The Cost of the National Low-Emissions Vehicle Program: A Case Study By Lori D. Snyder John F. Kennedy School of Government This case was prepared for teaching purposes. While the case is based on the methods EPA used to evaluate the costs of the National Low-Emissions Vehicle Program, the exact analysis and cost figures EPA stated in their Regulatory Impact Analysis could not be replicated because some of the data used to construct these figures were not available. In addition, some alterations to the estimation method were made for clarity of exposition.
1. Regulatory Background 1.1 Clean Air Act Amendments of 1990 The 1990 Amendments to the Clean Air Act established stricter tailpipe emissions requirements for cars sold in the United States. For passenger cars, the allowable emissions of hydrocarbons were cut by 26%, while the allowable emissions of nitrogen oxides (NOx) were cut 60%. Both hydrocarbons and NOx are contributors to ground level ozone, a precursor of urban smog. Additional restrictions on emissions of carbon dioxide, particulate matter, and toxics were also implemented. Emissions standards were either tightened or adopted for other categories of vehicles including lightweight trucks, mid-size trucks, and medium and larger heavy-duty trucks. Exhibit 1 contains a list of some of the key changes in emissions standards for all vehicle classes. Exhibit 1: Clean Air Act Amendments to Tailpipe Emissions Standards Hydrocarbons NOx Carbon Monoxide 1993 and earlier 0.41 1.0 3.4 1994 to 2006 0.25 0.4 3.4 These revised emissions standards, referred to as the Tier 1 standards, took effect beginning with the 1994 model year. The Environmental Protection Agency (EPA) had the statutory authority to change these emissions standards, if research suggested more stringent standards were warranted, but such Tier 2 standards could not take effect until the 2006 model year. There was an expectation that in 2006 more stringent Tier 2 emissions standards would be promulgated. 1.2 The California Exemption When Congress passed the Clean Air Act in 1970, all states lost their right to develop individual standards for air quality controls that were regulated by the federal government. The State of California, which at the time had the dirtiest air in the country, was granted an exemption from this restriction. California can develop and enforce their own standards as long the standards are, in the aggregate, at least as protective of public health and welfare. 1 In 1993, California petitioned EPA to waive the Federal Tier 1 standards in lieu of California s Low-Emission Vehicle Program (CAL LEV). The CAL LEV program established five categories of vehicles: Federal Tier 1 vehicles, transitional low emissions vehicles (TLEV), low emissions vehicles (LEV), ultra low emissions vehicles (ULEV), and zero emissions vehicles (ZEV). The tailpipe emissions standards are increasingly stringent for each vehicle category. The CAL LEV program did not require a specific percentage of vehicles be sold in each category, but rather established a fleet average 1 Section 209(b)(1) of the Clean Air Act. 1
emission standard that had to be met by all manufacturers. This average emission level decreases every year from 1994 to 2003. In addition, the CAL LEV program mandated that 10% of large manufacturers car sales be ZEV by 2003 (CARB, 1995). 2 Exhibit 2 provides data on emissions standards for each car category covered by the CAL LEV Program. Exhibit 2: Emissions Standards by Vehicle Category for the CAL LEV program Hydrocarbons NOx Carbon Monoxide Federal Tier 1 0.250 0.4 3.4 TLEV 0.125 0.4 3.4 LEV 0.075 0.2 3.4 ULEV 0.040 0.2 1.7 ZEV 0.000 0.0 0.0 EPA approved the California request for a waiver in 1993, thereby replacing the Tier 1 standards with the CAL LEV standards in California only. Other states had the option of adopting the CAL LEV standards in lieu of federal Tier 1 standards, but could not develop their own tailpipe emissions standards. 1.3 Ozone Attainment in the Northeast The Clean Air Act Amendments of 1990 also established the Ozone Transport Region (OTR), which consists of Connecticut, Delaware, the District of Columbia, Maine Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, and a portion of Virginia. The OTR was established because the ozone attainment strategies of these states were thought to be highly interdependent, due to the nature and direction of air transport of ozone in the region. For example, if New York City is out of attainment for ozone, New York may establish strict air quality standards to reduce ozone, but if New Jersey does not also adopt these standards, ozone attainment in New York may be compromised. Ozone policy in the northeastern states is determined by the Northeast Ozone Transport Commission (NOTC), which consists of representatives from each of the states in the OTR. The NOTC has the authority to develop additional air pollution control measures to bring the area into attainment with the 1990 Ozone standards. In February 1994, the NOTC recommended that EPA mandate a low-emissions vehicle program throughout the OTR based on the CAL LEV program. The vote was 9-4 with New Hampshire, Virginia, Delaware, and New Jersey dissenting. In December 1994, EPA approved the NOTC s recommendation and promulgated new standards for the OTR states, which required reductions for new motor vehicles that would be equivalent to the CAL-LEV program. 2 At the time of regulatory development, zero emissions vehicles were assumed to be electric vehicles. Thus, the 10% ZEV requirement is often referred to as the electric vehicle requirement. 2
In March 1997, the U.S. Court of Appeals for the District of Columbia vacated the EPA regulation. The court affirmed states' rights to adopt the CAL LEV program and also acknowledged that Federal Tier 1 emissions standards may not be sufficient for ozone compliance in the OTR, but argued that EPA did not have the statutory authority to require the OTR States to adopt a low-emissions vehicle program. 3 2. National Low-Emissions Vehicle Program Development of a national low-emissions vehicle (NLEV) program had been under discussion at EPA since the early 1990s. EPA officials felt that a NLEV program would increase environmental quality and do so at a lower cost than a low-emissions program that only applied to the northeast states. A national program would allow manufactures to produce low emissions vehicles in greater volume than could be supported by California and the OTR alone. It was widely expected that this increase in volume would lower compliance costs. A national program would also harmonize testing and certification standards, allowing a vehicle to be simultaneously certified for sale in California and the rest of the country. Finally, a national program would increase the effectiveness of ozone compliance efforts in the northeast by insuring that any cars brought into the states from other areas would meet the same low emissions standards (EPA, 1997). However, EPA did not have the authority to establish a national low-emissions vehicle program. This lack of statutory authority was re-affirmed by the 1997 court ruling in Virginia v. EPA. Therefore, a NLEV program could only be implemented upon voluntary agreement by all interested parties, namely the motor vehicle manufacturers and the States. Upon acceptance of the NLEV program, however, the program would be legally binding and enforceable. A voluntary program to produce and sell lower-emission vehicles may seem infeasible, because it requires the regulated entities to agree to be further regulated. However, the automobile manufacturers did have an incentive to work with EPA and the states to craft a NLEV program that was mutually agreeable. While EPA and the OTC had failed to mandate a low-emission vehicle program in the northeast, several northeast states had already adopted or were proposing to adopt the more stringent CAL LEV standards, complete with the zero emissions vehicle mandate. There was room to negotiate a lowemissions vehicle program that would reduce the releases of ozone generating chemicals in the northeast without requiring adoption of ultra low emission and zero emission vehicles that are part of the CAL LEV program. The proposed national low emissions vehicle program would require that on average all cars and light-duty trucks sold in the United States (except California) meet the California low emissions vehicle (LEV) standards beginning in the 2001 model year. 3 Virginia v. EPA, 108 F.3d 1397 (D.C. Cir. 1997). 3
That is, by 2001, all cars and light duty trucks could emit a maximum of 0.075 grams per mile of hydrocarbons and 0.2 grams per mile of NOx. The program would be phased in beginning with the 1999 model year. In 1999, 30% of the vehicles would be required to meet the current Federal Tier 1 standards, 40% would meet transitional low-emission vehicle (TLEV) standards, and 30% would meet the low emissions vehicle (LEV) standards. In the 2000 model year, 40% of the cars sold would meet the TLEV standard and the remaining 60% would comply with the LEV standards. By 2001, all cars sold would be required to meet the LEV standards. 3. Estimating the Cost of the NLEV Program 3.1 Technology Costs The California Air Resources Board (CARB) estimated the costs of manufacturing transitional low emission, low emission, and ultra low emission vehicles as part of their analysis of the CAL LEV program. These costs were based on the best estimates of the types of technologies that would be necessary to attain the tailpipe emissions standards associated with each class of vehicle. These estimates were revised in 1996 to reflect changes in technology and technology costs for motor vehicle emissions control. Typical technologies associated with reduced emission vehicles on the market today include variable valve timing (VTEC) engines, enhanced catalytic converters, and increased precision in the air-fuel ratio control mechanisms. In 1996, CARB estimated that it would cost $72 per vehicle to attain the TLEV standard, $120 per vehicle to attain the LEV standard, and $145 per vehicle to attain the ULEV standard. 4 It was generally believed that the per vehicle costs would be lower at national production levels than at production levels that meet California demand. This was due to economies of scale in production and the manufacturers ability to spread the fixed costs of research and development over more cars. Previous studies by EPA indicated that production costs fall by roughly 7% for every doubling of production levels (Pechan, 1997). Given data on the number of cars sold by state, EPA estimated the total cost of attaining TLEV, LEV, and ULEV standards under three sets of assumptions: (1) only California adopts a LEV program; (2) all the OTR states adopt a LEV program; and (3) a national LEV program is adopted. These estimated costs are presented in Exhibit 3. 4 These cost estimates were revised in 1996 based on changes in expectations regarding the technology that would be required to attain the LEV and ULEV standards. Previously, CARB had estimated the cost of complying with the TLEV standards at $105, the LEV standards at $146, and the ULEV standards at $214 (as cited in Sierra Research, 1994). The main reason for the change in cost estimates was that by 1996 engineers no longer believed that electrically heated catalysts and heated fuel preparation systems would be required to attain the lower emission standards (CARB, 1996). 4
Exhibit 3: Estimated Cost of Emissions-Control Equipment Per Vehicle Sold by California Air Resources Board, 1996 CA Sales Volume OTR Sales Volume National Sales Volume TLEV $72 $62 $53 LEV $120 $105 $95 ULEV $145 $135 $125 Several alternative estimates of the per vehicle cost of compliance were available to EPA, including estimates by Sierra Research Inc. that were substantially higher than the CARB estimates (Sierra Research Inc., 1994). Sierra Research Inc. obtained confidential estimates of compliance costs from Ford, General Motors, Chrysler, one major European manufacturer, and one major Japanese manufacturer. Manufacturers supplied information on the following types of costs associated with the LEV program: Variable costs of emissions control: Includes cost of parts and materials necessary to comply with the emissions standard, labor costs associated with modifying the vehicles, and incremental warranty costs required to support the vehicle. Engineering research and development costs: In addition to per vehicle estimates, the manufacturers supplied information on amortization rates and periods (typically 3-5 years). Capital costs: Includes costs for developing test facilities and capital necessary to complete the modifications in-house. Overhead allocation and dealer margin: Generally a fixed allocation to cover overhead expenses and dealer markup. Performance offsets: Additional costs that will be incurred to ensure compliance with other regulations. For example, if weight is added to comply with the LEV standard the manufacturer may have to find ways to remove weight from other parts of the vehicle to ensure compliance with the Corporate Average Fuel Economy (CAFE) standards (Sierra Research, 1994). Sierra Research could not reveal the specific cost data provided by the manufacturers nor the combinations of control equipment that manufacturers used in determining their cost estimates. However, Sierra Research did evaluate the data from all five manufacturers and determine the least-cost combination of equipment that could attain the standard. These costs were assumed to decrease over time as manufacturers became more familiar with the technology and the manufacturing techniques. Exhibit 4 provides the per vehicle compliance costs estimates obtained by Sierra Research. 5
Exhibit 4: Estimated Cost of Emissions-Control Equipment Per Vehicle Sold by Sierra Research Inc., 1994 California Nationwide TLEV $412 $302 LEV $935 $709 ULEV (small vehicles) $757 $610 ULEV (mid-size vehicles) $1,475 $1,347 Table adapted from Sierra Research Inc., 1994: pp 99. In their 1994 report, Sierra Research argued that the CARB estimates were substantially understated. As an example, CARB estimated the retail price increase due to the installation of an electrically heated catalyst system at $65, less than half the cost of the uninstalled parts, based on prices received from equipment suppliers. Similarly, CARB s technology specification for attainment of the LEV standard was likely to be sufficient only for smaller cars, with larger cars requiring substantially more expensive technology to comply (Sierra Research, 1994). In the Regulatory Impact Analysis (RIA) for the NLEV program, EPA chose to use the CARB estimates rather than the Sierra Research Inc. estimates. The following quote from the RIA provides EPA s justification for this decision: EPA has chosen to utilize the CARB estimates as the basis for this RIA, rather than the significantly higher Sierra estimates, due to the fact that other information suggests that the required technology will continue to become less complex and less expensive as development continues. CARB has noted that in the early years of its LEV program, manufacturers have generally not marked up their TLEVs to account for the additional control technology. Introduction of California TLEVs into the New York market has also not resulted in price mark-ups. (EPA, 1997) EPA also cites comments from the Manufacturers of Emission Control Association (MECA) as justification for using the CARB cost figures. 5 In those comments, MECA stated their belief that conventional technology could be used to achieve the standards in a greater percentage of automobiles than anticipated when CARB developed the CAL LEV program and estimated its costs. In addition, MECA noted that historically costs for emissions control systems have fallen rapidly in successive model years after their introduction (EPA, 1995). In their analysis, Sierra Research cautioned that any argument that LEV programs are inexpensive because manufacturers have sold cars that satisfy the requirements without a markup are flawed. Manufacturers have the ability to subsidize low emissions vehicles 5 These comments were issued with respect to the original proposal for a LEV program in the OTR states. See EPA, 1995. 6
for marketing purposes and, therefore, changes in the sticker price of a vehicle do not necessarily reflect changes in the cost of production 3.2 Defining the Baseline and Policy Scenarios After the D.C. District Court vacated the EPA mandate for adoption of a low-emissions vehicle program in the OTR, several states in the OTR unially adopted the CAL LEV program. New York and Massachusetts mandated the phase-in of low emissions vehicles beginning with the 1996 model year. Connecticut also adopted the CAL LEV program, with the phase-in scheduled to begin with the 1998 model year. Rhode Island, New Jersey, and Delaware s programs would begin with the 1999 model year, and Maine would begin a phase-in of low emissions vehicles in 2001. At the time of the proposed National LEV program, the remaining states in the OTR had not adopted a low emissions vehicle program. If the NLEV program was implemented, EPA assumed all states except California would adopt the NLEV program beginning in 1999, regardless of current low-emissions vehicle requirements. This program would continue until the 2006 model year, at which time EPA could unially impose the NLEV standards as Tier 2 standards. Prior to 1999, states would still follow their individually adopted LEV implementation programs. Beginning in 2006, the seven states that previously adopted the CAL LEV program would revert to the CAL LEV standards, while the other states remain on the NLEV (or Tier 2) standards. The implementation schedules for this policy scenario are presented in Exhibits 5-A, 5-B, 5-C, and 5-D. Program In Effect CAL LEV NLEV CAL LEV Exhibit 5-A: Policy Scenario Implementation Schedule for New York and Massachusetts Percent of Vehicles Sold Model Year Federal Tier 1 TLEV LEV ULEV 1996 80 20 0 0 1997 73 0 25 2 1998 47 0 51 2 1999 30 40 30 0 2000 0 40 60 0 2001-2005 0 0 100 0 2006 and 0 0 63 37 7
Exhibit 5-B: Policy Scenario Implementation Schedule for Connecticut Percent of Vehicles Sold Program In Model Year Federal Tier 1 TLEV LEV ULEV Effect CAL LEV 1998 80 20 0 0 NLEV 1999 30 40 30 0 2000 0 40 60 0 2001-2005 0 0 100 0 CAL LEV 2006 and 0 0 63 37 Program In Effect NLEV CAL LEV Program In Effect NLEV Exhibit 5-C: Policy Scenario Implementation Schedule for Delaware, Maine, New Jersey, and Rhode Island Percent of Vehicles Sold Model Year Federal Tier 1 TLEV LEV ULEV 1999 30 40 30 0 2000 0 40 60 0 2001-2005 0 0 100 0 2006 and 0 0 63 37 Exhibit 5-D: Policy Scenario Implementation Schedule for All Other States Percent of Vehicles Sold Model Year Federal Tier 1 TLEV LEV ULEV 1999 30 40 30 0 2000 0 40 60 0 2001 and 0 0 100 0 EPA compared the costs of the policy scenario to a baseline scenario where the seven OTR states that have adopted the CAL LEV program implement this program and the remaining states do not adopt any type of LEV program. The implementation schedule for the baseline scenario is given in Exhibit 6 below. 8
Exhibit 6: Baseline Scenario Implementation Schedule for OTR States that Adopt the CAL LEV Program Year in which Implementation Rate Percent of Vehicles Sold Applies by State MA and CT NJ, RI, ME Federal TLEV LEV ULEV NY DE Tier 1 1996 1998 1999 2001 80 20 0 0 1997 1999 2000 2002 73 0 25 2 1998 2000 2001 2003 47 0 51 2 1999 2001 2002 2004 22 0 76 2 2000 2002 2003 2005 0 0 94 6 2001 2003 2004 2006 0 0 86 14 2002 2004 2005 2007 0 0 80 20 2003 and 2005 and 2006 and 2008 and 0 0 63 37 3.3 Total Cost Estimates EPA utilized data on new car registrations by state from the American Automobile Manufacturers Association (AAMA, 1996). Total number of new vehicle sales, the percentage of these sales required to be sold in different LEV class, and the total cost of emissions-control equipment for each LEV class were used to calculate the total cost of the baseline and policy scenarios. For example, under the 1999 policy scenario, 40% of the cars sold in Ohio would need to meet the TLEV standard and 30% of the cars would need to meet the LEV standard. Assuming a national LEV program, the cost of attaining the TLEV standard is $53 per vehicle and the cost of attaining the LEV standard is $95 per vehicle. There were 414,304 new cars sold in Ohio in 1995. Assuming constant annual new car sales, the total undiscounted cost of attaining the policy scenario in Ohio in 1999 is $20.6 million dollars. 6 In the baseline, Ohio does not adopt any LEV program so the total undiscounted cost of the NLEV program relative to the baseline in this state is $20.6 million for model year 1999. A more complicated example would be the state of New York. There were 491,434 new cars registered in New York in 1995. In the policy scenario, 40% of these cars would be required to meet the TLEV standard and 30% the LEV standard in 1999. The total cost of these requirements is $24.4 million. However, in the baseline, cars sold in New York would be governed by the CAL LEV standard. In 1999, the CAL LEV standard would require 76% of the vehicles meet LEV standards and two percent meet ULEV standards. The total cost of the baseline scenario in New York (using costs associated with OTC 6 Total costs for Ohio = (414,304*0.40*$53)+(414,304*0.30*$95)=$20.6 million. 9
states only) is $40.5 million. The NLEV program results in reduced emissions control costs relative to the baseline case in New York in the amount of $16.1 million. These calculations were done for all states (except California) for each model year between 1996 and 2005. California was not included in the estimation because there is no difference in the policy and baseline scenarios in that state. Costs for each model year were converted to a present value using a 10 percent discount rate. Using a 10 percent discount rate, the net present value cost of the policy scenario was estimated at $2.7 billion. The net present value cost of low emission vehicle programs in the baseline scenario was estimated at $510.3 million dollars. Therefore, the net present value of the costs associated with adopting the NLEV program relative to the baseline was estimated at $2.2 billion. 10
References American Automobile Manufacturers Associations, 1996, Motor Vehicle Facts and Figures 96. California Air Resources Board, 1996, Low-Emission Vehicle and Zero Emission Vehicle Program Report, staff Report prepared by the Mobile Source Division, California Environmental Protection Agency. Environmental Protection Agency, 1997, Regulatory Impact Analysis: National Low- Emission Vehicle Program. Environmental Protection Agency, 1994, Low Emission Vehicle Program for the Northeast Ozone Transport Region: Summary and Analysis of Comments. Pechan and Associates, 1997, Analysis of Costs and Benefits of a National low Emission Vehicle Program, prepared for U.S. EPA under contract No. 68-D4-0102. Sierra Research Inc. and Charles River Associates, 1994, The Cost-Effectiveness of Further Regulating Mobile Source Emissions. 11