r 23940 LLa z0 I-4 O0 w]cnrfi7point 01~ Emission Control Masami Kojima and Privatizing Vehlie Dnspection and Reducing Fraud in Mexico City Robert Bacon The vehicle inspection program in Mexico City is regarded as one of Masami Kojima i's a seni or eMaerspecoialistasenr the most successful in a developing country. As the program evolved, energy specialist at the World Bank focusing on it suffered the problem common to most such schemes in developing joint energy, environment, oint eneryenvirnmen countries-high levels of evasion. But the program now operates well, and transport policy issues, especially fuel quality, through high-volume, test-only centers entirely in the hands of the vehicle emission control, private sector. As an increasing number of developing country cities fuel pricing policy, and the healthimpactofpollution. consider how to enforce emissions standards as a way to control She has worked on these transport-related air pollution, Mexico City offers valuable lessons on issues in Bank projects in Latin merica Southhow to ensure compliance. zLati n America, South _n Asia, Central Asia, and the Caucasus. Robert Mexico City accounted for 3 million of the esti- Mexico (Mexico City's two jurisdictions), have Bacon is manager of the mated 7.4 million gasoline-fueled vehicles in to display a sticker showing that their vehicle has oBacon is manager of the Bank's Oil and Gas Policy Mexico in 1999. These vehicles are believed to be passed an annual emissions test or risk a fine Division. He works on responsible for Mexico City's serious ozone prob- (US$85 in 2001) imposed by traffic police energy sector reform and lem.Theresultingproductivitylosses,hospitalvis- (later,,this became the work of ecological the links between reform its and admissions, and asthma attacks and police). The sticker also shows when the vehicle strategy and poverty respiratory problems cost the economy hundreds can be driven. Since 1996 Mexico City's "day IL alleviation; the impact of of millions of dollars a year. In response to the without a car" program has placed restrictions O sector reform on the worseningairpollution, the local governmentini- on when vehicles can be operated, depending macroeconomy; and oil tiated a voluntary vehicle inspection and mainte- on their emission levels. Today there are four product taxation and nance program in 1982. The program has since categories of emission levels, with cars in the z subsidies. He has recently undergone changes to ensure stricter and more cleanest two allowed on the road any day of the worked on Bank programs reliable testing and to reduce the number ofvehi- week. The emissions standards, which define in the Islamic Republic of cles obtaining a pass certificate despite emissions pass-fail as well as the four categories, initially Irani, Mexico, and Central that are too high-a "false pass" (table 1). applied to hydrocarbons and carbon monoxide. O Asia. They were progressively tightened between 3: How the system has evolved 1994 and 1999. The changes in the standards = All motorists, whether their vehicles are regis- were aimed at removing older and more pollut- I.- 0 tered in the Federal District or the State of ing vehicles from the city. EMISSION CONTROL PRIVATIZING VEHICLE INSPECTION AND REDUCING FRAUD IN MEXICO CITY ' Year Program measure 1982 Voluntary inspection program initiated, operated by the Mexico City government. 1988 Mandatory annual emissions inspection introduced for 1982 and earlier models. Test-and-repair centers authorized. 1989 "Day without a car" program started. 1992 Mandatory test introduced for all vehicles. 1993 Test-only centers operated by the Mexico City government closed, and multi-lane "macro-centers" opened. Dynamometer test introduced for all vehicles not privately owned. 2 1994 Emissions standards tightened. 1996 Test-and-repair centers closed. New "verificenters" authorized. "Double day without a car" program started. Emissions standards tightened further. 1997 "Clean" cars exempted from "day without a car" program. More verificenters authorized. Requirement that vehicles registered in the Federal District be tested there lifted. Hybrid testing protocol (based on acceleration simulation mode) started in second half of year. 1999 New testing procedure fully adopted. Replacing catalytic converter mandatory for 1993 models. Emissions standards modified. 2000 Standards for nitric oxide introduced. Replacing catalytic converter mandatory for 1994 and 1995 models. 2001 Requirement that vehicles registered in the Federal District be tested there reimposed. Source: Energy Sector Management Assistance Program, "Mexico Energy Environment Review" (ESMAP Report 241/01, Washington, D.C., 2001). Comparing test and test-and-repair centers (they were owned by a few industrial groups spe- The annual inspections, made mandatory in cializing in emissions inspection) aided the 1988 for vehicles of a certain age, were initially adoption of new technology and led to more conducted in test-only centers operated by the uniform results among centers. city government. But soon afterward independ- Over time the quality of testing by the test- ent test-and-repair garages were authorized. In and-repair centers deteriorated. The garages 1991 the government launched a program to soon found that they could offer lower prices if create independent, multi-lane, test-only they cheated in the emissions testing, which "macro-centers" in which some testing lanes allowed them to cut back on the repair services would be equipped with dynamometers (allow- performed. In a market with surplus capacity ing simulations of loads). By 1993 there were the desire to increase profits by expanding the 500 test-and-repair centers and 24 macro- volume of business was strong, and the chances centers in full operation, all privately owned.' of being caught were small. So, while the test- The test-and-repair centers were convenient and-repair garages were convenient for users, for vehicle owners. They provided a one-stop they had considerably less impact on emissions solution, eliminating the "ping-pong" effect of than the test-only centers. Indeed, as problems a vehicle owner being caught between a garage worsened, an estimated 50 percent of the vehi- arguing that it had correctly repaired and tuned cles going through the test-and-repair centers a vehicle and a macro-center reporting that the obtained their approval certificate fraudulently. vehicle exceeded the emissions standards. As a The public began to see the emissions control result, most private vehicles went to the test-and- program as highly flawed, and it came close to repair garages. Vehicles that were not privately being shut down. owned had to go to the macro-centers for the dynamometer test, which was unavailable at the Moving to test-only verificenters test-and-repair centers. These problems led to complete restructuring of Because the macro-centers limited their ser- the program inJanuary 1996. Despite the politi- vices to testing, they were far easier for the gov- cal implications, licenses were withdrawn from ernment inspectors to supervise. In addition, all 600 test-and-repair centers, while the number the concentrated ownership of these centers of test-only macro-centers was increased from 26 to 33, for a total of 180 test lanes. The program * Eliminating the test-and-repair centers caused introduced a series of stringent quality controls a large bottleneck in the verificenters. This and technical changes in the operation of the problem was particularly evident at the end of multi-lane centers and created a new public each month, when extremely long queues identity for these centers-as test-only "verifi- formed as irate vehicle owners tried to get a centers." test during the last few days of their assigned Besides making technical adjustments to the time slot. testing procedures, the verificenters intro- * The conditions specified in the testing proto- duced elaborate precautions to prevent testers col were insufficient to warm up and ignite the from giving false passes. These included catalytic converter on many vehicles, prevent- "blind" test lanes where the tester could not see ing testers from seeing whether it was working. the test results (available only at the station * The standards were easy to circumvent by exit), central computer and video monitoring tuning late and lean. Since nitric oxide was of testing, and technical audits of centers by not being measured, there were no controls government inspectors. As a result of these against this practice. actions, the share of failing tests increased sub- In addition, the testing protocol neither gen- stantially. In the second half of 1995-before erated sufficiently stable or replicable test the restructuring-the test-and-repair centers results nor produced sufficiently low levels of had reported a rejection rate of 5.8 percent, uncertainty to allow its use with the new, tighter and the macro-centers a rejection rate of 10.3 standards introduced in 1996. percent. In the first half of 1996, under the new operating rules, the verificenters had a rejec- Strengthening quality control tion rate of 22.5 percent. The program solved the problem of inadequate capacity by increasing the number of verificen- Facing persistent problems ters to 76 and authorizing 337 test lanes. The Still, even though an estimated 73 percent of higher number of centers provided a balance vehicles obtained correct emissions certificates between the quality of service to users (such as in the first half of 1997,8 percent obtained false waiting time) and the centers' profitability. If passes because of incorrect testing practices by there were too many centers, the waiting time the verificenter, and another 19 percent would be negligible, but so would the centers' received false passes because of incorrect prac- return on investment. The low returns would tices by the garage that tuned the vehicle before create pressure for unethical behavior if that the test. Tuning vehicles "late and lean," with could improve profitability. Indeed, some cen- late ignition timing and a lean fuel and air mix- ters did modify their procedures to attract more ture, became a common practice among clients-charging no fee for vehicles that failed garages, as did disconnecting air hoses from the to pass, turning a blind eye to visual inspection intake manifold. Once the vehicle passed the failures, and, in some cases, giving false passes. test, it would be retuned. These practices some- When there were fewer centers, the compa- times reduced engine power during testing to nies had been so profitable that they were will- an undrivable level and increased nitric oxide ing to police themselves to ensure that they emissions. But the practices effectively, though could remain in this excellent business. With temporarily, reduced hydrocarbon and carbon the increase in centers, the quality of service for monoxide emissions, and they could not be vehicle owners increased dramatically, but so detected by the testing procedures used. Even did the need for government supervision. so, these rates of false passes compare very favor- Two other changes probably led to higher ably with the rates of more than 50 percent esti- emissions than would otherwise have been the mated for the test-and-repair centers. case. First, in 1997 a requirement that vehicles Despite the great improvements in the emis- registered in the Federal District be tested there sions control program by 1997, several issues was lifted, allowing the owners to choose either still needed to be addressed: the State of Mexico or the Federal District for EMISSION CONTROL PRIVATIZING VEHICLE INSPECTION AND REDUCING FRAUD IN MEXICO CITY testing. In 1998 the State of Mexico authorized E A legal and regulatory framework that allows additional verificenters and test lanes, many of independent monitoring of the testing sta- them near the Federal District. By 1999 the two tions and sanctions for failure to carry out jurisdictions had a total of 154 verificenters the testing protocols correctly. operating 572 test lanes. But they did not share E Testing protocols designed to minimize the vuewpoint a similar determination to maintain high stan- chances of testers giving false passes. dards in the emissions inspection program, and o An easily monitored certificate for passing the many vehicle owners-particularly those with test, sufficient monitors (such as traffic the most polluting vehicles-sought out testing police) to ensure a high probability of catch- encourage dissemination of centers in the State of Mexico that would issue ing vehicles without such a certificate, and a public policy innovations for a pass more easily. Between the first half of 1997 fine for lacking a certificate that is high private sector-led and and the second half of 1999 an estimated enough to act as an incentive to pass the test. market-based solutions for 500,000 vehicles that earlier would have been o Testing technology capable of preventing development. The views tested in the Federal District were tested in the temporary tuning that enables a vehicle to published are those of the State of Mexico or not at all. The requirement pass the test but that cannot be sustained for authors and should not be was reimposed in January 2001, resulting in a regular driving. In the absence of such tech- attributed to the World significant increase in the vehicles tested in the nology, motorists and garages become adept Bank or any other affiliated Federal District and in the fines collected there at circumventing the purpose of the testing organizations. Nor do any of on vehicles that had not been tested. procedure-to identify the most polluting the conclusions represent Second, during 1998 and 1999, in a move to vehicles. official policy of the World reduce the public perception of corruption in the a Equally rigorous implementation of proto- Bank or of its Executive police force, traffic police were forbidden to stop cols and inspection of procedures at all test- Directors or the countries vehicles because of inspection sticker violations. ing centers. Otherwise, owners of the most they represent. This became the prerogative of the ecological polluting vehicles can easily identify the police, of which there were few at the time. As a "softest" centers. To order additional copies result, it became possible to drive around the city a The optimal number of centers relative to contact Suzanne Smith, without a sticker with little risk of being stopped. the volume of traffic to be tested. If there are managing editor, To address technical problems in the testing too many small centers, the tests tend to Room 19-017, procedures, in 1995-96 the Mexico City govern- become less rigorous as each garage tries to The World Bank, ment defined a new protocol-the acceleration increase market share. 1818 H Street, NW, simulation mode-and put it into effect for the Washington, DC 20433. second half of 1997. The changes in the testing Telephone: protocol were aimed at generating more certain 001 202 458 7281 test results, permitting the use of tighter stan- Note Fax: dards, and reducing false approvals. Today, test 1. As a result of strong lobbying by the independent 001 202 522 3181 centers in the Federal District caught not fol- garages, the city government had closed its own test-only Email: lowing the new testing or administrative proce- centers. Private operators may be better than public for ssmith7@worldbank.org dures face a fine of up to US$40,000. several reasons, including their risk of loss of license if During the first half of 2000 the government they fail to operate by the rules, their lack of access to soft established nitric oxide limits, eliminating the budgets, and their greater flexibility in hirinig. Printed on recycled paper possibility for polluting vehicles to pass by being tuned late and lean. The limits caused too high a Masami Kojima i .,. . i,. i . 1 and Robert Bacon failure rate, however, so the government relaxed (rbacon@worldbank.org). the standards by maintaining the same emission limits but slightly modifying the testing protocol. Conclusion The experience in Mexico City shows that, to be effective, a vehicle inspection and maintenance program needs several ingredients: This Note is available online: www.worldbank.org/htmilfpd/notes/