Throughout the 1980s, vehicle emissions inspection and maintenance (I&M) programs were established in those regions of the United States with the worst pollution problems. Contrary to expectation, the programs do not appear to have achieved large emissions reductions at low costs. To improve their performance, the U.S. Environmental Protection Agency has proposed a more extensive test of a vehicle's emissions equipment than the test currently used in most I&M programs. Two components of the test favored by the agency appear to be cost-effective in reducing emissions. A third component does not seem cost-effective but may be helpful in establishing the actual emissions reductions brought about by I&M programs.
Policies to curb vehicle emissions are becoming increasingly important as one option for achieving compliance with ambient ozone and carbon monoxide (CO) standards in areas of the United States where these standards have not yet been attained. Such policies are thought to be crucial because cars remain a major source of carbon monoxide, volatile organic compounds (VOCs), and nitrogen oxide (N0x) emissions despite strict new-car tailpipe standards. These standards required emissions from the tailpipes of new cars to be reduced by more than 95 percent during the 1970s and 1980s. Over time, they were expected to result in a similar reduction in emissions from all the nation's cars. However, the introduction of alternative gasolines, the failure of vehicle exhaust equipment, and tampering with such equipment have resulted in car emissions that are much higher than originally forecast.
It has become evident that vehicles often emit higher levels of pollutants than manufacturers claim. Even as they come off the showroom floor, vehicles purported to have low emissions have been found to have emissions that are on average one and a half to two times higher than their design standard. Some vehicles have emissions levels fifty times higher than their design standard.
To reduce vehicle emissions and to enforce strict new-car tailpipe standards, vehicle emissions inspection and maintenance (I&M) programs were established throughout the 1980s in those regions of the United States with the worst pollution problems. These programs were designed to identify highly polluting vehicles through the use of a tailpipe idle test that measured exhaust concentration. Using optimistic assumptions about the accuracy of the test and the willingness of car owners to take their vehicles to be tested and repaired so as to meet emissions standards, the U.S. Environmental Protection Agency (EPA) predicted that I&M programs would achieve large emissions reductions at relatively low costs. However, there is considerable evidence that these programs have not lived up to expectation. Emissions reductions have been considerably lower than anticipated. Drivers, mechanics, and state regulators, who bear the costs but perceive few of the benefits of I&M programs, have found many ways to circumvent I&M regulations. When the time and other costs borne by vehicle owners due to I&M programs are taken into account, it appears that the cost of achieving emissions reductions through I&M programs has also been higher than early studies projected.
Problems with existing I&M programs
Some of the problems with I&M programs relate to the 2500 idle test (referring to the 2,500 revolutions per minute of an idling motor) that most programs employ. This test measures only two kinds of emissions—VOCs and CO—from only one part of a vehicle's emission system: the tailpipe. New evidence suggests that levels of evaporative emissions—that is, vehicle emissions from areas other than the tailpipe—can be as high or higher than levels of tailpipe emissions. Because the 2500 idle test measures neither emissions of NOx from the tailpipe nor any evaporative emissions, such as VOCs that escape from a vehicle's emissions recycling system due to an improperly working purge cannister or lack of pressurization in the fuel tank, EPA has recommended that the idle test be replaced with a more comprehensive emissions test.
A host of other problems beset I&M programs. One is that I&M regulations, as promulgated by EPA in 1978, give states little incentive to enforce them. Another is that many states have waiver policies that allow owners of polluting vehicles to keep operating their cars when, after paying a minimum amount of money on repairs, emissions problems persist. The effect of these policies is to exempt from repair those cars that are most polluting. Yet another problem is that it is difficult to measure the benefits of I&M programs. For most states, data on whether such programs have actually reduced vehicle emissions are lacking. What data are available suggest that I&M programs have had little effect on the level of these emissions.
Partly in response to criticism of I&M programs by the General Accounting Office, EPA began an audit of the programs around the country in the mid-1980s. It investigated 75 vehicle inspection facilities in four states with decentralized I&M programs—that is, programs in which inspections are conducted at independent gas stations. It found that 69 percent of the facilities passed vehicles that had been intentionally set up to fail the emissions test. The agency also found that improper equipment was used for testing and that as little as 50 percent of the cars registered in states that require vehicles to bear inspection stickers were being brought in for testing.
At present, inspection and maintenance (I&M) programs do not measure vehicles' evaporative emissions, are not well enforced, and, in some states, allow exemptions for some of the most polluting vehicles.
EPA tests revealed a number of technical difficulties in identifying emissions problems. For example, the tailpipe idle test sometimes indicated that cars, particularly those manufactured after 1981, had higher emissions than allowed when in fact they did not. Because the test measures neither NOx emissions from the tailpipe nor evaporative emissions, many polluting vehicles passed the test while many less polluting vehicles failed it.
Enhanced I&M programs
The poor performance of I&M programs appears to have been acknowledged in amendments to the Clean Air Act in 1990. The amendments mandate the use of vehicle emissions I&M programs through the remainder of the decade but require enhanced programs in more than 70 regions with the highest ozone and carbon monoxide levels, They also prohibit states from waiving a vehicle owner's requirement to pass the emissions test unless the owner has spent at least $450 on vehicle repairs to meet emissions standards. Before this prohibition, the minimum amount car owners had to pay for such repairs ranged from $15 in some states to $200 in others.
To improve the performance of I&M programs, EPA recommends the use of the so-called high-tech I&M test. This test includes a pressure test and a purge test to measure emissions of VOCs from a vehicle's evaporative system. It also includes the IM240 tailpipe test (referring to the 240 seconds it takes to perform the test) to measure emissions of VOCs, CO, and NOx from a vehicle's tailpipe. EPA considers the IM240 test to be superior to the 2500 idle test, which does not measure NOx emissions and is slightly less accurate in identifying vehicles with high levels of VOC and CO emissions than the IM240 test.
In addition, EPA recommends that I&M programs be centralized within regions of states. Under centralized I&M programs, vehicle inspections are conducted at state facilities devoted to that purpose. The agency favors such programs over decentralized I&M programs, under which cars are both inspected and repaired at independent gas stations, because they are less likely to be fraught with cheating.
RFF model of vehicle emissions reductions
The EPA Office of Mobile Sources uses its fleet emissions model, MOBILE 5.0, to predict the emissions reductions that will result from various I&M programs. The model has been criticized on at least two grounds. First, studies have shown that it underestimates emissions of VOCs from vehicles. Second, it does not accurately measure the emissions reductions that result from ongoing I&M programs because it fails to account for the fact that vehicles not correctly identified as polluting in one year have some probability of being so identified in a subsequent year. In light of these criticisms, researchers at Resources for the Future (RFF) have developed an alternative model to examine the impact on emissions reductions of various emissions tests that might be used in enhanced I&M programs.
The RFF model tracks the emissions performance of vehicles over time. As vehicles age, malfunctioning equipment and worn parts cause both tailpipe and evaporative emissions to increase, sometimes quite drastically. While it is possible that some high-mileage vehicles will have relatively low emissions and some low-mileage vehicles will have relatively high emissions, it is generally the case that emissions increase over the operating life of a vehicle. For this reason, the RFF model assumes that a vehicle's age is an important factor in determining the probability that the vehicle will be scrapped or move from the group of cars identified as "clean" to the group of cars identified as "dirty" in its representative fleet of 1,000 cars of varying ages and with varying emissions characteristics.
In estimating emissions reductions, the RFF model also takes into account the rates at which emissions problems are likely to be identified and solved under any given I&M program. When a vehicle is identified as dirty, it will be either repaired—whereupon it is returned to the inventory of clean vehicles—or scrapped. To the extent that vehicles repaired in any year are still low emitters the following year, the RFF model assumes that the effect of an I&M program is likely to be cumulative. It also assumes that if a polluting vehicle is not identified as dirty in a given year because emissions tests are not accurate, there is some probability that it will be so identified in subsequent years.
Evaluating the effectiveness of emissions tests
To create enhanced I&M programs, EPA has focused on the development of an emissions test that will more accurately identify those vehicles that exceed acceptable emissions levels. It claims that the new IM240 tailpipe test can identify more high-emitting vehicles than the 2500 idle test. As proof, it points out that the IM240 test can identify 100 percent of super-emitting vehicles—that is, those vehicles that emit more than 10 grams of hydrocarbons per mile—while the 2500 idle test can identify only 95 percent of such vehicles.
Although the IM240 test can identify more high-emitting vehicles than the 2500 idle test, will it have a substantially greater impact on emissions reductions than the 2500 idle test? According to the RFF model, the reductions that can be expected to result from the use of the IM240 test would not be much larger than those that would result from the use of the 2500 idle test (see figure, p. 13). In an ongoing program, as noted above, a vehicle that is not correctly identified as dirty in one year is likely to be so identified in subsequent years. From year to year, then, the difference between the reduction in emissions that would result from a more accurate test and the reduction that would result from a less accurate test would be small.
The greatest determinant of the effectiveness of an I&M program may be the response of drivers, mechanics, and state regulatory agencies to the program. The kind of emissions test used in an I&M program will affect the probability that vehicle owners will tamper with emissions control equipment, mechanics will make mistakes during vehicle inspections, or state regulatory agencies will exempt certain vehicles from either testing or repair. Because a test that has some prospect of identifying excess emissions from every vehicle will discourage tampering, mistakes, and exemptions, it is likely to effect greater emissions reductions than a test that works only on some vehicles, even if it is more accurate for that subset of vehicles.
A test that has some prospect of identifying excess emissions from every vehicle inspected will reduce more emissions than a test that works only on some vehicles, even if it is more accurate for that subset of vehicles.
As an example, consider the difference in the performance of a test that has a 30-percent probability of identifying excess emissions from 100 percent of the vehicles inspected (a "30-percent-of-100-percent" test—test A) and a test that has a 100-percent probability of identifying excess emissions from 30 percent of the vehicles inspected but has a 0-percent probability of identifying such emissions from the other 70 percent (a "100-percent-of-30-percent" test—test B). According to the RFF model of emissions reductions, each test would lower emissions by the same amount in the first year of use. In subsequent years, however, test A would effect far greater emissions reductions than would test B. By the third year, test A would reduce emissions by 37 percent, while test B would reduce emissions by only 14 percent.
Comparison of annual emissions reductions from the IM240 test and the 2500 idle test
Marginal costs of the high-tech I&M test
EPA's analysis of the costs of and emissions reductions achieved by the high-tech I&M test, which includes the IM240 tailpipe test and the pressure and purge tests for evaporative emissions, is based on optimistic assumptions. According to the agency, the cost of reducing one ton of VOC emissions by implementing the test in a biennial, centralized I&M program would be between $500 and $800. According to the RFF model, however, the cost would be about $3,300. This estimate is significantly higher than EPA's estimate because, in part, the RFF model accounts for the estimated value to car owners of the time spent driving to and waiting at inspection stations. Time costs are an important consideration because they would be substantially increased if the high-tech I&M test is implemented. Because the test requires the use of expensive equipment, it would have to be performed at more centralized vehicle inspection and maintenance stations. Thus many car owners would have to drive long distances to get to the stations. Moreover, all car owners would spend more time at the stations than at present. This is because the high-tech I&M test takes 15 to 20 minutes, while the 2500 idle test takes only 3 minutes.
The above estimates of cost effectiveness reflect the combined cost of all three components of the high-tech I&M test. Because this combined cost obscures how cost-effective each component is in reducing emissions, RFF researchers calculated the marginal costs of each component—that is, the additional cost at which each of the three tests reduces an additional ton of VOC emissions. To do this, they first estimated the increases in cost and emissions reduction that would result when the pressure test is added to the 2500 idle test. They then divided the increase in cost by the increase in emissions reduction to obtain the marginal cost per ton of emissions reductions of adding the pressure test. They then calculated the marginal cost per ton of emissions reductions of adding the purge test to the pressure and 2500 idle tests. Finally, they calculated the marginal cost per ton of emissions reductions of adding the IM240 tailpipe test to the purge and pressure tests under a scenario in which the IM240 test replaces the 2500 idle test.
These calculations revealed that only two of the three components of the high-tech I&M test may be cost-effective in reducing vehicle emissions. The marginal costs of adding first the pressure test and then the purge test appear quite low. In fact, the marginal costs of adding both tests are lower than the average cost of the 2500 idle test alone, at a cost of about $1,000 each per ton of emissions reductions, they both appear to be real bargains. When added to the 2500 idle test, the pressure test achieves an additional reduction in the emission of VOCs of about 2.5 tons for every 1,000 cars inspected; when added to the pressure test, the purge test achieves an additional VOC-emissions reduction of 2.2 tons for every 1,000 cars inspected (see figure, p. 15).
By making it possible to obtain data on the actual emissions reductions due to I&M programs, the IM240 tailpipe test might focus attention on the programs' effectiveness in achieving ambient ozone and carbon monoxide standards.
In comparison with the 2500 idle test, the IM240 tailpipe test does not appear to be cost effective. The RFF model of emissions reductions reveals that, for every 1,000 cars inspected, it reduces only 0.4 tons more VOC emissions than the 2500 idle test. The reduction is so small because the RFF model realistically accounts for the probability that cars incorrectly identified as clean in one year will be identified as dirty in the future. Since the amount of emissions reductions achieved by the IM240 tailpipe test is small, the cost of each ton of emissions reduced through the use of the test is high—$12,000 as compared with $6,000 for the 2500 idle test (see figure, p. 15).
Although the IM240 tailpipe test appears to achieve reductions in emissions of VOCs at a high cost, it does have some advantages over the 2500 idle test. First, as noted above, the IM240 tailpipe test measures NOx emissions, which may be a large contributor to ozone problems in some regions of the country. Second, it measures mass emissions—that is, the number of grams of pollutants a vehicle emits during each mile it is driven. Mass emissions (grams per mile) can be directly converted into tons of emissions in order to calculate actual emissions reductions; however, pollutant concentrations (parts per million), which the 2500 idle test measures, cannot easily be converted into tons of emissions. Thus use of the IM240 tailpipe test may make it possible to include data on actual emissions reductions in evaluations of the effectiveness of I&M programs in achieving ambient ozone and carbon monoxide standards.
To date, states have been given "credits" toward emissions reductions from I&M programs on the basis of forecasts of these reductions from EPA models of vehicle emissions such as MOBILE 5.0. If these models do not reflect actual vehicle emissions and actual reductions in emissions due to I&M programs, they cannot predict with any certainty the impact of such programs. Efforts by state authorities to ensure that I&M programs are obtaining adequate emissions reductions have not been undertaken because they do not directly influence the emission-reduction credits granted to states for implementing I&M programs. However, such efforts may be quite important for ensuring that the programs are cost effective.
Data on actual emissions reductions, which might be obtained from the IM240 tailpipe test, would focus attention on the actual performance of I&M programs. Currently, there is little effort to evaluate this performance. There are virtually no studies of the effectiveness of centralized I&M programs. Of the states with decentralized I&M programs, only California has attempted to determine the actual emissions reductions resulting from its program. It has conducted roadside emissions tests and carried out undercover investigations of I&M stations to determine whether they are passing cars that should fail the emissions test. In general, states are reluctant to impose fines on stations that conduct improper inspections and to revoke the license of drivers who tamper with their car's emissions equipment.
Given the number of as yet unanswered questions that arise in the formulation of policies and regulations to reduce emissions from vehicles, it is not clear whether it would be prudent to use the proposed high-tech I&M test in every state where EPA requires an enhanced I&M program. The full cost of conducting the test is close to $50 for each vehicle inspected. If implemented on a biennial basis in all regions of the United States where enhanced I&M programs are required, the test might cost more than one billion dollars annually. There may be less expensive but equally effective ways to achieve reductions in emissions from vehicles.
Marginal cost of adding emissions-test components under a biennial inspection and maintenance program
Winston Harrington is a senior fellow in the Quality of the Environment Division at Resources for the Future. Virginia D. McConnell is a Gilbert F. White Fellow at RFF.
A version of this article appeared in print in the January 1993 issue of Resources magazine.
A version of this article appeared in print in the June 1993 issue of Resources magazine.
