must be retooled, emission standards must be met, and the cars must eventually be sold as consumers decide to purchase new vehicles (U.S. Department of Energy, 1989). A new vehicle takes 4 to 5 years in the United States to go from prototype to new product (Altshuler et al., 1984), and commercial aircraft take even longer (U.S. Department of Energy, 1989). Passenger car turnover rates are also slow, on the order of 7 to 8 years (Holcomb et al., 1987), and rates for other types of vehicles such as heavy trucks are even slower (12 to 15 years) (U.S. Department of Energy, 1989). If the selection of large vehicles is constrained, consumers may choose to retain less-efficient vehicles, and scrappage rates may be reduced. The following policy options could be used to encourage vehicle efficiency improvements.
Corporate Average Fuel Economy (CAFE) Standards
The most common method used to increase fuel economy in the past has been the Corporate Average Fuel Economy (CAFE) standards. The 1975 Energy Policy and Conservation Act required manufacturers to increase the sales-weighted fuel economy of their products from 14.2 mpg in 1973 to 18 mpg in 1978; this mechanism has continued to be used and reached 27.5 mpg by 1985 (U.S. Department of Energy, 1989). The purpose of the CAFE standards is to stimulate technological improvement and reduce fuel consumption. In particular, CAFE standards may have played a role during the middle to late 1980s, when fuel economy increased despite declining oil and gasoline prices.
Because CAFE may affect manufacturers' prices for large and small cars, the weight and horsepower of these cars, the share of imports, the rate at which new cars replace old cars on the road and the number of miles driven, and because consumers adjust their travel patterns as fuel prices change, CAFE increases cannot translate directly into changes in petroleum use. There is therefore no consensus in the scientific community on the net reduction in fuel use attributable to CAFE.
Studies of the effects of CAFE reach very different conclusions. Greene (1989) argues that the behavior of manufacturers in the 1980s proves that CAFE is far more effective in improving new car efficiency than are higher gasoline prices. Greene and Liu (1988) find that the engineering changes undertaken by domestic manufacturers between 1978 and 1985 to increase average fuel efficiency from 18.7 to 25.8 mpg cost only $200 to $500 per car ($0.22 to $0.55/gallon over the vehicle's lifetime assuming 10,000 miles per year at a discount rate of 10 percent). This is far less than the cost of fuel saved over the vehicle's lifetime. An earlier engineering study by the Congressional Budget Office (1980) estimated that the engineering cost of improving the current 27.5 mpg standard by 18 to 27 percent would be $560 to $620 (in 1980 dollars) per car.
Economists studying the vehicle manufacturers' and consumers' reactions to CAFE have argued that size mix changes, postponement of new car purchases, and increases in miles traveled offset a large share of the apparent gains from CAFE. Leone and Parkinson (1990) constructed a simulation model of the full effects of CAFE and concluded that past CAFE increases cost between $0.52 and $0.79/gallon saved above the cost of the gasoline and that such gains could have been achieved more efficiently with an $0.08/gallon tax on gasoline. Kwoka (1983) argues that under certain circumstances CAFE could actually increase gasoline consumption due to mix and sales volume changes. Kleit (1990) estimates that increasing CAFE could initially increase fuel consumption and that further increases would reduce consumption at a cost of about $10/gallon saved.
Many of the differences between the technological costing and energy modeling studies arise because the technological costing studies do not examine feedback effects of design changes on consumer demand, vehicle safety, or vehicle miles traveled. The energy modeling studies attempt to account for such effects, but must do so using a series of assumptions that may be difficult to verify or controversial. In its present state the literature does not provide a conclusive estimate of the implementation costs of past CAFE regulations nor an estimate of how much fuel has been saved.*
CAFE standards have some advantages as a mechanism for improving fuel economy. Because CAFE standards have been in place in the past, both the government and automakers have experience in implementing them. Past experience would also indicate that they can work selectively, having
*Robert Crandall, a member of the Mitigation Panel, believes that ''the panel's conclusions on the cost of energy conservation through improvements in vehicle efficiency are excessively optimistic and likely to be misinterpreted as reflecting the cost of further extensions of CAFE. Any estimate of the full costs of a policy designed to improve the fuel efficiency of new vehicles must include all such costs, including the feedback effects on vehicle mix, average vehicle age, and as some analyses indicate, on highway fatalities."
been responsible for some of the increase in automobile efficiency. A schedule of CAFE standards focuses automotive design decisions as product lines are planned. Fuel standards can also be set for vehicles other than automobiles, including light and heavy trucks as well as commercial aircraft, with potentially significant improvements (Greene et al., 1988; U.S. Department of Energy, 1989).
Carbon or Gasoline Taxes
Carbon (or gasoline) taxes constitute another mechanism for increasing vehicle efficiency. They could affect purchasing behavior by driving up the cost of operating inefficient vehicles. Owners of inefficient vehicles would thus have an incentive to replace their vehicles sooner. Although owners of more efficient vehicles would not be paying enough to induce them to trade in their cars, they would tend to drive fewer miles and thus lower CO2 emissions. Heavy truck and airline owners would be more likely than automobile owners to replace inefficient vehicles, due to the higher number of miles traveled by these vehicles (U.S. Department of Energy, 1989).
Fuel pricing alone, however, may not be capable of achieving potential vehicle efficiency improvements. Other industrialized countries with gasoline prices 2 to 3 times higher than those in the United States nevertheless have new vehicle fleet efficiencies only slightly higher than those in the United States. Also, as vehicles become more fuel efficient, the percentage of operating costs attributable to fuel falls off rapidly. At fuel economies of 30 mpg and above, even doubling or tripling fuel costs barely increases total automobile operating costs (Ross, 1989). However, as discussed later in the transportation system management section, increasing fuel efficiency does appear to affect the frequency of vehicle trips (Pucher, 1988).
The impact of a high fuel tax may cause inequitiesespecially for those who make their living from travel or live in areas where a great deal of travel is necessary. These inequities have been exhibited in congressional debates on fuel taxesresidents in the far-reaching western United States oppose such taxes because they would affect their states more than states in the East.
Tax/Rebate Vehicle Packages
Another alternative for reducing greenhouse gas emissions is a "guzzler/sipper" tax. In this case, the consumer can be encouraged to purchase a fuel-efficient ("sipper") vehicle through sales tax reductions, income tax credits, rebates, and so on. The lost income could be made up by the purchaser of an inefficient vehicle ("guzzler"), who would have to pay a higher sales tax.
Such packages have a number of desirable features. Because the guzzler tax applies only to new cars, low-income persons will be largely unaffected. Because such taxes involve a large penalty imposed at the point of purchase, they may have a stronger effect on purchasing decisions than the desire to avoid higher fuel costs over a long period of time.
The disadvantage of this method is that it would create a large used-vehicle market that might make it difficult for new car purchasers to sell their old cars and as a result nullify the price incentive. In addition, if the cars were sold overseas, greenhouse gas emissions might increase there. Another disadvantage of this method is the question of its effectiveness. A gas guzzler tax is currently in effect for new automobiles that get fewer than 22 mpg, but the tax has been found to have relatively little effect on consumer behavior because for the mostly imported, high-performance, expensive vehicles affected, this tax is small in relation to purchase price (U.S. Department of Energy, 1989). The tax may, however, have had the effect of increasing the fuel economy of the least-efficient vehicles offered to consumers.
Alternative Transportation Fuels
Although alternative transportation fuels have not made major inroads in the United States, they have in several other countries, including Canada, New Zealand, and Brazil. California does have a fledgling program in which a move to the alternative transportation fuel methanol is being considered. In each of these cases, the motive is to improve either energy security or air quality (Sathaye et al., 1988; U.S. Department of Energy, 1989).
Despite major programs in each of these countries, only Brazil has achieved success. In Brazil a program costing the government approximately $3.7 billion, and industry $2.7 billion, over 10 years has resulted in approximately 20 percent of the light vehicle fleet using ethanol (Sperling, 1988). However, although the New Zealand and Canadian programs (costing $25 million) have been in effect for 5 years, less than 1 percent of the Canadian and approximately 11 percent of the New Zealand light vehicle fleets use alternative fuels. Each country is employing a package of policy options to encourage fuel switching. The experience of these countries indicates that a sustained and costly effort is needed for successful substitution of alternative fuels for gasoline (U.S. Department of Energy, 1989).
Emission Control Methods
Three studies (DeLuchi et al., 1988; California Energy Commission, 1989; Ho and Renner, 1990) have examined the impact of using alternative motor vehicle fuels on greenhouse gas emissions. These are summarized in Table 23.4. Note that one must consider not only the type of fuel but also the
TABLE 23.4 Summary of the Relative Greenhouse Gas Emissions from Various Fuels
source of that fuel, illustrating the need for full-cycle emission accounting. For example, methanol from wood generates low greenhouse gas emissions, while methanol from coal generates almost twice that from gasoline. In addition, Table 23.4 contains some estimates that were derived by using what the panel considers to be more realistic estimates of vehicular N2O emissions (column 5). All of the estimates in Table 23.4 include the contributions of CO2, N2O, and CH4 not only from the exhaust, but also from losses associated with production of the fuel and emissions associated with transport and distribution of the fuel.
Although the emissions of CH4 and N2O are small in comparison with CO2 emissions, their much greater infrared absorption efficiencies make their contributions to vehicular greenhouse gas emissions measurable. For example, in the California Energy Commission (1989) "low estimate," the assumed exhaust emissions from a gasoline-powered vehicle are CO2 = 311 g/mi, CH4 = 0.04 g/mi, and N2O = 0.12 g/mi. When these are converted to CO2 equivalents, the rates become 1.1 to 3.3 g/mi for CH4 and 36 g/mi for N2O, which when combined represent 11 percent of the total greenhouse gas exhaust emissions. When the production losses of CH4 are included, the total CH4 emissions represent 6 to 15 percent of the greenhouse gas emissions, and the sum of the N2O and CH4 emissions accounts for 14 to 23 percent. The large range associated with the CH4 contribution occurs because of different assumptions regarding the relative greenhouse gas efficiency of CH4. The larger values do not take into consideration the lifetime of CH4, which is relatively short in comparison with that of CO2. As a result, calculations that have not considered lifetimes may erroneously inflate the CH4 contributions. Consequently, the numbers in Table 23.4 are based only on those calculations that considered the CH4 lifetime and used a CH4/CO2 greenhouse gas efficiency of 10 to 11 on a molar basis. The difference between 10 and 11 is negligible for the purposes of this discussion.
Other differences in the numbers reported in Table 23.4 are due to different estimates of production losses and of transportation and distribution emissions and to slightly different N2O, CH4, and CO2 exhaust emission estimates. In any case, these differing assumptions have a negligible effect on the conclusions that will be made from these data. There is one exception, however: the estimate for ethanol from biomass made by Ho and Renner (1990) assumes that the fuel used to produce, transport, and distribute the ethanol is not derived from biomass, whereas the same calculation done by DeLuchi et al. (1988), and by the panel (column 5), assumes that biomass-derived ethanol is used as the fuel. This one assumption has a dramatic effect on the viability of using ethanol as a fuel. In the "more reasonable" N2O emission scenario, the panel used an N2O emission rate of 0.04 g/mi (Science and Policy Associates, Inc., 1990) rather than the 0.12 g/mi used in the California Energy Commission (1989) report. Concerning the
data in Table 23.4, note that the greenhouse gas emissions are expressed relative to present-day gasoline, which is assigned a value of 100. For those fuels that have two or more estimates, the agreement is generally within a few percent (ethanol, an exception, is discussed above). The fuels can be divided conveniently into three categories: those that will reduce greenhouse gas emissions by less than 25 percent, those that will increase greenhouse gas emissions, and those that will completely or nearly eliminate greenhouse gas emissions. Those that will reduce the emissions by less than 25 percent relative to gasoline include the following:
• diesel fuel;
• natural gas;
• liquefied natural gas (LNG) from natural gas;
• methanol from natural gas;
• clean gasoline (spiked with 25 percent biomass-derived ether);
• electricity from new natural gas power plants;
• electricity from current power plant mix; and
• reformulated gasoline.
The "clean gasoline" contains 25 percent ether derived from biomass. The benefit from burning this is not lower CO2 emissions, but the use of a renewable biomass fuel. The "reformulated gasolines" are those being evaluated to see if they can produce lower emissions of volatile organic compounds, which react to produce photochemical smog. The amount of CO2 emitted during the combustion of gasoline is a function of the carbon/hydrogen ratio in the fuel. In the reformulated fuels program, this ratio is altered slightly by changing the relative amounts of aromatics and olefins in the fuels. The mean carbon/hydrogen ratio of commercially available gasoline is about 0.55. The lowest carbon/hydrogen ratio of a fuel being used in the Auto/Oil Reformulated Gasoline Program is about 0.52. This ratio would result in a maximum CO2 emission reduction in vehicle exhaust of about 1.7 percent. Consequently, reformulated gasoline will have a negligible impact on greenhouse gas emissions.
Fuels that could result in increased greenhouse gas emissions include the following:
• methanol from coal;
• electricity from new coal-fired power plants; and
• ethanol from biomass, but produced and transported using fossil fuel.
The fuels that eliminate or nearly eliminate greenhouse gas emissions include the following:
• methanol from wood using biomass fuel to produce and transport the fuel;
• ethanol from biomass using biomass fuel to produce and transport the fuel;
• hydrogen from nonfossil-fuel-generated electricity; and
• electricity from nonfossil fuels.
If fuel from biomass is going to ameliorate greenhouse warming, biomass fuel must also be used to produce and transport it. According to Ho and Renner's (1990) calculation, this may not be possible because it appears that the energy available from ethanol is less than the energy required to produce ethanol from corn. However, there is some disagreement as to the magnitude of the energy required, which may or may not be greater than the energy provided by ethanol (Ho and Renner, 1990; Marland and Turnhollow, 1990). In one scheme, where ethanol is produced from woody biomass employing an enzymatic hydrolysis process, no external energy is required (Bergeron and Hinman, 1990); however, this process is only in the research stage. At present, we lack both the production systems and the infrastructure needed for biomass fuels to replace gasoline. The environmental impacts of converting sufficient additional land to grow corn (the preferred ethanol feedstock) or fast-growing trees (the preferred methanol feedstock) need to be considered.
For hydrogen, several problems must be dealt with before it is a viable option. The primary options are compressed hydrogen (3000 psi), hydrides (Mg2Ni or FeTi), and cryogenic liquid (20 K). These three options have respective fuel storage volumes of 15, 4, and 4 times that for gasoline, and respective fuel storage weights of 25, 17, and 1.5 times that for gasoline. Cryogenic storage, the most practical based on volume and weight considerations, has the disadvantage of evaporative loss of hydrogen, a safety and fuel economy concern. A second major problem is the lack of an economical production system. Currently, most of the hydrogen in the United States is produced by steam reforming of CH4 (from natural gas), which produces H2 and CO2. Consequently, when the energy required to produce the steam is considered, this process produces more CO2 than if the natural gas was used directly as a vehicle fuel. In order to produce hydrogen from a nonfossil fuel source, it must be produced electrolytically using non-fossil-fuel-generated power. The cost per unit energy of electrolytically produced hydrogen has been estimated to be 3 to 5 times that of gasoline. There are also some combustion-related problems. Electric vehicles (EVs) are a viable option only if the electricity is produced from nonfossil fuels. However, further technological advances, particularly with batteries, but probably also involving the electrical charging infrastructure, are needed to produce electric vehicles that will be as convenient to use as today's fossil-fuel powered vehicles. For nearly a century, battery limitations have prevented the EV from satisfying personal transportation demands on a widespread basis. The
principal problem is the batteries' low (compared to liquid fuel) specific energy, i.e., energy per unit mass. This is manifested in the short range (distance traveled before recharging is necessary) of the vehicle. Even when the thermal efficiency of the internal combustion engine is taken into account, the specific energy storage of gasoline is nearly an order of magnitude greater than that of the best contemporary battery (Amann, 1990). The specific power is also low in comparison with that provided by the combustion of liquid fuel. This manifests itself as inferior vehicle performance and long times required to recharge the batteries (hours compared to minutes to fill a gasoline tank). However, it appears that progress is being made in some of these areas. General Motors claims its "Impact" EV can accelerate from rest to 60 mph in 8 s and has a cruising range of about 190 km between charges. Nevertheless, a concerted effort has been mounted by the major auto companies and the U.S. government to explore alternative battery technology that will produce improvements in the areas cited above.
The panel attempted to determine the cost-effectiveness of these alternative transportation fuels, but found the task to be too formidable because of the numerous assumptions involved, more than for any other mitigation option discussed in this report. To properly calculate the gasoline-equivalent cost-effectiveness for alternative transportation fuels, assumptions need to be made as to the feedstock costs, production costs, capital charges, long-distance shipping costs, distribution costs, retail markup, and fuel/gasoline energy conversion rate. The cost of each of these is affected by such factors as vehicle design, available subsidies, location of markets, technological development and trade-offs, timing, magnitude of development, and many other factors. A change in any one of these can make tremendous differences in determining the cost-effectiveness. Therefore the panel decided that development of a single cost-effectiveness number would be inaccurate and that to develop one for the range of possibilities was not possible in the time available in this study and also would involve forecasts the panel was unwilling to make. This was also the conclusion of a recent (and concurrent) Office of Technology Assessment study (U.S. Congress, 1990) on this issue. Chapter 21 of that report discusses these difficulties for the case of substituting methanol for gasoline (which has more information available for it than the three fuels mentioned above).
Barriers to Implementation
Increased use of any of the alternative fuels will require the resolution of a "chicken and egg" problem: the automobile industry is reluctant to produce vehicles capable of running on alternative fuels if motorists cannot easily purchase such fuels, and the fuel industry is reluctant to create a new infrastructure for the distribution of alternative fuels if few vehicles can use
them. If alternative fuel vehicles are used first in centrally fueled fleets, however, the initial distribution infrastructure could be smaller.
One option to encourage the use of alternative fuels is a carbon tax. The impact of such a tax is likely to be immediate and to encourage the development and use of alternative fuels. The research and development of alternative fuels would be encouraged even more if the revenues of the carbon tax were used to subsidize alternative fuel development. The tax could also be used to subsidize the reduction of alternative fuel costs (U.S. Department of Energy, 1989). Past experience indicates that if an alternative fuel is not priced within a few cents of gasoline, it is unlikely to gain acceptance (Greene, 1989).
Transportation System Management
Transportation system management involves activities such as the construction of new mass transit facilities and high-occupancy-vehicle (HOV) lanes, land use planning, and development of a jobs/housing "balance" that can potentially reduce the energy consumed in transportation.
Cars and light trucks drove more than 1.7 trillion miles in the United States in 1987double the amount driven in 1965 (Motor Vehicle Manufacturers Association, 1989). The number of vehicle miles traveled (VMT) grew at a rate of 3.5 percent per year during the early 1980s, but the annual increase should slow to 2.4 percent annually as the growth in newly licensed drivers levels off due to the end of the baby boom and the saturation of women entering the labor force and becoming drivers (Ross, 1989). Nevertheless, even at this lower projected growth rate, VMT would double again by 2020.
One particularly noticeable trend is the growth in sales of, and energy use by, light trucks, a category that includes pickups, vans, and utility vehicles. Since 1982, light truck sales have increased at an average annual rate of 12 percent; they now account for nearly one-third of the light-duty vehicle market (Davis et al., 1989). Reasons for the shift to using pickup trucks as passenger vehicles may include the decreasing number of passengers in typical trips, the longer lifetime of light trucks, and the fact that they are less regulated with respect to fuel economy, emissions, and safety than cars (Ross, 1989).
Because light trucks stay on the road longer and are less fuel efficient
than automobiles, they account for about half of the fuel consumed for personal transportation (Bleviss and Walzer, 1990). Surveys indicate that approximately three-quarters of light trucks are being used essentially as carsfor commuting, passenger travel, and errands (as opposed to hauling or other activities requiring a truck's special features). The U.S. motor vehicle gasoline consumption would have been 39 percent lower in 1987a savings of nearly 50 billion gallons of gasolineif more-fuel-efficient automobiles had been used for the approximately 312 billion VMT by light trucks in situations where automobiles could perform the same task (Motor Vehicle Manufacturers Association, 1989; Ross, 1989).
Increases in VMT by both automobiles and light trucks have significantly outpaced highway construction. The number of new highway miles rose 9 percent from 1960 to 1987, a period during which VMT increased 168 percent (U.S. General Accounting Office, 1989). By 1987, nearly two-thirds of peak-hour travelers on urban interstate highways experienced delays (U.S. Department of Transportation, 1990). Even if roadway capacity is expanded by 20 percent, urban freeway congestion is projected to increase by 300 percent by 2005 (U.S. General Accounting Office, 1989). Congestion in turn will lower fuel economy, which can decline by 30 percent when an automobile is traveling at 15 mph in heavy traffic rather than at 55 mph on the open road (Davis et al., 1989).
Both demand-side and supply-side approaches are being used to deal with the growth in VMT and traffic congestion. Interest has grown in transportation demand management as a means of slowing VMT growth and reducing congestion. A growing number of governmental and private sector programs are using public information, financial incentives, parking restrictions, and provision of transportation alternatives such as carpools and vanpools to reduce peak-hour vehicle trips. On the supply side, traditional highway construction and expansion programs are being supplemented by construction of exclusive HOV lanes, park-and-ride lots, and new mass transit capacity. The most ambitious such plan has been established in southern California, where a combination of aggressive transportation demand management (to create 1.6 million new ridesharing trips and 1.4 million new mass transit trips) and construction of more than 1200 miles of HOV lanes is projected to reduce base-case VMT growth by 24 percent and hours of traffic delay by 90 percent by 2010 (Southern California Association of Governments, 1989). Evaluation of these programs and enforcement of their provisions are, of course, still in the infancy stage.
Emission Control Methods
Transportation system management uses both demand-side and supply-side measures to create a least-cost system for meeting mobility needs.
Generally, management measures are designed to divert automobile driversespecially solo commutersinto high-occupancy vehicles (buses, vanpools, and carpools) or onto mass transit (Suhrbier and Deakin, 1988; Hillsman and Southworth, 1990).
Transportation system management is useful in a CO2 emission reduction strategy for several reasons. First, shifting automobile and light truck users, particularly solo commuters, to other transportation modes can reduce energy use and thus CO2 emissions. Second, without control measures, growth in VMT could completely offset the CO2 reduction effects of improved fuel economy. Third, reducing travel demand and commuting distances can make alternative fuels more practicable by reducing the amount of fuel needed and shortening the required range of alternative fuel vehicles.
During the 1980s, substantial increases in vehicle fuel economy were more than offset by increases in vehicle use, producing net increases in transportation sector fuel use and CO2 emissions. The VMT increase of 56 million miles of automobile travel and 27 million miles of light (two-axle, four-wheel) truck travel that occurred from 1987 to 1988 (Motor Vehicle Manufacturers Association, 1989) produced nearly 40,000 t CO2 emissions. The amount by which these emissions could have been reduced depends on the alternative transportation modes the drivers would have chosen. In general, however, alternatives are less energy intensive than the automobile. Table 23.5 presents information on the different amounts of energy and vehicle occupancy for different modes of travel. As shown here, a commuter automobile carries, on average, 1.15 occupants and uses 50 to 60 percent more energy per passenger-mile than bus or rail transit.
Barriers to Implementation
A number of structural barriers, regulatory hurdles, and market imperfections must be overcome before substantial changes will occur in transportation demand, fuel use, and CO2 emissions. Among the most significant barriers to efficient transportation management are the inaccuracy of price signals in reflecting the full social cost and the lack of more-efficient transportation alternatives.
Americans seldom see the full costs of their transportation choices (whether automobile or transit) and virtually never pay these costs as transportation user charges. Instead of user fees, public and private funds pay for a substantial part of the infrastructure needed to extract and distribute gasoline, to construct and operate highways and transit systems, and to provide parking. With respect to the automobile, these hidden costs include the following:
• Highway construction and maintenance: User charges and earmarked taxes at all levels of government financed only 46 percent of highway expenditures in the United States in 1987; the remaining $36 billion came from general tax revenues (U.S. Department of Transportation, 1989). This situation is reversed in Western Europe, where revenues from taxes on gasoline and automobile ownership far exceed government expenditures for highways (Pucher, 1988).
TABLE 23.5 Energy Intensities for Passenger Travel by Mode and for Peak Travel Time
• Parking: Seventy-five percent of all commuters in the United States park in ''free" employer-provided parking spaces that actually cost employers
$1,000 to $15,000 per space to construct (plus annual maintenance and operation costs). Tax policy encourages these expenditures by treating free parking as a tax-free benefit for employees and a tax-deductible expense for the businesses that provide it (Pucher, 1988; Institute of Transportation Engineers, 1989).
• Foregone tax revenue: Close to half of the land in a typical American city is devoted to the infrastructure needed to support the automobile. More than 60,000 square miles of the United States2 percent of the total surface area of the country and the equivalent of 10 percent of all arable landis paved. Much of this land has been removed from the tax rolls and is thus indirectly subsidized by local municipalities (Ketcham and Pinkwas, 1980; Renner, 1988).
Americans have no way of knowing that they are indirectly paying these costs to support their driving habits. Indeed, the out-of-pocket costs of driving have fallen substantially in recent years. Americans paid only $0.96/gallon for gasoline in 1988, less (in real dollars) than they paid at any time during the previous decade and perhaps the lowest price ever. The cost of operating an automobile has fallen to $0.32/mi (1987 dollars), with the costs of fueling the vehicle dropping 8.4 percent annually from 1985 to 1988 (Davis et al., 1989). It is not surprising that American drivers responded to these declining out-of-pocket costs by driving more. When apparent automobile operating costs are low, Americans have no good economic reason to purchase dual- or alternative-fuel cars, to worry about the fuel efficiency of their automobiles, or to use mass transit.
Another barrier to effective transportation system management is the lack of safe, convenient, and cost-effective mass transit services in most areas of the country. Although U.S. mass transit programs served almost 9 billion riders in 1988, ridership was concentrated in a few large markets. Americans averaged only 36 transit trips per capita in 1987 (American Public Transit Association, 1989; Davis et al., 1989), one-fourth to one-half as many as citizens of other industrialized countries (Pucher, 1988).
The inadequacies of the U.S. mass transit system reflect the low priority the federal government has given these transportation modes relative to the private automobile. This priority is also reflected in the amount of public money spent to subsidize these modes. Mass transit systems, for example, received only $10.5 billion in local, state, and federal capital and operating assistance in 1988, less than one-third of the $36 billion spent on highways in 1987 (U.S. Department of Transportation, 1989; U.S. Department of Transportation, Urban Mass Transit Administration, 1989b).
A number of transportation system management options can be used to reduce greenhouse gas emissions. These include altering demand through
more accurate pricing mechanisms, through parking and transportation demand management, and by increasing the supply of HOV lanes and mass transit.
Behavioral studies indicate that out-of-pocket transportation costssuch as gasoline, parking, tolls, and transit faresare explicitly considered in daily travel decision making and that changes in such costs strongly influence travel choices (Wachs, 1989). Pricing strategies can also be used to make drivers, particularly solo commuters, responsible for more of the out-of-pocket and social costs imposed by their automobile use. The policy objective would be to ensure that each class of transportation system user pays the full marginal social costs they impose on the system. At least three types of pricing strategies are available: increasing the cost of gasoline through taxation, increasing the cost of driving through higher tolls and roadway pricing, and increasing the cost of purchasing automobiles through taxes and higher registration fees (Deakin, 1990). (Increased parking costs are treated below as a parking management measure.)
Gasoline prices in the United States have always been low in comparison with those of other countries, and they dropped drastically during the 1980s. A gallon of gasoline cost less in 1989 in real terms (corrected for inflation) than it did in 1979. In 1986, gasoline and motor oil accounted for only 15 percent of total car operating costs per mile in the United States, down from 26 percent in 1975 (Davis et al., 1989). Taxation policies strongly affect the cost of gasoline. Federal, state, and local fuel taxes in the United States total only $0.20/gallon (U.S. Department of Transportation, 1989). Per-gallon taxes on gasoline are 3 to 7 times higher in European countries, and the difference is overwhelmingly due to deliberate taxation policies, not to differences in the cost of petroleum. This price differential is, in part, responsible for the fact that cars are used for only 50 percent of trips in most Western European countries and less than 40 percent of trips in Switzerland, Sweden, Italy, and Austria, but are used for 80 percent of all U.S. trips (Pucher, 1988), although it does not appear large enough to increase incentives for vehicle efficiency improvements (Ross, 1989).
Higher gasoline prices may help reduce vehicle usage because they are the type of out-of-pocket cost consumers weigh when making travel decisions. There are, however, limits to reducing VMT through gasoline pricing, especially for owners of fuel-efficient cars for which gasoline costs are a small proportion of total operating costs. Even with gasoline priced at $2.00/gallon, the owner of a car with in-use fuel economy of 30 mpg will save only $28 per month in fuel charges by eliminating a daily one-way commute of 10 miles.
A variety of proposals have been advanced for increasing the cost of vehicle use through roadway pricing mechanisms such as tolls and congestion
charges. Graduated fee schedules could be used to encourage multioccupant and off-peak travel. One roadway pricing approach is currently being applied in the Singapore central business district (Southern California Association of Governments, 1989). Research and empirical experience in other settings are needed to determine elasticities and project drivers' responses to roadway pricing changes so that the potential emission reductions and cost-effectiveness can be calculated.
Pricing strategies that increase the cost of purchasing automobiles, such as sales taxes or registration fees, may be used to modify both the type (size and/or fuel economy) and the number of vehicles purchased. One reason for VMT growth is growth in the number of vehicles in operation in the United States. The number of vehicles per household increased from 1.55 in 1970 to 1.87 in 1987; there have been more vehicles than licensed drivers in the United States since 1985 (Davis et al., 1989). Reductions in the number of vehicles purchased due to higher sales taxes and registration fees may account, in part, for the smaller share of travel by automobile in Western European countries, which generally charge a higher sales tax on new cars than the United States does. Sales taxes in the United States average only 5 percent, while those in Europe range from 33 percent in France to 47 percent in the Netherlands to 186 percent in Denmark (Pucher, 1988).
Parking and Transportation Demand Management
Parking availability is a crucial element in the calculus of commuting: a worker who cannot inexpensively park his or her car in close proximity to work is less likely to commute alone on a daily basis. Parking for U.S. workers is frequently subsidized by employers and therefore inexpensive. Approximately 75 percent of all commuters park in free employer-provided parking spaces, and another 11 percent use free off-street parking (Pucher, 1988). Case studies representing large and small employers, from both suburban and urban areas in all parts of the United States and Canada, indicate that the provision of free parking leads to more solo commuting. When parking fees are imposed, the percentage of employees driving to work alone falls as the price of parking rises (Willson et al., 1989). Indeed, making parking very costly may work better than raising fuel prices to discourage single-occupant vehicle use.
Increasing parking costs, limiting parking, and providing incentives to carpoolers and transit users can encourage the use of mass transit and HOV modes such as carpooling and vanpooling. The removal of parking subsidies or the introduction of parking fees can reduce the mode share of solo driving by as much as 30 percent. Two recent U.S. Department of Transportation surveys have concluded that parking management is an essential element of any broader transportation demand management program (U.S.
Department of Transportation, Urban Mass Transit Administration, 1989b; U.S. Department of Transportation, Federal Highway Administration, 1990).
A variety of parking management strategies have been identified. These include the following:
• establishing a surcharge on parking for single-occupant vehicles and/or a discount for multioccupant vehicles;
• requiring preferential parking for ridesharers;
• eliminating peak-period on-street parking and/or free employer-provided parking; and
• capping the number of parking spaces permitted in a zone or for a particular project (Southern California Association of Governments, 1989).
Strategies that eliminate parking carry a negative cost for employers and developers: they save $1,000 to $15,000 for each parking space they do not provide (Institute of Transportation Engineers, 1989). Alternately, converting free or subsidized parking into paid parking provides a revenue stream for the employer, developer, or local government (if a parking tax or surcharge is used).
One possible parking management strategy would use public and private measures to reduce the amount of, and impose user charges on the remainder of, all currently "free" employer-provided parking in U.S. metropolitan areas by 1995. The 48 million workers in U.S. metropolitan areas (Pisarski, 1987) are provided approximately 36 million free parking spaces by their employers. Under the parking management program described in Appendix F, 25 percent, or 9 million, of these spaces would be eliminated. The remaining 36 million spaces would be priced (through a combination of fees, taxes, and surcharges) at a level designed to reduce the solo commuting share by 15 percent, from 65 to 50 percent. Such a parking management program would produce a cumulative CO2 reduction of 49 Mt annually at a cost of -$4.75 billion to $2.6 billion, for a cost-effectiveness value of -$97/t CO2 to $53/t CO2.*
The term "transportation demand management" (TDM) encompasses a variety of actions whose purpose is to alleviate traffic and congestion problems
*Panel members Robert Crandall and Douglas Foy have further views on the practicality of this option. Crandall believes that the existing economic literature on modal choice in transportation indicates reason for skepticism. He notes the panel's failure to measure, through existing modal-choice models, the magnitude of the fees, taxes, and surcharges that would be required to shift 9 million commuters from private automobiles to various collectivized transit modes, many of which are already heavily subsidized. Were such an analysis done, he believes that the results would show extremely large costs, not cost savings. Foy believes that there is ample empirical evidence that commuters' modal choice can be varied through relatively low cost policies affecting variables such as parking pricing and availability. He also notes that the high occupancy vehicle and transit modes to which automobile commuters switch are often publicly subsidized, but no more (and probably less) than automobile commuting.
through improved management of demand for vehicle trips. These TDM measures may be designed to eliminate trips (through telecommuting), shift automobile trips to other modes, move trips to off-peak periods, or decrease the length of trips (through jobs/housing balance mechanisms).
Transportation demand management efforts have generally been targeted at reducing traffic during peak commuting hours, when congestion is at its worst, and vehicle occupancy at its lowest. The target of most peak-hour TDM efforts is single-occupant commuting vehicles, which account for up to 60 percent of trips during these times (Pisarski, 1987; Davis et al., 1989). Reductions in such trips are relatively easy to achieve (because the occupant is traveling between the same two points at the same time daily) and are desirable because the commuting takes place during the most congested time of day (U.S. Department of Transportation, Federal Highway Administration, 1990). Shifting such trips onto transit or increasing vehicle occupancy are more productive approaches than simply moving the trips to off-peak periods.
Transportation demand management strategies now in use by public and private sector organizations include the following:
• creating a transportation management organization at the work site, with a coordinator to provide information and services to employees;
• promoting carpooling through preferential parking for carpools and computerized matching services;
• promoting vanpooling by providing vanpools (both subsidized and self-supporting);
• encouraging use of mass transit with incentives such as free or subsidized passes or shuttle buses from nearby transit stations;
• encouraging use of bicycles through provision of secure storage areas and shower facilities;
• providing a guaranteed ride home for mass transit, carpool, and vanpool users in emergency situations; and
• eliminating commuting through work-at-home and telecommuting programs (U.S. Department of Transportation, Urban Mass Transit Administration, 1989a; Southern California Association of Governments, 1989; Deakin, 1990; U.S. Department of Transportation, Federal Highway Administration, 1990).
Intensive automobile use in the United States has been driven in part by land-use and tax policies that encourage construction of low-density communities of single-family homes. Because travel demand is heavily influenced by such development patterns, reshaping land use and growth to support a variety of commuting alternatives may be critical to the success of travel demand management efforts. Land use and growth management TDM techniques are, however, future-oriented and cannot always address existing traffic system problems (Deakin, 1990).
Land use planning strategies that local governments and planners can use to manage travel demand include requiring land use plans and zoning to be consistent with transportation capacity, planning future development densities to accommodate present and future transportation capacity, and restricting uses that generate excessive numbers of trips (Deakin, 1989). Another set of regulatory approaches focuses on reducing trips associated with new developments by requiring developers to meet certain site design criteria, provide for transportation and other infrastructure, and comply with trip reduction ordinances (Suhrbier and Deakin, 1988). The Southern California Association of Governments (SCAG) has proposed one of the most far-reaching land-use-oriented TDM programsthe redistribution of some population growth into job-rich areas and some jobs growth into housing-rich areas in order to promote jobs/housing balance (Southern California Association of Governments, 1989).
Employer-based TDM programs have successfully changed commuting behavior and reduced the use of single-occupant vehicles. A recent survey by the Federal Highway Administration concluded that TDM programs, especially when undertaken in conjunction with parking management programs, should normally be able to achieve trip reductions in the range of 20 to 40 percent (U.S. Department of Transportation, Federal Highway Administration, 1990). TDM programs can be even more effective if they address some of the reasons commuters prefer to drive (e.g., the need to leave children at day care and concern that the car may be needed during the day for emergencies) through methods such as provision of day care at the work site and guaranteed ride home programs for emergencies (U.S. Department of Transportation, Urban Mass Transportation Administration, 1989a; U.S. Department of Transportation, Federal Highway Administration, 1990).
The cost of TDM programs varies widely depending on the measures used, the size of the employer, and other factors. Costs are frequently in the range of $15 to $55 per employee, but may be as high as $100 per employee per year (Municipality of Metropolitan Seattle, Service Development Division, 1989; U.S. Department of Transportation, Urban Mass Transit Administration, 1989a).
Using TDM to increase average vehicle occupancy could be a cost-effective way of reducing CO2 emissions. In the United States the average occupancy level of commuter vehicles is only 1.1 persons per vehicle. If this could be increased to 2.1 persons per vehicle, 30 to 40 million gallons of gasoline would be saved daily (American Public Transit Association, 1989). While increasing the peak-hour vehicle occupancy rate to 2.1 persons per vehicle would be a daunting (some would say impossible) task, achieving this long-term goal would result in an annual CO2 reduction of 98 to 131 Mt. Even if the TDM costs needed to achieve this goal are as high as $100 per employee per year for each of the 96.7 million workers in the United States (Pisarski, 1987), the offsetting gasoline savings of $10.95 billion per year ($1/gallon for 30 million gallons/day) would eliminate 98 Mt of CO2 emissions at a cost of -$13/t.
Pricing measures, parking management, and transportation demand management are all designed at least in part to increase average vehicle occupancy during peak commuting hours. One way of accommodating the increased number of multioccupant vehicles would be to increase the supply of specially designated high-occupancy-vehicle (HOV) lanes. Such lanes are built for the exclusive use of buses, vanpools, and (sometimes) carpools carrying three or more persons.
In recent years, construction of HOV lanes has gained acceptance as an effective means of reducing commuting time, encouraging ridesharing, and increasing vehicle occupancy rates (Southern California Association of Governments, 1989). HOV lanes can reduce commuting time on congested highways by 45 to 50 percent and also reduce the total VMT (Davis et al., 1989; Institute of Transportation Engineers, 1988). The Institute of Transportation Engineers has concluded that HOV lanes are often the lowest-cost type of fixed transit facility, with benefit-cost ratios in excess of 6 (Institute of Transportation Engineers, 1988, 1989). Despite these advantages, HOV lanesespecially when converted from existing general-purposes laneshave met with political opposition from drivers upset by the sight of seemingly unused capacity in the more free-flowing HOV lanes.
High-occupancy-vehicle lanes often cost far less than equal amounts of new highway capacity because each HOV lane (both at and well before the point at which the incentive to drive in the lane reaches zero) carries more persons than a regular highway lane. Costs generally range from $30,000 to $2 million per lane-mile when existing lanes are converted to HOV lanes and from $4 to $8 million per lane-mile when HOV facilities are constructed in existing or separate rights of way (Turnbull and Hanks, 1990).
Other Benefits and Costs
By reducing fuel use, vehicle use, and traffic congestion, transportation system management measures produce a variety of benefits in addition to CO2 emission reductions. These include reducing dependence on imported oil, lowering emissions of conventional and toxic air pollutants, and avoiding productivity losses from delays in the movement of goods and people. These benefits may total tens of billions of dollars. At the same time, there are costsincluding potential consumer welfare losses from required life-style changesincurred in adopting transportation demand measures.
Reducing transportation energy use could significantly improve the U.S. balance of trade and energy security position. The transportation sector accounts for approximately two-thirds of all U.S. oil use, an amount greater than that produced domestically (Davis et al., 1989). The direct costs of
importing petroleum products account for nearly 40 percent of the U.S. trade deficit. The United States incurs other costs because of this dependence on imported oil, including those related to using the military to protect strategic supply regions, maintaining the Strategic Petroleum Reserve, and living with the risk of supply disruptions. These costs have been estimated at between $21 billion and $125 billion annually (Greene et al., 1988), or $7 to $40 for each of the 8.3 million barrels of oil imported daily during the first four months of 1990.
Reducing VMT will reduce emissions of air pollutants, for which permissible standards are set in terms of grams of pollutant emitted per mile. The transportation sector, particularly light trucks and automobiles, is responsible for a large proportion of U.S. air pollution, particularly in urban areas. Motor vehicles emit five of the six "criteria" pollutants regulated under the Clean Air Act, including fine particulates and
• 70 percent of CO;
• 40 percent of volatile organic compounds, which are precursors to the formation of ozone, the primary component of urban smog;
• 40 percent of NOx, regulated pollutants that are precursors to both acid rain and photochemical air pollution; and
• 35 percent of lead, a potent neurotoxin particularly dangerous to young children (MacKenzie, 1988; U.S. Office of Technology Assessment, 1989). Motor vehicles are also responsible for a variety of toxic emissions (especially benzene and formaldehyde) from exhaust and gasoline vapors, all of which also could be reduced by greenhouse gas mitigation measures.
Air pollution adversely affects human health and the environment. Estimates of the total societal cost of pollutant emissions from refineries, fueling stations, and motor vehicles range from $11 billion to $187 billion annually, depending on assumptions about both the numbers of deaths and illnesses and the monetary value assigned to the loss of life (Greene et al., 1988).
Another benefit of transportation system management is that measures limiting the number of VMT or reducing the degree of congestion encountered while traveling avoid substantial productivity losses from delays in the movement of goods and people. National estimates of delay on urban highways and resulting productivity losses have been prepared by the Federal Highway Administration and reviewed by the General Accounting Office. In 1987, highway delays cost motorists $7.8 billion in wasted time. By 2005 the cost of delays is projected to increase to $29 billion to $43 billion. Another type of lost productivity is the cost of delaying trucks carrying goods. The Federal Highway Administration has estimated that the total annual cost of truck delay on highways alone is $4.2 billion to $7.6 billion; additional trucking productivity losses from congestion on urban
streets would add $19.4 billion to $22.9 billion to this total (U.S. General Accounting Office, 1989).
These estimates of $52.6 billion to $73.5 billion in annual productivity losses at current congestion levels do not include delays faced by commuters on local streets or the costs of tardiness to employers. When surveyed, employers frequently cite reductions in tardiness and absenteeism as reasons for sponsoring ridesharing programs. One survey of 20 such programs found that 18 of 20 ridesharing programs had cost-benefit ratios greater than 1 when the benefits of reduced tardiness, absenteeism, and need to construct parking facilities were calculated; cost-benefit ratios were frequently in excess of 10 (Wegmann, 1989).
Despite these benefits, there are costsprimarily in the form of consumer welfare lossesthat will likely occur due to efforts to reduce solo commuting and vehicle usage. For many workers, commuting could take longer and result in lost productivity. A typical mass transit commute in the United States averages 13 to 14 mph, whereas commuters driving alone or in a carpool travel at 31.5 to 33.5 mph while covering a comparable distance (Pisarski, 1987). (Although such losses cannot be quantified, commuters and other drivers will also suffer "consumer welfare" losses from real or perceived reductions in personal mobility and freedom.)
These consumer welfare losses can, however, be minimized by focusing transportation system management efforts on reducing commuters' travel time. Transit can be made more convenient and faster; HOV lanes can be used to reduce travel time for buses, vanpools, and carpools. Differences in commuting time between automobiles and other modes will also narrow if congestion increases automobile travel times: average vehicle speeds in southern California are projected to fall to 19 mph in the absence of transportation system management (Southern California Association of Governments, 1989). In addition, behavioral studies indicate that commuters are more concerned about travel time reliability than total elapsed travel time (Wachs, 1989); thus minimizing variations in travel time on mass transit and in the use of HOV lanes should reduce consumer welfare losses.
Research and Development Needs
The research and development needs for each of the three primary mitigation options described above are listed in this section. Recently, the National Research Council's (1990) alternative energy research and development study made the following recommendations:
• Improve batteries for vehicle propulsion to achieve higher performance and durability and to reduce costs.
• Adapt alternative fuels (e.g., alcohols) to engines and vehicles.
• Reduce emissions from efficient power plants such as the diesel.
• Evaluate vehicle systems to ensure the safety of smaller cars built with lightweight structural materials.
• Investigate innovative electric transportation systems.
The Mitigation Panel would also include hydrogen in the list of alternative fuels that need to be adapted to current engines and vehicles and would note that the focus of these alternative fuel efforts should be generation of fuels (such as methanol) from sources such as biomass and nonfossil electricity that do not increase greenhouse gas emissions. Since the method by which primary energy is generated is such an important factor in alternative transportation fuels, a number of research and development recommendations in the energy supply sector (see Chapter 24) are also applicable here.
Transportation system management has, to date, been a largely empirical exercisegovernments and private employers put measures into effect and only subsequently evaluate which work or why. Surveys, pilot programs, and other research tools should be used to predict and evaluate which methods are most effective in altering transportation behavior. In particular, the relative efficacy of different pricing mechanismsgasoline taxes, parking fees, tolls, sales taxes, and registration feesshould be evaluated (Deakin, 1990; Hillsman and Southworth, 1990).
The transportation sector accounts for 30 percent of CO2 emissions. Changes in transportation over the next several decades could substantially affect the nation's contributions to greenhouse warming. Moreover, such changes are likely to influence practices elsewhere as well. For these reasons, a focus on transportation is an important component of policy.
Of the three different types of methods for reducing greenhouse gas emissions from the transportation sector, vehicle efficiency has the greatest potential. However, implementation of vehicle efficiency measures will be slow because of the low turnover of vehicles in the marketplace and the time delay between design and production of vehicles.
Each of the alternative transportation fuels has problems. Biomass fuels require production systems and infrastructures not currently available. In addition, there is concern about the ability to convert sufficient land to grow corn for ethanol or woody biomass for methanol or ethanol.
For hydrogen, the production systems and infrastructure would have to be created. There are storage and safety concerns to be resolved for vehicles having on-board reservoirs of hydrogen. If hydrogen is to be used to reduce greenhouse gas emissions, it must be made by using nonfossil-fuel-generated energy.
Electric vehicles are a viable option only if the electricity is produced from nonfossil fuels. Further technological advances, particularly with batteries,
but probably also involving the electrical charging infrastructure, are needed to produce electric vehicles that will be as convenient to use as today's fossil-fuel-powered vehicles.
If alternative fuels are to serve as a significant alternative to gasoline in the next century, the federal government will have to both increase funds for research and development programs and tax gasoline to reflect its full social costs and make alternative fuels cost-competitive. In the interim, alternative fuels should not be used if the switch from gasoline increases life-cycle carbon emissions.
In the area of transportation system management, supply-side measures such as providing HOV lanes and increasing mass transit capacityand demand-side measures involving pricing, parking availability, and transportation demand managementcan cost-effectively and substantially reduce vehicle miles traveled, traffic congestion, fuel use, and CO2 emissions. One particularly attractive feature of transportation demand management is that CO2 emission reductions could occur during the period when fuel economy changes are being phased in; these measures will also protect fuel economy improvements in the longer term. To gain these benefits, barriers such as inappropriate transportation pricing and lack of alternatives must be overcome. In addition, steps must be taken to reduce potential consumer welfare losses due to increases in commuting time and reductions in personal freedom. Transportation system management has the potential, however, to be a cost-effective CO2 reduction strategy for small amounts of CO2 reduction while fulfilling its primary goal of providing mobility for people at the least total cost to society.
Major uncertainties stand in the way of large-scale reductions in greenhouse gas emissions from transportation. Several are discussed here because they are especially important to U.S. national policy:
• Uncertainty over the degree of willingness of vehicle owners (households and firms) to purchase smaller, more fuel-efficient vehicles when fuel prices are low poses a significant barrier to the adoption of more fuel-efficient vehicles. Ways of overcoming this barrier while preserving the dynamics of competition in the U.S. vehicle market will have to be identified, evaluated, and implemented.
• Policies aimed at altering vehicle use, including fuel pricing, transportation demand management, and changes in urban design, also face large uncertainties. Promising attempts are beginning in southern California, motivated by concern over air quality. Careful monitoring and deliberate experimentation in this and other metropolitan areas could yield important findings over the next decade.
Perhaps the largest uncertainty is the degree to which consumers and voters will be willing to respond to changing price and policy signals by
buying more fuel-efficient cars, changing commuting patterns, and otherwise altering their life-style. One should bear in mind that the emergence of current vehicle use patterns in industrialized societies occurred over the last half-century, a time span comparable with that in which greenhouse warming may occur.
1. Throughout this report, tons (t) are metric; 1 Mt = 1 megaton = 1 million tons; and 1 Gt = 1 gigaton = 1 billion tons.
2. Subsequent to the preparation of these results, K. G. Duleep (co-author with Difiglio and Greene) has refined the estimate of the ''maximum technology" point. Duleep estimates that in 1996 all available technologies could produce a CAFE mpg of 29.3 (22.5 on-road) and in 2001 all available technologies could produce a CAFE mpg of 36.0 (27.7 on-road) (Plotkin, 1991).
Altshuler, A., M. Anderson, D. Jones, R. Roos, and J. Womack. 1984. The Future of the Automobile. Cambridge, Mass.: MIT Press.
Amann, C. A. 1989. The automotive spark-ignition engineAn historical perspective. In History of the Internal Combustion Engine, ICE, Volume 8, Book No. 100294-1989, E. F. C. Somerscales and A. A. Zagotta, eds. New York: American Society of Mechanical Engineers.
Amann, C. A. 1990. Technical options for energy conservation and controlling environmental impact in highway vehicles. Paper GMR-6942 presented at the Energy and the Environment in the 21st Century Conference, Massachusetts Institute of Technology, Cambridge, Mass., March 26–28, 1990.
American Public Transit Association. 1989. Transit Fact Book. Washington, D.C.: American Public Transit Association.
Atkinson, S. E., and R. Halvorsen. 1984. A new hedonic technique for estimating attribute demand: An application to the demand for automobile fuel efficiency. Review of Economics and Statistics 66(3):416–426.
Atkinson, S. E., and R. Halvorsen. 1990. Valuation of risks to life: Evidence from the markets for automobiles. Review of Economics and Statistics 72(1):137–142.
Bergeron, T. W., and N. D. Hinman. 1990. Fuel Ethanol Usage and Environmental Carbon Dioxide Production. Golden, Colo.: Solar Energy Research Institute. January.
Bleviss, D. 1988. The New Oil Crisis and Fuel Economy Technologies. New York: Quorum Books.
Bleviss, D., and P. Walzer. 1990. Energy for motor vehicles. Scientific American 263(Sept.):103–109.
California Energy Commission (CEC). 1989. Comparing the Impacts of Different Transportation Fuels on the Greenhouse Effect. Sacramento: California Energy Commission.
Carlsmith, R. S., W. U. Chandler, J. E. McMahon, and D. J. Santini. 1990. Energy
Efficiency: How Far Can We Go? Report ORNL/TM-11441. Prepared for the Office of Policy, Planning and Analysis, U.S. Department of Energy. Washington, D.C.: U.S. Department of Energy.
Congressional Budget Office. 1980. Fuel Economy Standards for New Passenger Cars After 1985. Washington, D.C.: U.S. Congress.
Davis, S. C., D. B. Shonka, G. J. Anderson-Batiste, and P. S. Hu. 1989. Transportation Energy Data Book: Edition 10. Report ORNL-6565 (Edition 10 of ORNL-5198). Prepared for the U.S. Department of Energy. Washington, D.C.: U.S. Department of Energy.
Deakin, E. A. 1989. Land use and transportation planning in response to congestion problems: A review and critique. Pp. 77–86 in Congestion, Land Use, and Transportation Planning. Transportation Research Record No. 1237. Washington, D.C.: Transportation Research Board, National Research Council.
Deakin, E. A. 1990. Suburban Traffic Congestion, Land Use and Transportation Planning Issues: Public Policy Options in Traffic Congestion and Suburban Activity Centers. Transportation Research Circular No. 359. Washington, D.C.: Transportation Research Board, National Research Council.
De Luchi, M. A., R. A. Johnston, and D. Sperling. 1988. Transportation fuels and the greenhouse effect. Transportation Research Record 1175:33–44.
Difiglio, C., K. G. Duleep, and D. L. Greene. 1990. Cost effectiveness of future fuel economy improvements. The Energy Journal 11(1):65–86.
European Economic Community (EEC). 1988. Energy analysis of the transport sector: Determining factors and forecasts. In Energy in Europe: Energy Policies and Trends in the European Community. Brussels: Directorate General for Energy, European Economic Community.
Godek, P. E. 1990. The Corporate Average Fuel Economy Standard 1978–1990. Working Paper, October.
Greene, D. L. 1989. 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):37–58.
Greene, D. L., and J-T. Liu. 1988. Automotive fuel economy improvements and consumers' surplus. Transportation Research 22A:203–218.
Greene, D. L., D. Sperling, and B. McNutt. 1988. Transportation energy to the year 2020. In A Look Ahead: Year 2020. Washington, D.C.: Transportation Research Board, National Research Council.
Hillsman, E. L., and F. Southworth. 1990. Factors that may influence responses of the U.S. transportation sector to policies for reducing greenhouse gas emissions. Pp. 1–11 in Global Warming: Transportation and Energy Considerations 1990. Transportation Research Record No. 1267. Washington, D.C.: Transportation Research Board, National Research Council.
Ho, S. P., and T. A. Renner. 1990. The global warming impact of attainment strategies using alternate fuels. Paper presented at the Air and Waste Management Association Conference on Tropospheric Ozone, Los Angeles, Calif., March 19–22, 1990.
Holcomb, M. C., S. D. Floud, and S. L. Cagle. 1987. Transportation Energy Date Book: Edition 9. Report ORNL-6325. Prepared for the U.S. Department of Energy. Oak Ridge, Tenn.: Oak Ridge National Laboratory.
Institute of Transportation Engineers. 1988. The Effectiveness of High-Occupancy Vehicle Facilities. Washington, D.C.: Institute of Transportation Engineers.
Institute of Transportation Engineers. 1989. A Toolbox for Alleviating Traffic Congestion. Washington, D.C.: Institute of Transportation Engineers.
International Energy Agency (IEA). 1984. Fuel Efficiency of Passenger Cars. Paris: International Energy Agency, Organization for Economic Cooperation and Development.
Ketcham, B., and S. Pinkwas. 1980. Beyond autocracy: The public's role in regulating the auto. In Technology, Government and the Future of the Automobile, D. Ginsburg and W. Abernathy, eds. Cambridge, Mass.: Harvard Graduate School of Business Administration.
Kleit, A. N. 1988. The impact of automobile fuel economy standards. Working Paper 160. February 1988. Bureau of Economics, Federal Trade Commission, Washington, D.C.
Kleit, A. N. 1990. The effect of annual changes in fuel economy standards. Journal of Regulatory Economics 2(2):151–172.
Kwoka, J. E., Jr. 1983. The limits of market-oriented regulatory techniques: The case of automotive fuel economy. Quarterly Journal of Economics 97:695–704.
Ledbetter, M., and M. Ross. 1989. Supply curves of conserved energy for automobiles. Draft paper prepared for Lawrence Berkeley Laboratory by the American Council for an Energy-Efficient Economy, Washington, D.C.
Leone, R. A., and T. W. Parkinson. 1990. Conserving energy: Is there a better way? May 1990. Paper prepared for the Association of International Automobile Manufacturers, Arlington, Va.
Lyman, F. 1990. The Greenhouse Trap. Boston: Beacon Press.
MacKenzie, J. J. 1988. Breathing Easier: Taking Action on Climate Change and Energy Insecurity. Washington, D.C.: World Resources Institute.
Marland, G., and A. Turnhollow. 1990. CO2 Emissions from production and combustion of fuel ethanol from corn. Report ORNL/TM-11180. Oak Ridge, Tenn.: Oak Ridge National Laboratory.
Motor Vehicle Manufacturers Association. 1989. MVMA Motor Vehicle Facts and Figures '89. Detroit, Mich.: Motor Vehicle Manufacturers Association.
Municipality of Metropolitan Seattle, Service Development Division. 1989. Transportation Demand Management Strategy Cost Estimates. Seattle: Service Development Division, Municipality of Metropolitan Seattle.
National Research Council. 1990. Confronting Climate Change: Strategies for Energy Research and Development. Washington, D.C.: National Academy Press.
Pisarski, A. 1987. Commuting in America: A National Report on Commuting Patterns and Trends. Westport, Conn.: Eno Foundation for Transportation, Inc.
Plotkin, S. E. 1991. Testimony before Senate Committee on Energy and Natural Resources. Washington, D.C.: Office of Technology Assessment. March 20, 1991.
Pucher, J. 1988. Urban travel behavior as the outcome of public policy: The example of modal split in Western Europe and North America. The Journal of American Planners Association 54(4):509–520.
Renner, M. 1988. Rethinking the Role of the Automobile. Washington, D.C.: Worldwatch Institute, 1988.
Ross, M. 1989. Energy and transportation in the United States. Annual Review of Energy 14:131–171.
Sathaye, J., B. Atkinson, and S. Meyers. 1988. Alternative Fuels Assessment: The International Experience. Report LBL-24736. Berkeley, Calif.: Lawrence Berkeley Laboratory.
Science and Policy Associates, Inc. 1990. The Greenhouse Effect and Motor Vehicle Emissions of Greenhouse Gases. Washington, D.C.: Science and Policy Associates, Inc.
Shackson, R. H., and H. J. Leach. 1980. Using Fuel Economy and Synthetic Fuels to Compete with OPEC Oil. Pittsburgh, Pa.: Carnegie-Mellon University Press.
Southern California Association of Governments. 1989. Regional Mobility Plan. Los Angeles: Southern California Association of Governments.
Sperling, D. 1988. New Transportation Fuels: A Strategic Approach to Technological Change. Berkeley: University of California Press.
Suhrbier, J. H., and E. A. Deakin. 1988. Environmental Considerations in a 2020 Transportation Plan: Constraints or Opportunities. Proceedings of the Conference on Long-Range Trends and Requirements for the Nation's Highway and Public Transit Systems. Washington, D.C.: Transportation Research Board, National Research Council.
Turnbull, K., and J. Hanks. 1990. A Description of High-Occupancy Vehicle Facilities in North America. College Station, Tex.: Texas Transportation Institute.
U.S. Congress, Office of Technology Assessment. 1990. Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles. OTA-E-364. Washington, D.C.: U.S. Government Printing Office.
U.S. Department of Energy (DOE). 1989. A Compendium of Options for Government Policy to Encourage Private Sector Responses to Potential Climate Change. Report DOE/EH-0103. Washington, D.C.: U.S. Department of Energy.
U.S. Department of Transportation. 1989. National Transportation Statistics. Washington, D.C.: U.S. Department of Transportation.
U.S. Department of Transportation. 1990. Moving America: New Directions, New Opportunities. Washington, D.C.: U.S. Department of Transportation.
U.S. Department of Transportation, Federal Highway Administration. 1990. Evaluation of Travel Demand Management Measures to Relieve Congestion. Washington, D.C.: U.S. Department of Transportation.
U.S. Department of Transportation, Urban Mass Transportation Administration. 1989a. An Assessment of Travel Demand Approaches at Suburban Activity Centers. Washington, D.C.: U.S. Department of Transportation.
U.S. Department of Transportation, Urban Mass Transportation Administration. 1989b. National Urban Mass Transportation Statistics, 1988, Section 15, Annual Report. Washington, D.C.: U.S. Department of Transportation.
U.S. General Accounting Office (GAO). 1989. Traffic Congestion: Trends, Measures and Effects. Washington, D.C.: U.S. General Accounting Office.
U.S. Office of Technology Assessment. 1989. Traffic Congestion: Trends, Measures, and Effects. Report GAO/PEMD-90-1. Washington, D.C.: U.S. Office of Technology Assessment.
Wachs, M. 1989. Transportation demand management: Policy implications of recent behavioral research. Paper presented at the Symposium on Transportation Demand Management, Los Angeles, October 12–13, 1989.
Wegmann, F. 1989. Cost-effectiveness of private employer ridesharing programs: An employer's assessment. Transportation Research Record 1212:88–100.
Willson, R., D. Shoup, and M. Wachs. 1989. Parking Subsidies and Commuter Mode Choice: Assessing the Evidence. Los Angeles: Southern California Association of Governments.