• Workers inspect a car on the production line in a Volkswagen factory in Poznan, Poland.

    Workers inspect a car on the production line in a Volkswagen factory in Poznan, Poland.

    Full Screen

Study: Volkswagen’s emissions cheat to cause 60 premature deaths in U.S.

Workers inspect a car on the production line in a Volkswagen factory in Poznan, Poland.

Timely vehicle recall by German automaker would avoid some 130 early deaths, researchers say.

Press Contact

Kimberly Allen
Email: allenkc@mit.edu
Phone: 617-253-2702
MIT News Office

Media Resources

1 images for download

Access Media

Media can only be downloaded from the desktop version of this website.

Volkswagen’s use of software to evade emissions standards in more than 482,000 diesel vehicles sold in the U.S. will directly contribute to 60 premature deaths across the country, a new MIT-led study finds.

In September, the Environmental Protection Agency discovered that the German automaker had developed and installed “defeat devices” (actually software) in light-duty diesel vehicles sold between 2008 and 2015. This software was designed to sense when a car was undergoing an emissions test, and only then engage the vehicle’s full emissions-control system, which would otherwise be disabled under normal driving conditions — a cheat that allows the vehicles to emit 40 times more emissions than permitted by the Clean Air Act.

That amount of excess pollution, multiplied by the number of affected vehicles sold in the U.S. and extrapolated over population distributions and health risk factors across the country, will have significant effects on public health, the study finds.

Assessing health outcomes

According to the study, conducted by researchers at MIT and Harvard University and published in the journal Environmental Research Letters, excess emissions from Volkswagen’s defeat devices will cause around 60 people in the U.S. to die 10 to 20 years prematurely. If the automaker recalls every affected vehicle by the end of 2016, more than 130 additional early deaths may be avoided. If, however, Volkswagen does not order a recall in the U.S., the excess emissions, compounding in the future, will cause 140 people to die early.

In addition to the increase in premature deaths, the researchers estimate that Volkswagen’s excess emissions will contribute directly to 31 cases of chronic bronchitis and 34 hospital admissions involving respiratory and cardiac conditions. They calculate that individuals will experience about 120,000 minor restricted activity days, including work absences, and about 210,000 lower-respiratory symptom days.

In total, Volkswagen’s excess emissions will generate $450 million in health expenses and other social costs, the study projects. But a total vehicle recall by the end of 2016 may save up to $840 million in further health and social costs.

Steven Barrett, the lead author of the paper and an associate professor of aeronautics and astronautics at MIT, says the new data may help regulatory officials better estimate the effects of Volkswagen’s actions.

“It seemed to be an important issue in which we could bring to bear impartial information to help quantify the human implications of the Volkswagen emissions issue,” Barrett says. “The main motivation is to inform the public and inform the developing regulatory situation.”

Cheating (and) death

To estimate the health effects of Volkswagen’s excess emissions, Barrett and his colleagues at MIT and Harvard based their calculations on measurements by researchers at West Virginia University, who found that the vehicles produced up to 40 times the emissions allowed by law. They then calculated the average amount that each vehicle would be driven over its lifetime, and combined these results with sales data between 2008 and 2015 to estimate of the total excess emissions during this period.

The group then calculated the resulting emissions under three scenarios: the current scenario, in which 482,000 vehicles have already emitted excess emissions into the atmosphere; a scenario in which Volkswagen recalls every affected vehicle by the end of 2016; and a future in which there is no recall, and every affected vehicle remains on the road, continuing to emit excess pollution over the course of its lifetime.

The group then estimated the health effects under each emissions scenario, using a method they developed to map emissions estimates to public exposure to fine particulate matter and ozone. Diesel vehicles emit nitrogen oxides, which react in the atmosphere to form fine particulate matter and ozone. Barrett’s approach essentially maps emissions estimates to population health risk, accounting for atmospheric transport and chemistry of the pollutants.

“We all have risk factors in our lives, and [excess emissions] is another small risk factor,” Barrett explains. “If you take into account the additional risk due to the excess Volkswagen emissions, then roughly 60 people have died or will die early, and on average, a decade or more early.”

Barrett says that, per kilometer driven, this number is about 20 percent of the number of deaths caused by road transport accidents.

“So it’s about the same order of magnitude, just from these excess emissions,” Barrett says. “If nothing’s done, these excess emissions will cause around another 140 deaths. However, two-thirds of the total deaths could be avoided if the recalls could be done quickly, in the course of the next year.”

Daniel Kammen, the editor-in-chief of Environmental Research Letters and a professor of energy at the University of California at Berkeley, says the group’s study provides a “rigorous evaluation of the scale of the impacts, which are potentially exceedingly serious.

“The analysis demonstrates the value of policy-inspired fundamental research where the air quality and health impacts of transgressions such as the VW issue can be calculated, and made available for public discussion,” says Kammen, who did not contribute to the research.

Topics: Automobiles, Emissions, Pollution, Chemistry, Data, Earth and atmospheric sciences, Environment, Health, Industry, Research, Technology and society, School of Engineering, Aeronautics and Astronautics, Greenhouse gases, Oil and gas


I'd like to see this compared to how many deaths and health problems can be attributed to other vehicles on the road and their emissions.

That's a very interesting thesis. As the owner of an offending vehicle, I accept that I may rationalize my objections to the basic tenets.
If we are interested as a society in saving lives related to NOx emmisions, then let I eagerly await the follow-up article addressing the 80% problem, diesel trucks.
I'm sure the EPA standard for trucks (including the blurred line between truck and SUV) is more than the 0.04 gm/mi standard applied to the EPA definition for a "passenger" vehicle.

sqwirk - this study only covers the additional damage caused by the defeat device. Yes, it's a tiny percentage compared to overall gas- or diesel-related deaths, but it's still additional deaths caused by cheating that otherwise would not have happened. No matter what, it's unacceptable.

This is a crap study which is not scientific because there is no estimate of the accuracy of the findings. Also, diesel tractor trailers present on the road today are not regulated and likely put out so much NOX that the presence of the TDI NOX would make an immeasurable difference in the health of individuals.

60 premature deaths? My own studies show 47 people, 32 birds, a cat, 17 honey bees, and a formica tree. Yeah a Formica Tree!!!

Anyone, who uses an oil furnace for heating his/her house, its emissions are pouring out much more pollution in the air than these VW diesel vehicles. In fact for every gallon of heating oil burned in a home oil furnace, the pollution is far greater than a gallon of diesel fuel burned in a car. There aren't catalytic converters or particulate trap filters or any other pollution control devices on home oil furnaces either.

When a home oil furnace is burning oil to heat a home, its emissions spews out far more pollutants per gallon in the air than a gallon of oil burned in a VW diesel vehicle on the road. There are no catalytic converters or particulate trap filters or any other pollution control devices on home oil furnaces. In fact it is known that living by an interstate highway/high traffic areas shortens an individual's life span too.

Don't forget, there is a link between air travel and deaths on the ground, as noted on the article located on the upper right side of this particular MIT News web page--entitled "A Link Between Air Travel and Deaths on the Ground."

I have a 2011 Volkswagen Jetta TDI which I bought brand new and since buying this car I have come down with Breast Cancer and problems with my lungs...I am wondering if this car has anything to do with me getting breast cancer. According to my doctors genetics did not play a factor in my breast cancer....

The EPA Renewable Fuels Standard causes up to 245 premature deaths per year due to increased particulate and ozone emissions. This is the result of EPA's own analysis backed up by findings of the National Academy of Sciences and is included in the fine print of the 2010 RFS2 final rule (p. 7 of http://www.epa.gov/otaq/renewa... ). Whatever penalty is applied to Volkswagen should be multiplied by a factor of four and applied to EPA leadership.

In order to value, how dangerous the additional NOx-emissions by VW-diesel passenger cars with defeat device are, I'd like to ask the makers of this study, to calculate how many deaths are caused within one year by the NOx-emissions of all diesel trucks running in the United States. Thank you!

I first came across this story on a link from Google News to a web site where it was very obvious that English was not the first language of the writers and editors, so I had to check it out from the source, MIT. I thought the figure of "60" might be 60,000 or more, as it seemed like a typo.

60 premature deaths? Every life is important, yes, but billions of dollars spent recalling and repairing VWs to prevent 60 deaths?? I agree with other commenters - I'm sure there are thousands more illnesses and deaths that could be prevented if the awful diesel exhaust clouds from trucks were cleaned up. While in my 2015 VW Beetle, I've been driving behind diesel pickup trucks and the semi tractor-trailers that are obviously emitting much more pollution than my car is, and this angers me. My car is considered a polluter and they go on their merry way, leaving me coughing and gasping in their clouds of black smoke.

are you people insane? with this logic how many people is boeing killing with air craft emissions? Is every one who drives a car pre 2008 epa laws liable for air quality? If you cut your grass are you killing people? What about heating your home? heating oil is burned with no emissions. What does MIT use for heating?

I'll get back on my old soap box. We should be getting heavy trucks off the road by giving the railroads aid to improve their rail system so they can deliver long haul loads faster and cheaper than long haul trucks. This would do much more to reduce premature deaths than VW spending billions to correct their emissions problems.

The comments that preceded this one make a valid point. Yes, there is reason to believe that some people will die prematurely. But what is objected to is that the analysis is utterly devoid of any sort of relative ranking. It's something like saying smokers should not eat buttered toast, because studies show that butter can cause premature death. Yes, but compared to smoking?

So take the data from this study and calculate the "cost" of one life-year. Then compare it to other costs, and see how it ranks.

Richard Iredale BSEE '77

How did the researchers get from tiny increases in ozone levels to number of deaths? Did they perhaps use the highly suspect dose linear assumption? This is the assumption that if a lot of something is lethal then a little distributed over a lot of people is just as toxic per pound of exposure. Apply this to table salt and you see the problem. Humans have spent 10,000 generations around smoky fires and probably have defense mechanisms to cope with low particulate levels.

Since minimizing NOx emissions causes loss of performance and reduced fuel economy, how many automotive traffic deaths will result from the newly underperforming cars? How many deaths will result from the additional CO2 and global warming? Might there be more actual deaths than the projected NOx deaths? Is this like the EPA mercury standards to save 0.001 IQ points in the offspring of women who eat 500 lbs of lake fish during their pregnancy?

What complete and utter bunk. I am so disappointed that MIT puts their name on this pseudo-science. The VW kerfuffle needs to be put into context.

All the emissions of the entire VW fleet are not 1% of the emissions of the school busses around the country. Shall we end school bussing to avoid premature deaths? They are not 1% of the emissions of the trucking industry--should we ban trucks now?

This is just plain goofy.

The excess deaths they reference are premature deaths, probably of sensitive population receptors with COPD or some other chronic, terminal ailment. I have read the paper and it appears that they did not prorate the result to reflect the predicted amount of life lost. For example, if the premature deaths are lives that are curtailed by an average of, say 1 year, the result should be multiplied by approximately 1/80 yielding less than one equivalent life lost. I have made various comments on this to MIT, the principal author, and numerous media outlets. No reply. Also, the quick publication turnaround and apparent lack of peer review are suspicious. Apparently they used standard risk assessment and exposure modeling methodology, but omitted the proration as I have discussed here. I wonder if the lead researcher, an MIT engineering lecturer, is truly qualified to conduct this study - certainly he is not a toxicologist or health physics expert.

Back to the top