10

Volkswagen’s Other Bug

the health impacts of diesel engine exhaust extend far beyond miners working underground. The same tiny diesel exhaust particulates that burrow into the recesses of the lungs, along with the same collection of nitrogen oxide compounds (NOx—pronounced “knocks”), are billowed into the air wherever there’s a diesel engine at work—in a commercial-trade truck, a school bus, or a diesel passenger vehicle. Whatever their source, the particulates in these diesel exhausts are known to cause cancer.

In fact, the most brazen attack on EPA diesel standards was done in the name of one company’s interest in diesel passenger cars, an episode that evolved into “the Volkswagen scandal” (aka, inevitably, Dieselgate). This is a story of profound, cinematic dishonesty. While the tricks and practices revealed in the corporate fight against the science about the particulates involved statistical manipulations of existing epidemiological studies, this new story demonstrates how the ethics of testing laboratories themselves can be easily compromised when the checks are signed by corporate sponsors.

Dieselgate is a textbook example of the pitfalls inherent in corporate funding of scientific research—and, importantly, ethical issues around animal testing. As we’ll see, it’s an important litmus test for real, well-intended science versus mercenary, paid-for science: the former is increasingly sensitive to ethical issues, while the latter doesn’t give a damn. After all, ethics are, by definition, never its priority.

In Europe, gasoline is taxed heavily and therefore very expensive. Diesel engines put out less carbon dioxide—a greenhouse gas that contributes to climate change—and European automakers convinced the regulatory authorities long ago to encourage the use of diesel fuel by taxing it at a lower level. The result is that diesel vehicles are the more economical choice in Europe. In the United States, where diesel isn’t similarly subsidized, gas prices are far lower, and so diesel vehicles have always been less popular.

This brings us to Volkswagen Group, a multinational auto manufacturer whose executives didn’t hide their ambition to become the leading auto manufacturer in the world. To do this, they needed to increase market share in the United States, where they had been relatively unsuccessful. VW’s managers in Germany recognized that sales growth across the Atlantic would depend on selling reasonably priced cars that got superior mileage and were fun to drive. Rather than take the route of Toyota, which developed the popular Prius hybrid, VW believed they could achieve the same success with diesel, an older technology they understood very well.

There was only one problem: Even in small passenger cars, a big portion of diesel exhaust are the NOx compounds. These nasty molecules irritate the airways, inflame the lining of the lungs, and increase people’s susceptibility to respiratory infections. They also trigger asthma attacks, which makes outdoor play dangerous for many kids. When NOx levels rise, emergency room visits and hospitalizations, especially for respiratory conditions, increase. NOx is also a precursor in the formation of smog and ozone. Exposure to these pollutants increases the risk of heart disease, stroke, and chronic obstructive lung disease. There is compelling evidence that exposure levels below our current standards are still causing thousands of people to get sick. The impacts are worse among the poor and vulnerable populations.¹ And as the epidemiology improves, we are seeing new associations; there is now pretty convincing evidence, for example, that exposure to pollution in utero or early in life is causally related to autism. Nor are the effects limited to humans. The nitrogen gases react to form acid rain, which acidifies lakes, damages trees and other flora, and corrodes outdoor surfaces—like limestone and marble buildings and monuments. And airborne nitrogen deposited in water systems can lead to dangerously polluted drinking water, as well as algae growths that endanger aquatic life.

While manufacturers of the big commercial diesels were making great strides in improving that heavy-duty technology, VW’s engineers teamed with the engineers at the German manufacturing firm Bosch to improve its “light-duty” (passenger car) technology. More efficient fuel injection and advanced software produced a smoother, quieter ride, with less vibration and better mileage. Most important for the atmosphere, the smoke issuing from the exhaust pipe became less and less obvious, less and less unhealthy.

But at the same time, the Environmental Protection Agency in the United States was setting increasingly stringent emission limits for NOx. Equally challenging for diesel manufacturers, the California Air Resources Board—presiding over one of the country’s largest markets for car-buying—had its own strong standards and significant penalties for breaking its rules. (In Europe, regulatory standards affecting the atmosphere are weaker than in the United States. The European Commission regulators were also strengthening their standards, but they had little teeth and the regulators had no authority at that time to issue fines for noncompliance.)

In 2008, Volkswagen rolled out a new diesel vehicle engine with great fanfare—but failed to meet the stringent new U.S. standards. Faced with pressure to make good on its technological investment, Volkswagen had an imperative to make it work.

One option was a technological fix called the Selective Catalytic Reduction (or SCR), which is essentially a filter that employs a urea-based chemical solution to trap and sequester the NOx. It would add perhaps $500 to the price of each car, and in addition motorists would be inconvenienced—they would be responsible for topping off the urea solution (a compound also found in human urine) on a regular basis.

Another option was to install the NOx trap in tandem with a special “defeat device”—software whose sole purpose was to detect and deceive the equipment in the authorized testing stations. When this software sensed that the car was undergoing emissions testing—situated on rollers, and with the steering wheel never turning—it would kick the pollution control system into high gear, injecting more of the urea solution into the NOx trap. Presto: The engine’s emissions as measured by the testing machine would be commendably low—forty times lower than under normal driving conditions in the real world. Moreover, the “defeat device” program assured that the NOx trap went into action only during testing. It would never run down and need to be refilled. The rest of the time, the car was a flagrant, deceitful polluter; the company was apparently willing to allow for that. (Why not run the NOx trap all-out, all the time to make the car practically pollution free? To keep the cost of the diesel cars competitive with gasoline-powered ones, VW engineers designed the emission control system to last only a few hundred miles, rather than the 120,000 required by U.S. regulation.)

I consider myself something of an expert on corporate wrongdoing, with an excess of great exhibits to choose from. But for clever, almost unbelievably reckless skullduggery, Volkswagen may claim the grand prize. Of course, their recklessness probably didn’t seem so at the time. Since the cars were never tested for emissions while actually driving on a real road, the engineers who designed the software understandably thought they had little to worry about. And Volkswagen certainly didn’t invent defeat devices: Many of the truck-engine manufacturers worked with similar tricks in the 1990s, varying the timing of the electronic fuel injection while the engine was being tested, thereby improving the emissions numbers. The settlement cost those firms about $1 billion, including some fines, but mostly in commitments to develop cleaner engine technologies. And they promised to never do it again.

Maybe those truck companies didn’t do it again—but Volkswagen did. The company’s illicit strategy worked for quite a few years. It was able to market its cars as both green and efficient, since they got terrific gas mileage. As such, VW became an exemplary global citizen, touting its commitment to sustainability. The “nonpolluting” claim was particularly important in the United States, where the price differential for diesel fuel was not as great as it was in Europe. Customers needed another reason to choose diesel, and the “green” claims provided it.

Sales boomed. Between 2007 and 2013, diesel-powered passenger vehicle sales in the United States jumped by a factor of six. In 2016, the German automaker that had first come to the world’s attention with its cute, oddly shaped little Beetle (initially prized famously by 1960s hippies) surpassed mighty Toyota as the number-one car manufacturer in the world. Life was good.

Too good. Back in Europe, VW, along with other manufacturers of diesel passenger cars, was still lobbying against stronger emission standards in the E.U., an odd way for them to spend their time and money given that their cars were meeting the considerably tougher requirements in the States. A much-respected advocacy group for global clean air, the International Council on Clean Transportation (ICCT), had the smart idea of using the success of the low-pollution European diesel in the United States to leverage tougher standards for Europe. And why not? If the manufacturers had, as they claimed, perfected the technology at reasonable cost, it would be a campaign for an attainable greater good.

In 2014, ICCT hired a group at West Virginia University to find some European-manufactured diesel cars stateside, hook them up to a portable testing device (one that registered real-world emissions, not those of a car on rollers) and take them on the road. The researchers—grad students, actually—had to go to California to find some European diesels because none were available in their home state. By luck (the investigators’ luck, not Volkswagen’s), they happened to select two Volkswagens: a Jetta because it had a lean NOx trap, and the Passat, equipped with an SCR. When they then drove the diesels up and down the length of that state and around Los Angeles, the NOx readings were so high everyone thought the testing devices weren’t functioning correctly. The California Air Resources Board (CARB) had a few questions for everyone involved.

This story is pretty wild, with a lot of drama, and it was dramatically laid out by New York Times reporter Jack Ewing in his book Faster, Higher, Farther: The Volkswagen Scandal.² Suffice to say that VW continually lied and gave evasive answers to the CARB staff that was trying to understand how these astronomical NOx emissions under actual driving conditions were so much higher than those measured in the testing stations. The VW officials, led by Oliver Schmidt, who headed its environmental and engineering operations in the United States, repeatedly claimed that the whole study was flawed, the calibrations were off, anything but the real reason—the defeat device. At one point, VW claimed to find some software problems and announced a recall to fix them. Incredibly, technicians then installed new software that actually made the cheating worse. It was an improved defeat device, one that better recognized when the engine was undergoing emissions testing.

By the summer of 2015, both the EPA and the California board recognized, because of the ICCT’s work, that the emissions testing results didn’t reflect real performance. But the agencies still could not completely understand why that was happening. Finally, the EPA pulled out its biggest stick: it threatened to deny certification of the company’s 2016 model diesel autos for sale in the United States.

While VW had to reveal the defeat device and admit some culpability, it initially insisted that low-level employees were responsible; ensuing investigations made it increasingly clear that this was not true. Frustrated and increasingly convinced that officials high up the ladder at the esteemed German automaker were lying, the EPA held firm, determined to find out who the individuals were. Now facing the possibility of being unable to sell any cars in the United States, Schmidt, who had earlier been transferred back to Germany, returned to California and confessed to his role in the whole dirty business.

Schmidt and other VW executives had reason to believe that by confessing, their personal and corporate penalties wouldn’t be too onerous. After all, the largest fine the EPA up to that time had been the $100 million levied against the Korean auto manufacturer Hyundai-Kia, for overstating fuel economy and understating greenhouse emissions. (That amount sounds punitive, but for a big company, it’s not much more than a slap on the wrist.) After his confession, Schmidt returned to Germany. Believing he was safe from prosecution, he returned to Florida for a vacation with his wife in January 2007. The U.S. Department of Justice had other ideas, and Schmidt was arrested at the Miami airport on his way home. He was held without bail, eventually pleaded guilty, and in December 2017 received a sentence of seven years in prison. As I type these words, he resides in the low-security federal penitentiary in Milan, Michigan. Five other VW executives have been indicted by the United States but they are unlikely to ever see jail (unless they’re dumb enough to vacation in the country where they’ve committed federal crimes).

When the dust from Dieselgate finally settled, it appeared that VW had installed defeat devices in about 11 million cars worldwide, including 8 million in Europe and about half a million in the United States. We know that when they were on the road, the NOx emissions belched into the air by VW’s diesel cars were about 40 times higher than they were when measured with their defeat devices running, and that that extra NOx has killed people; we just don’t know exactly who and how many. The best estimates indicate that approximately 1,200 people in Europe, and 59 people in the United States, died prematurely because of the excess pollution put out by the company’s cars.³,⁴ Through 2018, VW has spent more than $32 billion in criminal penalties, civil compensation, and restitution to federal and state authorities as well as consumers, and they are facing lawsuits for another $10 billion.⁵ Customers have received thousands of dollars per car in compensation for a variety of losses, including the deception itself and diminished resale value.

VW also agreed to buy back 400,000 cars from U.S. customers, most of which are sitting in 37 parking lots across the country.⁶ The company intends to remove the defeat device software from these cars, upgrade the emissions system with new hardware and modified software, and try to resell them over time, slowly, so as to not overwhelm the market. VW also agreed to pay a €1.2 billion fine in Germany.⁷ They sold these dirty diesels globally, of course, and are being punished globally as well. VW has agreed to pay up in Canada and was forced to suspend sales in Korea for two years.⁸

VW wasn’t the only European automaker that fiddled with their new diesels’ software to keep emissions remarkably low during testing. The ICCT has calculated that actual on-the-road NOx emissions of most diesel cars sold in Europe are 6 to 7 times the maximum level permitted under European standards.⁹ In January 2017, days before VW’s huge plea deal, the EPA accused Fiat Chrysler of illegally manipulating NOx emission on more than 100,000 Jeep Grand Cherokees and Dodge Ram 1,500 diesel vehicles.¹⁰ The French government has also begun prosecution of the automaker Renault¹¹ (which also manufactures Peugeot and Citroen), alleging the installation of defeat devices on almost two million diesel vehicles,¹² and both BMW and Daimler (Mercedes-Benz) cars have been accused by German authorities of doctoring their diesels.¹³

In each of these cases, the firms have denied breaking any laws. And they may be able to make a case for that, at least in Europe, where defeat devices are prohibited, except when “the need for the device is justified in terms of protecting the engine against damage or accident and for safe operation of the vehicle.” There is similar language in U.S. regulations, although the implementation of those requirements differs dramatically between the United States and Europe. The Europeans have little ability to issue fines, so that route was left to the Americans, who were willing to use it—at least until President Donald J. Trump came to power.

I don’t know of a single significant case of manufactured scientific doubt in any industry that does not also rely on product defense professionals—attorneys, mercenary scientists, and PR strategists hired to run interference, concoct science, provide cover. In the diesel car scandal, the main such facade was called the European Research Group for the Environment and Health in the Transport Sector (EUGT), a collaboration between Volkswagen, BMW, Daimler, and the engine manufacturer Bosch. It was established in 2007, likely not long after VW’s engineers realized that their new diesel engines would be unable to feature both high mileage and low emissions and came up with the defeat device. As such, the European automakers’ new “sound science” team was charged with “examining the effects of and interaction between emissions, air pollution, and health and to find ways of avoiding possible health consequences.”¹⁴ A more accurate description was later offered by the German news magazine Der Spiegel, which in its assessment of EUGT’s mission called it “a joint lobby organization that was disguised as a research institute.”¹⁵

The names and backgrounds of the group’s leadership staff demonstrated that the new outfit was in no way interested in independent research. Its director, Michael Spallek, had previously spent years employed as an occupational medicine physician at Volkswagen. (He retained his VW email address, even after his move to EUGT.)¹⁶ The results of the institute’s research were accordingly one-sided. Spallek himself co-authored several of the lobby’s attacks on the classification of diesel exhaust particulate as a human carcinogen, along with other pieces that questioned whether breathing diesel exhaust was even so bad for people.

The low-emission zones in cities that place restrictions on driving cars with high emissions? There’s no proof they’re effective, according to one study Spallek and his coworkers produced.¹⁷ (Establishment of these urban zones is one of the most promising policies to reduce exposure to diesel emissions. Starting in Stockholm in 1996, scores of cities across Europe have designated areas limiting entrance of various vehicles depending on the pollution they cause.¹⁸ Unfortunately, one of the first evaluations of the zones—negative, of course—was the one published in 2014 by EUGT scientists and disseminated by the industry’s lobbyists and PR teams.) Nighttime noise pollution from cars? It’s no problem, as long as the racket is continuous. Do diesel emissions cause cancer? Can’t be proved. Also recall from Chapter 6 that Spallek’s EUGT was one of the funders of the mercenary re-analyses meant to demonstrate the uncertainties of the U.S. government’s study that found elevated lung cancer risk among diesel-exposed miners.¹⁹ In addition to Spallek, other EUGT scientists also authored papers attacking the classification of diesel particulates as carcinogenic.²⁰

There’s a famous quote from the American writer Upton Sinclair that says, “It is difficult to convince a man of something if his salary depends on him not believing it.” Psychologists label this phenomenon “motivated reasoning.” Our motivations influence our reasoning. In humans, it’s a hard fact of life. There is no question that being paid by a polluter (in what is probably a lucrative financial relationship) changes the way a scientist looks at the scientific literature. And maybe those scientists working for EUGT truly believed that all the studies completed by academic and government scientists were wrong—that exposure to the soot spewing out of diesel tailpipes containing dozens of carcinogens does not increase risk of cancer. Maybe. But when their research is paired with the flagrant and concurrent scam of installing defeat devices in new diesel engines, it is certainly reasonable to ask what the EUGT leadership knew. After all, it is now clear that hundreds of high-level executives and managers at the German automakers were aware of the subterfuge. And we now also know that even after the scandal was front-page news, EUGT scientists were doing their best to pollute the scientific literature with additional rigged studies, ones designed to make the new “clean diesel” engines look safe—even as their dangers were slowly being revealed.

Much about the VW/EUGT episode was flagrant, but one particular episode demonstrates how far at least some of the industry’s collaborating scientists were willing to go to ensure that the scientific literature was polluted with phony studies exonerating diesel exhaust. It ended very badly for the company and the industry, scientifically and ethically. It’s worthy of a very close look.

In 2012 the World Health Organization’s prestigious International Agency for Research on Cancer (IARC) surveyed the rapidly accumulating studies around diesel exhaust and was prepared to officially label the particulates as a known human carcinogen. Several of the industry’s consulting scientists (including those paid by EUGT) attended the IARC meeting where the announcement was to be made. Since IARC limits the role of scientists with conflicts of interest, these attendees were present only as nonvoting observers. These consultants had already proclaimed that any evidence linking diesel exhaust to cancer was faulty, and on the very day of the IARC announcement their employers were prepared with press releases that challenged the new rating and offered expert interviews rebutting IARC’s claims to interested media. Their primary message was that the diesel engines whose exhaust was carcinogenic were the old dirty ones, not the newer ones made by modern engine manufacturers—a standard industry talking point for years.

But the automakers wanted more than a strenuous rebuttal. They also wanted a positive PR message. They wanted to emphasize that the new engines constituted progress. The U.S. corporate communications manager of VW subsidiary Audi sent an email up the chain, asking for help internally with “counter messaging.”²¹ And the best way to do that: a laboratory study where humans would be exposed to exhaust from the new “clean” engines to demonstrate their safety. But there was a major hurdle: The IARC classification concerned lung cancer, which develops over many years. To disprove any cancer link, a study would therefore have to give the human subjects huge doses of particulate-filled diesel exhaust and follow them for decades. The auto manufacturers didn’t have time for that.

Instead, VW decided to study something more innocuous: short-term exposure to NOx emissions, which were well controlled in the new technology engines (at least when their defeat devices were operating). The original plan was to find human volunteers, put them in a chamber where they would ride stationary bicycles to increase respiration, then expose them to air mixed with diesel exhaust (with the particulates filtered out) at levels not thought to cause permanent lung damage. This might have been a reasonable study, yielding what should be an obvious and predictable result, if it had been a true comparison of the old (dirty) and new (cleaner) technologies.

But that was never the idea. VW executives did everything in their power to rig the results. And in doing so, at least some of their staff demonstrated a complicity with the defeat devices that were then in mass production. While VW assigned the task of commissioning and overseeing the study to the EUGT, the company was intimately involved from the start. It was their attorneys who vetoed the idea of using human volunteers on exercise bicycles, perhaps because research involving putting people in sealed chambers and then pumping in gas might bring back memories that VW, a firm whose history was so closely entwined with the Nazi regime, would prefer to forget.²²

Plan B was to substitute monkeys for humans. Spallek, the Volkswagen physician turned EUGT’s head man, reached out to the Lovelace Respiratory Research Institute, a private nonprofit research lab in New Mexico also featured in Chapter 6. In exchange for $718,572, Lovelace agreed to a study involving 10 male monkeys. The object would be to compare the effects of breathing air containing exhaust emissions from a new-technology diesel to the exhaust from an old-technology diesel. Both emission samples were diluted with filtered air. Some of the monkeys were exposed to the old engine exhaust, some to the new technology, and some to filtered air only. After each exposure, the animal would undergo medical testing, with a particular focus on the presence of any lung inflammation.

The deal between EUGT and Lovelace was signed in August 2013. From the outset, the ground rules were scientifically and ethically dubious. For starters, the contract required Lovelace to provide a final report, but also required strict confidentiality concerning the results.²³ There’s precedent for why this is frowned upon: in 2001, following a series of scandals in which drug companies refused to let researchers publish findings the firms didn’t want public, the editors of 13 of the world’s leading biomedical journals announced they would henceforth publish only studies conducted under contracts in which the investigators are “free of commercial interest.” In short, those editors would no longer accept papers presenting findings from studies such as this Lovelace one on diesel exhaust—that is, ones performed under contractual terms that allowed the sponsor to control whether the results could be published. In a joint statement in 2001, the journal editors asserted that such contractual arrangements “not only erode the fabric of intellectual inquiry that has fostered so much high-quality clinical research, but also make medical journals party to potential misrepresentation, since the published manuscript may not reveal the extent to which the authors were powerless to control the conduct of a study that bears their names.”²⁴

So while EUGT’s deal with Lovelace was bound to raise significant ethical concerns about both parties, the eyebrow-raising didn’t stop at the terms of the contract. The most impassioned reporting began with the English-language edition of Der Spiegel in July 2018, a year after the study wrapped up. Other newspapers followed, and their reporting focused on the problematic study design. First, these were monkeys, raising ethical issues about using nonhuman primates to test toxic exposures. Second, they watched cartoons on TV while they breathed, because the viewing calmed them down. The image of the laboratory monkeys watching TV while breathing engine exhaust makes an unforgettable impression.²⁵

What the journalists could not have known is that VW and EUGT rigged the study with the default device. (This was a couple of years before that scam was discovered and revealed.) VW worked closely with Lovelace researchers to ensure that the experiment would give VW leadership exactly the result they wanted, allowing them to point to what looked like a reputable study, falsely assuring us that their engines didn’t cause health damage. It did not work out as they planned.

To do this, the Volkswagen-funded EUGT had to provide the researchers with appropriate vehicles that would give the desired results. The Germans insisted that the old-technology diesel chosen as the comparison engine should not be made by a German manufacturer, because they didn’t want to be associated with an admittedly dirty vehicle.²⁶ So they found and bought a 1997 Ford F-250 pickup truck, which by then was 15 years old. For the new engine they went with a VW diesel Beetle. James Liang, a Volkswagen engineer who helped get the Lovelace study off the ground, selected and personally drove a new red diesel Beetle from Los Angeles to the Lovelace lab in New Mexico. To make sure the defeat device was functioning to perfection and delivering the minuscule NOx emissions numbers the company needed, Liang requested that Lovelace install a signal booster to help transmit real-time data from the engine directly to him in his California office. (This was an added cost, not to mention a wild breakdown in the wall between funder and experimenter, and Lovelace received assurance that EUGT would pay for that equipment.²⁷)

VW engineers also helped obtain the equipment for the study that mimicked conditions in emissions-testing facilities: wheels moving, but remaining in place, signaling to the Beetle’s defeat device to kick on. Stuart Johnson, who worked with Oliver Schmidt in the engineering and environment office of the Volkswagen Group of America, and who would later replace Schmidt when Schmidt was sent back to Germany (before he was arrested at the Miami airport), picked out much of the equipment used in the testing.²⁸ Liang made sure it was properly installed and shipshape, most especially the defeat device, which would guarantee its NOx emissions would be tiny.²⁹

This was not Liang’s first experience with these sorts of efforts. He helped design VW’s original defeat device software years earlier. After that triumph, he was sent to the firm’s testing facility in Southern California to calibrate the defeat device sold in the States so it would recognize the U.S. test’s drive cycles. His title at the company: Leader of Diesel Competence. Presumably, the scientists at Lovelace did not know about the defeat device. Presumably, they did not know that this engine’s exhaust emissions were guaranteed to be very low.

The 2018 Der Spiegel story on the Lovelace study mentions a small-town Virginia attorney, Michael Melkersen, who was one of the many lawyers suing VW on behalf of consumers misled into purchasing cars that were heavy polluters. Before Melkersen had come across a reference to the study in one of the documents he was reviewing for the case and started to issue subpoenas for additional ones. Eventually, he took depositions of several of the actors in the drama.³⁰ Thanks to the unethical confidentiality clause in the contract, this story might never have seen the light of day if not for the lawsuits and, more specifically, Melkersen’s digging through the evidence. The materials Melkersen uncovered document the comedy and tragedy of the ill-fated research.

Setting aside the ethical issues of using monkeys, not to mention the problematic terms of the funding and contract, the actual design for the study was reasonably competent. But reading through the memos and the depositions, especially those of Lovelace’s lead scientist, Jacob McDonald, one can reasonably conclude that the actual execution of the study was a comedy of errors throughout. Most ridiculous were the results: even though it was carefully rigged to “prove” that the new diesel engines, such as those in the red Beetle, were exponentially cleaner than the old engines, like the one in the Ford truck, the results demonstrated just the opposite: when inhaled by the monkeys, the new Beetle’s exhaust caused more lung inflammation than the old truck’s exhaust, which was 180 times more powerful in terms of NOx exposure. This made no sense whatsoever.

By 2015, when the Volkswagen scandal was breaking and Lovelace had yet to publish anything from their study, the New Mexico researchers learned of the defeat device the same way everyone else did: in the news. As McDonald, Lovelace’s lead scientist in the study, said at the time, “I feel like a chump.”²⁹ But inexplicably, rather than just determining that their VW diesel had been tampered with and the study was therefore invalid, the Lovelace researchers continued with the project. EUGT made clear that it had paid good money and wanted what it had paid for, even as the walls were closing in on Volkswagen.

I’ve seen a lot in the product defense arena in my time. Very little compares to what is contained in the memos and drafts that were uncovered via Melkersen’s subpoenas. The public likes to think that scientific research is a straightforward exercise, where the scientists accurately report their findings without shading the results. But through these documents we see how wanting to please the sponsor (and get paid) changes how the results are reported. I can imagine a little of how the Lovelace staff felt, because their world had just been turned upside down. They were dealing with two seemingly contradictory facts, both of which made them look like chumps. For one, they learned that the VW’s NOx emission levels were fraudulently low, and their study was rigged from the start. And at the same time, the results that their execution of the rigged study produced were the opposite of what they expected: the fraudulently low levels of emissions from the new Volkswagen diesel seemed to cause more inflammation than the poison spewing from the old pickup truck.

What could Lovelace do, especially considering that EUGT was withholding $71,000 (10 percent of the contractual amount for the study) until the lab fulfilled the contract’s requirements, one of which was to publish the results in a peer-reviewed journal? Well, one thing they could do, and did, was continue analyzing the data and preparing a report and abstracts, reporting with a straight face the rigged study’s finding. The Lovelace staff prepared the first draft of an abstract for presentation at the Society of Toxicology’s 2016 annual meeting. It ended with the statement, “Sample analysis continues but, contrary to the hypothesis, [new technology diesel] appears to have induced great inflammation by measurement of several key parameters.”³¹

Perhaps recognizing this was not what the study’s sponsors were looking for, someone at Lovelace suggested removing the word “great” and disguising their observation that the new engine caused more inflammation by “stating that inflammation was observed after both exposures.” The final version of the abstract that appeared in the meeting’s actual proceedings—the only version seen by the public—more or less gave EUGT what it wanted. The investigators dropped any mention of having even tested the new-technology diesel exhaust, reporting instead that old-technology diesel caused somewhat more inflammation than simply exposing the monkeys to regular filtered air with no exhaust at all.³² It is hard to believe this could ever be accepted in a scientific journal; the results are painfully predictable and reviewers would correctly ask how Lovelace’s Institutional Animal Care and Use Committee, which reviews studies to protect the welfare of laboratory animals, would have let this study be conducted.

The final reports Lovelace prepared for EUGT also went through a series of changes to make it acceptable to their German funders. The first draft acknowledged that the inflammation was worse in the monkeys exposed to the ridiculously low levels of NOx. The researcher who prepared it explained in an email to McDonald, “I was trying to soften the blow from the results of the study without saying it was a bad study.”³³

Nice try. When that version of the final report was shared with EUGT, the response was predictably negative. The trade group responded with a series of questions and comments that they insisted be addressed before Lovelace could be paid in full. (Along with Spallek, the board that issues these demands included several staunch diesel defenders, who individually had written several papers raising doubts about the studies linking diesel particulates with lung cancer. My favorite is entitled “The European ‘Year of the Air’: Fact, Fate or Vision.”³⁴)

Did EUGT and its chief Michael Spallek know from the beginning that the Lovelace study was conducted with rigged equipment meant to produce convenient results? We don’t know and no one from EUGT has been charged with any wrongdoing. But the documents show that, after the widespread rigging of Volkswagen diesels became front page news across the globe, Spallek still pressured the investigators to publish a paper making the new technology engine appear to be safe to human health. Even though the contract said that the final 10 percent would be paid on submission of the final report, McDonald explained to the contracts office at Lovelace, “We submitted this final report several months ago and they have disputed it because it did not meet their expectations in terms of outcome.” In other words, EUGT was holding the $71,000 until they received the final report with the results they wanted.³⁵

So what went wrong with the Lovelace test? How was it possible for the monkeys with the minuscule NOx exposure to have more lung inflammation than those heavily exposed? It may have been the decision to purchase 10 female monkeys rather than the male monkeys stipulated in the contract. From the Lovelace lab’s point of view, there were practical reasons for this substitution—females are less expensive and less aggressive, therefore easier to handle—but there is also more variability in female monkeys’ inflammatory lung response, especially during menstruation.³⁶ Alternatively, the finding may have been the result of poor execution by the investigators. For each monkey, the scientists did a baseline lung lavage—essentially washing out the lung with a saline solution to measure inflammation—the days before they were exposed to the NOx mixture. It is possible that testing after exposure measured not new inflammation caused by NOx, but inflammation triggered by the baseline examination. Whatever it was, the researchers decided the data were, in lead scientist McDonald’s succinct assessment, “garbage.”³⁷

Garbage in, garbage out? The old saying is true and, from Lovelace’s perspective, another reason to throw this whole thing out. Many scientists would have done just that. But in New Mexico, under pressure to modify the study’s findings and prepare a manuscript for publication in order to get that final payment of $71,000, and only two months after McDonald’s one-word condemnation, he wrote to EUGT’s Spallek that he was modifying the final report with additional “end points where we saw the increase in ‘effect’ in the old technology. The endpoint a [sic] we observed are consistent with our hypothesis about diesel and lung injury.”³⁸ (Keep in mind, this is a full year after the scandal broke.)

In other words, this was classic product defense manipulation of the raw data. Spallek approved fiddling with the data, of course, but still wanted the Lovelace team to get the results into the scientific literature before he’d cut that final check.³⁹ McDonald wrote to another Lovelace scientist in an email, “I need to publish a paper and basically I will have to throw out the lavage data[,] then I have three figures … and a bunch of aerosol stuff … so I am trying to see if I can squeeze out something else that may be interesting and says ‘old diesel bad, new diesel good’ so I can win the nobel prize. [ellipses original]”³⁷

I believe we can take McDonald’s Nobel Prize reference as deeply sarcastic. The final report from the Lovelace study, sent on June 30, 2017, by McDonald to Stuart Johnson at VW and Michael Spallek at EUGT, provides the opposite conclusion of the first draft. In fact, Lovelace gives VW and EUGT the conclusion they wanted from the start: “Based on the results shown here, the [old technology diesel—the 15-year-old Ford F250] showed an increase in inflammation systematically and in the lung, while [new technology diesel—the new VW Beetle] did not.” The email accompanying the report included a commitment from McDonald to submit the data to the journal Inhalation Toxicology and requested yet again the final payment of $71,000.⁴⁰ But Lovelace was out of luck. Pilloried in the press for sponsoring the monkey study, EUGT could no longer serve a useful role for the German automaker. It was dissolved days before McDonald sent in the final version of the report, the one he thought the German automakers would find to their liking. It has never been published, beyond the misleading abstract published for the 2016 conference.

I shed no tears for Lovelace. They may have gotten a small raw deal in the end, but it also seems apparent that the principals at the lab were willing to try their best to publish misleading results in order to get their final $71,000. They did not succeed, but this episode raises important questions about testing labs in general.

When the tests with the monkeys made world headlines, animal rights groups were furious. Studies had already been done exposing human volunteers to low levels of NOx; there was no scientific need for additional studies with monkeys to learn the same thing. Executives with VW, BMW, and Daimler were shocked, shocked that the study even took place. Volkswagen CEO Matthias Muller called the study “unethical and repugnant.”

And how about “rigged”? The monkeys had been put through an experiment with software rigged by VW that ensured a fraudulent result, even if Lovelace was kept in the dark about the emissions control deactivation in actual driving conditions. By this point in what had become the long-running VW scandal, it might have been hard to drive the company’s reputation further down, but this story did so. VW apologized profusely: “Volkswagen Group explicitly distances itself from all forms of animal cruelty. Animal testing contradicts our own ethical standards.” But VW wasn’t blaming its executives, of course. “We ask forgiveness for this bad behavior and for the poor judgment of some individuals.”⁴¹

Which individuals did Volkswagen mean to call out? Initially it was some junior employees at VW and Daimler, who were suspended. This maneuver comfortably shifted the blame from upper management, including those who had attended the regular EUGT meetings where the study was discussed. Subsequently, CEO Muller announced that Thomas Steg, VW’s head of foreign relations and sustainability, had known about the study since May 2013 and will “assume full responsibility” for the scandal.⁴² Steg was suspended from his position but was fully exonerated by a VW internal audit and returned to his post less than six months later. As the U.S. and German investigations proceeded, though, more details have emerged, contradicting VW’s claims attributing the misbehavior to lower level employees. According to the U.S. Securities and Exchange Commission (SEC), then–Volkswagen CEO Martin Winterkorn and other VW executives first learned about the defeat device in 2007, before the first device was even installed. The SEC alleges they were warned that selling vehicles with these devices would be problematic for the car manufacturer if their subterfuge was discovered, but those concerns were ignored.⁴³

James Liang, the engineer who wrote the original defeat device software and made sure it was working perfectly in New Mexico, was the first VW employee convicted on criminal charges in Dieselgate. In August 2017, he pled guilty for his role in the conspiracy to defraud both U.S. regulators and VW customers. The judge presiding over the case felt the extent of the fraud was so significant that, even though Liang cooperated with prosecutors to help in their case against Volkswagen (which paid $4.3 billion in civil and criminal penalties in the United States alone) and its executives, he sentenced Liang to 40 months in prison and imposed a fine of $200,000.⁴⁴ He and Schmidt will be deported to Germany when they finish their prison time.

The biggest indictment in Dieselgate came in May 2018: former CEO Winterkorn, for fraud and conspiracy relating to his role in deceiving U.S. regulators. He remains a free man in Germany.

Dieselgate has had repercussions. Claims raised since by industry scientists are now looked upon with much more skepticism. European cities, eager to rid their air of dangerous chemicals and furious at the subterfuge advanced by the auto manufacturers, are implementing or expanding the Low Emission Zones that three EUGT scientists had somewhat successfully discredited. The German city of Hamburg has now banned older diesel vehicles from much of its downtown. London, which has long had a Low Emission Zone, has announced an “Ultra Low Emission Zone” in certain urban centers.

On the other hand, the scandal did not have a huge impact on Volkswagen’s overall sales. Yes, it lost its one-year reign as the world’s leading automaker by volume, but it is still solidly second place. I should also note that the year VW did top the list, 2016, was after the scandal broke.

VW may return to its old spot as the world’s leading automaker, but the sales should be only for gas and hybrid and electric cars. It is time to recognize that diesel-powered passenger cars have little future. New gas-powered engines with catalytic converters put out a small fraction of the NOx coming from even the new-technology diesel engines. The diesels simply can’t be both clean and high-mileage, especially in comparison with electric and hybrid vehicles, which are getting better and cheaper. When European authorities stop subsidizing the price of diesel at the pump, there will be no reason left to purchase a diesel-powered automobile. In the United States, there’s not one now.

Have I saved the best for last? Maybe. In May 2018, the same month former VW CEO Winterkorn was indicted in the United States, about three years after the dominoes started falling in Dieselgate, VW subsidiary Audi was pulled over by German regulators on suspicion of employing a new defeat device in some of its high-end diesel models sold in Europe, and Rupert Stadler, its CEO, was arrested the following month in connection with the earlier Volkswagen scandal. Audi accepted corporate responsibility for the illegal software and paid a fine of €800 million, or about $930 million, in civil penalties. The prosecution of the now former Audi CEO is still under way.⁴⁵