Capitalisn’t: The sticky side of Teflon

Jan 31, 2019

Sections Economics

On this episode of the Capitalisn’t podcast, hosts Kate Waldock and Luigi Zingales discuss the health hazards and economic costs of air pollution and contaminated drinking water from the toxic chemical PFOA (C8), which has been used in the manufacture of Teflon. How did DuPont skirt regulation and avoid corporate responsibility for so long?

Luigi: Hi, this is Luigi Zingales at the University of Chicago.

Kate: This is Kate Waldock from Georgetown University.

Luigi: And this is Capitalisn’t.

Kate: A podcast about what’s working in capitalism today.

Luigi: And, most importantly, what isn’t.

Speaking of what is not working, just before Christmas, I went to a conference in New Delhi, and when I landed, the air was unbreathable. I had been there 20 years ago, and it was bad, but this time, it was really, really bad, so I started to look at some numbers. One way you can measure pollution is by how much small particulates are in the air, and, generally, the healthy stuff is between one and 100. In California, when there were the fires, the big fires, it reached 200. That day I was in New Delhi, it was 450. I had to buy a mask just to survive.

Kate: Yeah, I mean, that’s insane, but I think another insane thing about that story is that, as an American, when I think about traveling to other countries, I’m automatically wary. I think, “Oh, I need to get a mask,” and, “Oh, I need to get water tablet pills,” to make sure that I don’t die of malaria or something, but I take for granted that here in the US, everything is clean. We don’t have those problems. Those are just foreign problems, when, in actuality, that’s not really the case.

Luigi: To be fair, our problems are much, much lighter than the one in India, but they do exist and not only with regards to the PMI index, which is this particulate matter, but other things we’re going to discuss in this episode.

Kate: Just as a disclaimer, when we talk about pollution or when you hear the word “pollution,” you might think that’s really a topic that concerns climate change and global warming. On this episode, we’re not going to talk about that sort of pollution. We’re going to focus on environmental health hazards like contaminants of your groundwater and your air quality. At a later time, we’d like to fully flesh out issues relating to climate change and capitalism, but we hope to discuss those issues in the future. 

Luigi: Let’s start with one of the leading causes of death in the United States and in the world: small particles that are brought into the air by burning fossil fuels or by burning crops. They’re so small that they enter our lungs and they cause a number of respiratory diseases, but, most importantly, they are a primary cause of cancer.

Kate: Now, one of the challenges for economists and scientists is to try to figure out how bad these particles actually are. Of course, a natural thing to do would be to collect data about areas where there are lots of these particles, and then you collect data about deaths in those areas, or respiratory diseases or cancer, and then you come up with some sort of correlation between the two. People have done that, and there seem to be pretty high correlations between respiratory disease and particulate matter. However, that still doesn’t really address the issue of whether the particulate matter is causing the disease or the death, because there’s so much else going on in the environment that we don’t know or we can’t measure.

Luigi: Recently, as an economist, I resorted to a technique with a fancy name, but with a very simple idea. It’s called regression discontinuity, and let me try to explain the ideas in simple words. The example we’re going to use was used by a colleague to estimate how much the increase in this particulate matter increased death.

In medical science, if you want to see what generally cures a disease, you do what is called a controlled experiment. You have a treatment sample and you have an untreated sample. You make sure that the two samples look identical, and then the difference in results between the two gives you the impact, for example, of a new drug. Especially when it comes to pollution, you are not allowed, thank God, to randomly treat people and say, “To you, I give you a lot of pollution, and, to you, I don’t,” and see what happens. However, there are some policies that come pretty close to doing that. 

In early communist China, Mao Zedong decided that everybody who lived north of the Wei River needed to be heated, and so they gave away coal to heat their houses. But if you were just south of the Wei River, it was warmer and you didn’t need to be heated, so the Communist Party would not give you any coal to heat. The people just north of that river were treated with a massive amount of pollution from coal, because everybody was heating their houses with coal, and coal is actually the biggest source of this particulate matter. As a result, data show that people living north of the river had a 46 percent higher level of this particulate matter in the air than people living south of it, and life expectancy was actually a year shorter just north of the river versus just south of the river.

Kate: I appreciate studies like this. I think that they are clever and I think that they help establish causality in environments where it’s really difficult to do so. I have a general criticism with this approach and the way of thinking of a lot of economists. I don’t think that they appreciate correlations enough, and I think that they put too much emphasis on these types of studies. If we have overwhelming evidence based on correlations, that should be enough for us to really get a good sense of how much this particulate matter coming from coal can impact life expectancy. OK?

Luigi: Actually, I recently discovered that the Nazis discovered that smoking kills you with a correlational study in 1939. As a result, Hitler prohibited everybody around him to smoke cigarettes, and he was actually proud that the fascist leaders were not smoking, Hitler, Mussolini and Franco, while the allies were all smoking, from Roosevelt to Churchill to Stalin. 

Kate: Wow. That’s a super interesting fact. 

Luigi: After the war, all this research was thrown away with all the Nazi research, and it took more than 10 years for the Americans to figure out that, actually, smoking kills you.

Kate: Hitler’s revenge. 

Luigi: Anyway, this is to say that, you are right, sometimes correlation studies can be very useful in identifying a problem before you do these random treatments. But I think that these are useful in trying to get closer to the standard of causality that other sciences use, and, in economics, it’s more difficult to use it.

Kate: Yeah.

Luigi: Now, what is interesting is that Michael Greenstone, who is one of the authors of this study and a colleague here at Chicago, went farther than that. He actually used satellite data to measure the level of this PMI index in a lot of cities around the world. As a result, I can tell you, if you live in Los Angeles, you end up living one year less than the average. If you live in New Delhi, you end up living 10 years less than your life expectancy.

Kate: That’s crazy.

Luigi: It’s a pretty dangerous proposition to live there.

Kate: Yeah.

Luigi: The other thing which is quite interesting is that, until 2008, the worst place was probably Beijing. Now, Beijing has made a very concerted effort in trying to improve pollution. I think one aspect that is often ignored is that these studies are very useful in focusing people’s attention, because if you just have an idea, “Oh, yeah, pollution is bad,” that is one thing. If I tell you that you live 10 years less because of pollution, the government would pay attention, so I think that it’s certainly an important step in trying to create the demand for change.

Kate: Aside from just the underemphasis, I think, on pure correlational studies, another bone to pick that I have with this experimental design is that you have a cutoff, and you’re looking on one side of the cutoff versus the other side. In this case, the cutoff is people living above the river and people below the river. 

In order to estimate the impact of pollution on, say, life expectancy, you want to see some difference in life expectancy around the cutoff. In order to do that, you’re collecting data on either side of the river, and what economists and data scientists often do is that they have to fit some function to either side of the cutoff. I have issues with the incentives that are in place in fitting those functions, because the incentive is to choose a function that will maximize the difference in outcomes. 

You really want to say that there’s a big impact of pollution on life expectancy, and so you always see these graphs that make it look like the authors have absolutely tried to pick the function that makes it look like there’s the biggest jump in outcomes on life expectancy when, really, maybe there was a jump, but it wasn’t that huge. I would say that there’s some of that going on in this paper. I’m not sure that it really looks like there is a full year of difference here, but there is a difference, so I will grant them that.

Luigi: Also, I think, Kate, you’re right that there are these incentives. However, there’s a tension between saying nothing precisely or precisely nothing, and I think there’s a lot of value in having some estimates that might not be perfect. But if you just see an index that goes to 450, you can’t really relate it to your life. If they tell you that you’re losing 10 years of your life and you could only lose one if you move to LA, that makes a big difference, and maybe you’re losing between nine years and nine months, but it’s still a huge difference.

Kate: Maybe the air quality in the US isn’t quite as bad as it is in New Delhi, but there are still cities in the US where particulate matter is a problem—particularly in LA, for example—and one of the issues is that we’re not really moving in the right direction. The EPA under Trump is starting to roll back some of the measures that are designed to protect people and the health of their lungs rather than strengthening them, and I think part of the reason for that is because of the people we see leading the EPA. 

You would expect someone who is head of the Environmental Protection Agency to actually be concerned with protecting the environment, but, in fact, we had . . . Formerly, there was Scott Pruitt, who was on the record saying that he wasn’t convinced that carbon dioxide was necessarily a contributor to global warming, and now we have Andrew Wheeler, acting head of the EPA, who has an established record as a lobbyist for big coal.

Luigi: This is very important because, of course, companies could use more efficiently and at a lower cost if they don’t have to worry about the externality they generate through pollution. Regulation does increase costs, but also, regulation has an impact on human lives. 

The function of the EPA is to put rules in place so that companies can focus on producing, and know what they can and cannot do in terms of polluting the environment. The problem is that this regulation does not seem to be working very well, and Kate is absolutely right. One of the problems why this regulation does not work is a very intense revolving-door system. If you look at all the EPA administrators since the early ‘80s, most of them either became board members of major polluters, from chemical companies like Monsanto or DuPont to oil companies like Conoco. Others started consulting on environmental business, probably having these people as clients, if not going directly to work for those companies. I think that the revolving door is massive at the top, but it is massive even at the intermediate level.

The principal deputy on the Office of Chemical Safety and Pollution is a certain Nancy Beck, who for many years had been an executive of the American Chemistry Council, so she was basically a lobbyist for the chemical industry to reduce pollution, and she has been put in charge of the henhouse. 

It’s not just a problem of incentives. As an economist, of course, I think incentives are very important, but it’s also a question of your way of thinking and, basically, a lifetime of research. If you are somebody who worked for the coal industry all his life, you probably think that coal is the greatest thing since sliced bread and that all the environmental issues are second-order.

Kate: I think we should talk about the DuPont case, because it’s a pretty good illustration of what’s broken in terms of environmental pollution and trying to protect citizens against it. For those of you who are not familiar, DuPont is a chemical company. By the way, it merged with Dow Chemical in 2017 to create DowDuPont, and it’s now the biggest chemical company in the world. One of its flagship products is Teflon, which you’re probably familiar with in terms of the liner of your pots and pans. It’s a nonstick product. DuPont has actually been making Teflon since the ‘40s, and one of the chemicals involved in making Teflon is called PFOA, or C8. It used this product in the manufacturing of Teflon for over five decades.

Luigi: For the record, the division that was producing Teflon has now been spun off by DuPont, and its name is Chemours, and it is separately traded on the NYSE, so all the information we’re describing today does not regard DuPont directly. However, what makes this case very interesting is that litigation brought up a vast number of internal documents that allowed people to make a calculation of what decisions were made back in 1984, when DuPont became aware of some problems, and, I will say in a second, pretty big problems of C8. It decided not to stop production, not to introduce a form of control or reduction of pollution, but actually to double production.

By the early ‘80s, DuPont either knew directly or was made aware by 3M, which was producing the C8, that C8 was a toxic product that needed to be handled with care. It also knew that, in mice, the exposure to C8 could lead to birth defects. After receiving this study, DuPont went and looked at all the women that were working in the plant that was producing Teflon. Of the seven women who had been pregnant, two had children with birth defects, and the same kind of birth defect identified in the mice, so much so that they decided to remove all women from the production of Teflon. But then, they also sent some employees to sample the water supply in the area in West Virginia where the plant was, and they found abnormal levels of PFOA in the water supply. 

In spite of all this, they didn’t do anything. They didn’t report to the EPA. They didn’t do anything. They decided to double production of PFOA. Now, what’s interesting is, several years later, a farmer who happened to have a farm next to the plant started to see his cows die. They died with green eyes and very strange behavior. He immediately thought that the cause was a landfill by DuPont that was nearby. He tried to get some compensation from DuPont, and DuPont said, “Go away. You don’t know how to raise your cows.” He got lucky because he remembered that, when he was a kid, there was a guy playing with him that later became a lawyer in a major law firm in Cincinnati, so he reached out to him. 

This lawyer, whose name is Robert Bilott, in spite of the fact that he was a top corporate lawyer, decided to take this plaintiff case just out of friendship. He understood from the internal documents of DuPont what PFOA was and started a major case. This case started in the late 1990s and the last settlement took place in 2017.

Kate: I also think it’s worth mentioning that this C8 didn’t just happen to get into the water supply. It wasn’t just an unfortunate accident. DuPont was just dumping the C8 into the Ohio River through tubes. They didn’t care. They were just like, “Let’s get rid of it. We don’t even want to bother securing it. We’re just going to dump into the river.” They were also creating these landfills that were unmarked, so they were intentionally trying to keep them secret. They were just burying the C8 and they were also emitting it through their smokestacks, so it was basically getting into the environment in the worst ways imaginable. 

Luigi: To be fair, DuPont is not the only producer of C8 and is not the only one that polluted America, because I’m sorry to report that 99.7 percent of Americans have C8 in their blood. It also came from the 3M Scotchgard, from a lot of flame retardants, and it’s unfortunately used a lot by the military as anti-fire stuff, so I think that pollution is all over the place at levels that are quite dangerous.

Now, if it wasn’t for Robert Bilott, we’d never have discovered the effects of C8. Because C8 is a substance that is called biopersistent, which means that it’s not disintegrating in the environment and it accumulates in your body. Nobody really knew what that accumulation led to. Bilott had the brilliant idea to ask for an independent study. The problem with this independent study is where you get the data. 

Bilott used some of the settlement money in one of the earlier cases to pay people in the affected area to donate blood. They went near Christmas with some trucks, asking people to donate their blood in exchange for $400. As you can imagine, $400, especially in West Virginia, especially next to Christmas, goes a long way. Out of the 70,000 people that were affected, he was able to collect 68,000 blood samples.

I don’t know whether you remember the movie Erin Brockovich, but, in that movie, there is a similar situation in which the main character is able to collect an enormous amount of data. We now know PFOA is responsible for an increase in the numbers of testicular cancer, kidney cancer, hypertension and another set of less lethal diseases.

Kate: Backing up a minute, I think it’s important to go through the thought process of the Tennant family. If you just realized that all your cows have died, and you’re trying to figure out what to do, what remedies do they have? Here, I think, there’s numerous ways in which the system just broke down for them. The way that they ended up getting DuPont is that they won based on a negligence suit. These types of lawsuits are in the realm of what’s called toxic torts. They were suing DuPont for being negligent, but, in order to do that, you have to prove a couple of things. First, you have to show the defendant was negligent and that they caused them harm. 

In this case, it was up to the Tennant family to show that DuPont had caused the Tennant family’s cows to die. Now, what sort of information did they have at the time? Absolutely nothing. It’s, in fact, really hard to show proof. You have to know exactly what the chemical was that caused the harm and you have to be able to prove that chemical was directly responsible for the harm, and from the point of view of the Tennants back then, it was virtually impossible.

Luigi: Just to give you a sense, there are now roughly 85,000 substances, chemical substances, that have been used commercially, and every year, 1,500 are added. Unlike new drugs that go through a very lengthy process to establish that they are safe to use, basically, in the United States, new substances are considered safe for use unless proven otherwise. The burden of the proof is on the plaintiff or the people who are damaged or the EPA, who, as we said, is fairly cozy with the largest producers.

Kate: On top of that, there was another element to this case concerning the EPA, which was that, back in 1976, the EPA passed the Toxic Substances Control Act—which you might think, based on the title, would have allowed the EPA to regulate this sort of substance. The problem was, if materials or chemicals existed prior to 1976 when the act was passed, they were grandfathered in, as long as they weren’t already proven harmful at that point, which C8 wasn’t, because we just didn’t know enough about C8 and its health effects. And DuPont had no incentive to try to establish them, because, otherwise, it would have opened the chemical up to not being grandfathered in. The EPA at this point didn’t even really have the authority to try to test whether this was, in fact, a harmful substance.

Luigi: Recently, this regulation called TSCA has been revised. In fact, in a rare bipartisan move in 2016, a new TSCA law was passed in which, on the one hand, there would be more intervention by the EPA in terms of testing. On the other hand, there would more grandfathering and tolerance for the existing substances. Unfortunately, after the industry got in this nice spot, the administration changed, and the new EPA seems not so eager to do the tests they were supposed to do. So, while the regulation was meant to be at least a partial improvement, it might end up being a step back.

Kate: Another reason why this case, the Tennant case, was great that it won, but not the best example of an everyday pollution toxic tort case, was that the lawyer that you mentioned earlier, Luigi, Robert Bilott, wasn’t exactly your typical lawyer. In most cases like this, if you have enough people who have been harmed by environmental pollution, they would get together and form what’s called a class-action lawsuit to sue the company that was responsible for the damages. But the thing is that class-action lawsuits are expensive, and they take a long time, and they’re risky, and a lot of the financing for class-action lawsuits comes about through debt. That makes it such that the lawyers who are responsible for pursuing these types of cases tend to be risk-averse. They don’t necessarily want to go after cases where there is not much information and where the company has an established record of trying to hide that information. So, if this had been pretty much any other family without those sorts of connections, it would have been very unlikely that a class-action suit would have even been launched against DuPont.

Luigi: What is scary here is that DuPont is not your typical fly-by-night company that does pollute the environment and run away, et cetera. DuPont is probably the most blue-blooded corporate entity in America. It’s one of the oldest companies still in business and is praised for the way they’ve treated the environment, the research they have done on environmental risk for their employees. Even one of the CEOs during all this debacle, Charles Holliday, wrote a book called Walk the Talkdescribing how you can be a chemical company and be very good for the environment. Ironically, all these people, from the CEO to the directors to the people who made the decision in 1984, are either dead or retired, and they’re paying no reputational cost for what they have done.

Kate: Yeah. I think another interesting element of the reputational cost story is that DuPont is a chemical manufacturer. It’s not like I go to CVS and I buy DuPont goods all the time. When you think about reputational damage, for example, like Wells Fargo and their fake-accounts scandal, as a consumer, I might have the direct ability to change my bank, or I might have the direct ability to stop buying that good. But for a company like this, where what it produces is often an intermediary in the production of another good, the consumers don’t really have that much power to stop buying goods, because they’re not direct.

Luigi: Normally, in corporate finance, we think that what keeps companies at bay is a combination of legal liability, regulation, and reputation. Kate just explained to you why reputation does not seem to work. We discussed why regulation is not very effective, in part because of this revolving-door policy. Even legal liability does not seem to be that great.

In a study I’ve done of this DuPont case, we looked at . . . Imagine the company had known what the liability would be in the future. What was the probability of being caught that would make it optimal from a shareholder’s point of view to actually pollute? What we find is this probability is quite high, it’s 20 percent. So, if the risk of being caught is less than 20 percent, it was optimal for the company to pollute. 

Kate: Another point related to this problem of information disclosure is that part of what allows them to protect this information is this idea that the chemicals that they’re manufacturing are trade secrets. If they were to release information to the EPA or to government regulators about even the name of the chemical, let alone its molecular compound, then that would jeopardize its competitive advantage. They wouldn’t be able to make profits from making that chemical anymore or using the chemical in its Teflon manufacturing process. And that, therefore, would disincentivize them from doing research and development. 

I’m so sick of hearing this trade-secret story used to justify all of these corporate misdeeds. I mean, whether it’s from employment contracts and restricting labor mobility to just outright pollution, it’s ridiculous that this guise of trade secrets should allow us to protect corporations from revealing any information whatsoever.

Luigi: From an economic point of view, it’s important to understand that the optimal amount of pollution is not zero. Reducing pollution has a cost. We don’t want our house to be super safe at the cost of having no windows, because a house without windows would be safer, but it’s not a house we want to live in. In the same way, we don’t want to reduce pollution to zero, because the cost is prohibitive. However, it’s important to figure out what are the costs and benefits of pollution. 

Kate: I’m glad you raised that point, because, I agree, it would be ridiculous to try to make pollution absolutely zero. But, at the end of the day, you can’t do a cost-benefit analysis if you don’t know what the pollutants are, and you don’t know what their costs are. I think that the real problem here isn’t necessarily the pollution itself, even though, obviously, excessive pollution can be a problem. It’s the fact that our institutions aren’t incentivizing these companies or aren’t requiring them to disclose enough information for us to be able to make that assessment.

Luigi: Absolutely. I think that, in an ideal world, the incentives provided by litigation, regulation, and reputation should make companies behave as if they were doing the cost-benefit analysis from a societal perspective. We need to think about how to realign these incentives. Unfortunately, this is not specific just to the particulate matter in the air or DuPont PFOA. It’s a more general problem, and to see another case of this in the next episode, we are going to interview Carey Gillam, an investigative journalist who has written a book, Whitewash, about Monsanto and a pesticide that is widely used around the world, Roundup. We’re going to discuss with her the possible solutions to these problems.

Kate: Before we go, I just wanted to insert a quick thank you to all of our listeners who responded to our feedback survey, some of whom got Capitalisn’t T-shirts. It really meant a lot to us, and it really helped us shape the direction of the show, so, again, thank you.