Freakonomics Radio - 472. This Is Your Brain on Pollution

Episode Date: August 12, 2021

Air pollution is estimated to cause 7 million deaths a year and cost the global economy nearly $3 trillion. But is the true cost even higher? Stephen Dubner explores the links between pollution and cognitive function, and enlists two fellow Freakonomics Radio Network hosts in a homegrown experiment.

It's worse than cigarette smoking. It's worse than wars. It's worse than auto accidents. Wow. What's worse than wars and car crashes and smoking? Here, I'll give you a hint. Imagine you were getting ready to leave your house for work or school, maybe to go for a run.

There is some standard information that most of us seek out before leaving home. There's this. We've got partly sunny skies. It's 85. South winds at 14. And there's this. We've got multiple accidents, stalled vehicles causing major delays. It makes sense to check the weather and traffic before leaving home, but there's information we don't usually check that could be just as important, if not more so. What if this is what you heard in the morning?

The level of particulate matter in the air today is above the recommended World Health Organization guidelines. Or even this. If your child has an important test today, or you're giving a big presentation at work, you might want to consider rescheduling. Or even this. The Supreme Court will be delaying oral arguments until next week because of a high particulate matter count in Washington, D.C. It's well established that air pollution has significant negative effects on the human body. And many places do require a public announcement when pollution levels are high.

But is it possible that on a given day, high pollution can affect your brain, your cognitive abilities? So I can't say I've heard many more theories. That would surprise me more if they were true. Today on Freakonomics Radio, this is your brain. The top card is written in black. The top card is written in black. The bottom card is written in blue. So I'm going to say yes.

And this is your brain on pollution. It's one of the hardest things I've ever done in my life. My God. Is pollution making us more stupider? This is Freakonomics Radio, the podcast that explores the hidden side of everything. Here's your host, Stephen Dubner. Whoa, whoa, whoa. Ooh, ooh. Here's your host, Stephen Dubner. Andrea Linneaus is an economist at the University of Queensland in Australia.

I'm an environmental economist, which means that I use data and the tools of economics to understand the causes of environmental problems, and to think about policy solutions. Air pollution is, of course, a long-standing environmental problem. Chemicals like ozone and carbon monoxide, and also what's called particulate matter, or PM2.5. Tiny little particles in the air that are a diameter that is less than 2.5 micrometers. So it's more than 100 times thinner than a human hair. And those particles can come from natural sources like dust and smoke, but also from things like the combustion of fossil fuels.

So that's one of the most concerning forms of air pollution. Particulate matter can be invisible. So unless it's really bad, you can't just tell by looking at the sky whether the air you're breathing is polluted. But the odds are that it is. The World Health Organization estimates that roughly 90% of people around the world breathe polluted air. The WHO has different guidelines for different pollutants. For particulate matter, anything above 10 micrograms per cubic meter on average over a year is considered polluted.

The average across China is 35 micrograms. The average across the U.S. is 12, still above the WHO threshold, but much better than it was just a few decades ago. Accordingly, our concern about pollution has been falling. 30 years ago, more than 60 percent of Americans said they had a great deal of concern about pollution has been falling. 30 years ago, more than 60% of Americans said they had a great deal of concern about air pollution. Today, that number is just 43%. Here's one of those 43%.

I think air pollution is the greatest single threat to human health on the planet. Michael Greenstone is an economist at the University of Chicago, where he directs the Energy Policy Institute and co-directs the Climate Impact Lab. He also spent a year in the Obama White House working on climate policy. One of his creations is called the Air Quality Life Index. The Air Quality Life Index uses satellite data to say how much longer would people in any part

of the world live if their area was brought into compliance for what air pollution should be. So how does air pollution affect life expectancy? The average person on the planet is living 2.2 years less than if where they lived complied with WHO standards. Which is what leads Greenstone to say this. It's worse than cigarette smoking. It's worse than wars. It's worse than cigarette smoking. It's worse than wars.

It's worse than auto accidents. The World Health Organization estimates that roughly 7 million people die every year from exposure to fine particles in polluted air. That's at least double the number of people who died globally from COVID in 2020 and more than five times the number of people killed every year in

car crashes. The more proximate causes of the pollution deaths include pneumonia, stroke, and heart disease. The economic costs of pollution are also massive. One estimate puts it at nearly $3 trillion a year, more than 3% of global GDP. Andrea Linneaus again. The interest in economics is fundamentally about the productivity impacts. And so part of the reason we're interested in cognition is that if cognition affects productivity, then the costs of exposure to air pollution may be much, much larger than we had previously estimated. In other words, we know that air pollution is dangerous to our physical health, as evidenced by millions of deaths around the world. And as Linos tells us... There are a number of papers that all point in the same direction, in that it does appear that

there are cognitive impacts of exposure to high levels of air pollution. So how significant are the cognitive impacts of air pollution? The U.S., remember, is a relatively low pollution country, in part because of domestic policies like the Clean Air Act, but also because we have offshored so much of our manufacturing and the pollution that goes with it. A study published in the journal Nature in 2007 found that more than 50% of China's air pollution at the time was associated

with goods and services consumed outside the provinces where they were produced, and that 11% of Chinese air pollution deaths could be traced to goods and services used in the United States and Western Europe. So the U.S. has had the luxury to worry less about the physiological effects of air pollution, but should we worry more about the cognitive effects? Before we answer that question, let's take a look back at how the U.S. got to where it is. So the Clean Air Act, I think, is one of the most beneficial pieces of legislation that was ever passed. Michael Greenstone again. It was passed in 1970. It was President Nixon who signed a Clean Air Act into law. It was amended several

times, almost always on a bipartisan basis. The Clean Air Act essentially sets limits on the amount of pollution that can be released into the air via manufacturing, transportation, and so on. How effective has it been? Everyone has probably seen pictures of Delhi today, and there were many parts of the United States that looked like that in the late 60s and early 1970s. One of my favorite anecdotes from that period is that white-collar workers in Gary, Indiana, as a regular matter of doing their job, brought a second shirt. And so these high levels of pollution that we're seeing in other parts of the world, they once exist in the United States. And the reason they don't exist in the United States anymore is largely due to the Clean Air Act. There are many benefits of cleaner air, even beyond the obvious. A recent study in the scientific journal Environmental Research Letters found that American crop yields are significantly higher than they were 20 years ago, thanks to fewer pollutants in the air, but that some pollutants, especially particulate matter, are still hurting crop yields. The Central Valley in California remains pretty polluted. There's

parts of the Midwest that remain pretty polluted. But relative to the WHO standard, the United States is very clean. The majority of the problem is concentrated in Asia, especially in India, Bangladesh, China, and in some parts of sub-Saharan Africa. The primary factors that drive pollution in those Asian countries are power generation, home heating, transportation, and as I mentioned earlier, manufacturing, including a lot of manufacturing that used to be done in places like New York and Chicago and Los Angeles. So it's a bit rich for the U.S. to criticize developing countries for their high pollution, especially since most of our environmental regulation came along after we built out our infrastructure and cities. Pollution is a natural byproduct of

civilization building, and there's plenty of historical evidence. Blackened lungs in mummified tissue from Egypt, Peru, and Great Britain point to wood fires from ancient homes. Complaints about air pollution date back at least to ancient Rome, when the smoky cloud hanging over the city was called infamous air and heavy heaven. But air pollution really took off with the invention of the steam engine and the Industrial Revolution, starting in England. The UK today produces less than 3 million tons of coal a year with the goal of getting to zero. At its peak in the early 1900s, they produced nearly 300 million tons a year. The UK burned so much coal that the natural ecosystem adjusted. There's a story of microevolutionary biology, which is about the peppered moth.

I'm not sure if you've ever heard about this. That is Stefan Hiblich, a German economist who teaches at the University of Toronto. As for the peppered moth. The peppered moth appears in the UK in two varieties, a darker and lighter variety. And it's well known that before industrialization in the north of England, the lighter variety was the predominant species.

And this was basically because it could hide on trees from predators. But then as cold smoke started turning trees darker, we see a rise in the instance of this darker version of the peppered moth. So the darker version of the peppered moth was a byproduct of heavy air pollution, kind of like those white shirts worn by office workers in Gary, Indiana. For Hiblich and his fellow researchers, the moth would be a useful indicator in a much larger story about pollution. It's a story that involves geography, poverty, and wind, a westerly wind, to be precise. In cities in the Western Hemisphere, winds blow from the west to the east,

and you might observe that in a lot of these cities, east sides are more deprived. Deprived meaning lower income. There are, of course, exceptions, but the general rule is that the east side of many cities in the Western Hemisphere are poorer than the west side. We started wondering if this was driven by cold smoke during industrialization and a sorting of poor people into the east side and rich people away from the East Side. And we wanted to understand if this has long-lasting effects. If pollution could have an evolutionary effect on the color of a moth species,

could it be that prevailing winds carrying coal smoke could change the demographics of a city? Hiblich and his co-authors Alex True and Janos Zilberberg began to assemble data from 70 cities across England, starting before coal was heavily used as a fuel for industrialization and extending through its heyday. This was not a simple task, and it required a fair amount of creativity. For instance, they hunted down the locations of industrial smokestacks. We started looking into historical maps and found out that Victorian cartographers were absolutely stunning in the level of detail that they drew into their maps. We found the exact location of industrial smokestacks within factory buildings.

These factories were the sites of steel production and other processes that burned massive amounts of coal. We basically found across all these 70 cities. In England, we found about 5,000 chimney locations, like the exact geolocations. They were literally like a historical version of Google Maps. The researchers also incorporated census data, like baptismal records, to get at the economic demographics of the English population.

But what about pollution data? Victorian England may have had brilliant cartographers, but they didn't have monitors to measure particulate matter. This is where the peppered moth comes in handy. Andy. Using the geolocations of the old smokestacks to pinpoint the pollution source, the researchers used an algorithm to model how that coal smoke was carried eastward on the wind, and they confirmed the model's prediction by aligning it with the historical ratio of dark to light peppered moths in a given area, since there were more dark moths in high pollution areas. Clever, yes? Hiblik and his co-authors recently published their findings in a paper called East Side Story, Historical Pollution and Persistent Neighborhood Sorting. What did they find?

So after coal smoke came in, we see a resorting of poor households into the east side. We have data from 1817, which is before coal smoke was a main fuel for industrialization. And we find that in 1817, the wind direction where coal smoke would blow to doesn't have an effect. Meaning that in 1817, before heavy coal use, the east sides of cities were not systematically poorer than the west sides. Then they looked at the data from 1881. They chose that particular year because they had really good data. We had a census where we had all the names and addresses transcribed. And because there was by now a lot of coal being burned.

It's pretty much just before the heyday of industrialization. And what did they see in 1881? In 1881, we see a pronounced pattern where there's a much higher share of low-skilled workers on the east side of the city. What's your best evidence that this relationship is causal and not just a correlational finding? If you draw, let's say, a small circle around a chimney, you would expect in general to have a higher instance of low-skilled workers just because commuting at that time was walking and they have to live somewhere close. But even if you hold distance constant and draw a circle, you would then see that as you walked along the circle, once you get to the east, you will see that the instance of low-skilled workers is in the range of one or two percentage points higher.

Are the low-skilled workers low-skilled because of the coal smoke, or are they living there because they're low-skilled workers? I cannot tell for sure if it's because of the coal smoke. I think in the past it was mostly a sorting into industries. From today's evidence, we know that there might be intergenerational effects, and pollution might also have longer-lasting effects that might affect cognitive capacities. Longer-lasting effects that might affect cognitive capacities, that is, effects that outlast the original 19th century pollution. The idea is that children who grow up in those polluted areas suffer negative effects that lead to worse outcomes in education, health, and income, even if they were to move away later.

The UK, like the US, began cracking down on air pollution in the mid-20th century. But here's the thing. Piblick found that the effects of neighborhood sorting didn't go away. What we're seeing is that really polluted and really unpolluted neighborhoods, they are basically becoming even more extreme, either richer or poorer. What we're finding is that one standard deviation increase in pollution would lead in the past to about 15% higher share of low-skilled workers in neighborhoods. And then today, we would see that this would go up to 20%. The likely explanation is a classic case of path dependence.

You have the causes initially that the east side had these negative effects of pollution. Poor people sorted there. And then the effects were cemented over time by additional investments, right? Maybe you had the highway cutting off the east side from the west side, or you have poorer building structure. As a result of that, you have a certain composition of residents. You have less funding for schools. You have less funding for other amenities. And this is then the snowball effect.

In our paper, we find, for instance, that test scores in these east sides are lower and that crime instances are higher. Okay, lower test scores and higher crime in the areas that have historically had high pollution. But again, how can you untangle cause from effect? Does pollution itself lower people's cognitive abilities, or do people with lower cognitive abilities sort into polluted areas? Lower cognitive abilities may mean lower incomes, which may mean fewer options when it comes to where you live. And how can you untangle this question in the face of snowball effects like school funding? This brings us back to Andrea Linneaus. I had been reading the literature on the effects of air pollution on productivity,

but also other behaviors, for example, crime. And knowing that a leading hypothesis for those effects was really this cognitive impact. So there's a literature showing that the test scores of high school students is negatively impacted by exposure to particulate matter. But we didn't at that stage have much evidence for the cognitive effects in adults. And that makes sense because we don't regularly sit high school exams every year as adults. There was one piece of evidence for the cognitive effects of pollution on adults. It came from a paper that analyzed baseball umpires. Yeah, who said economics isn't fun? This was researched by James Archsmith, Anthony Hayes, and Sudeh Sabarian. They're able to compare quality of umpires' decisions on days of high pollution exposure and low pollution exposure.

And they did find that umpires made more mistakes when they were in a place that had a high pollution level on that day. That's a really important paper. It demonstrates that there is an impact on performance of really highly skilled professionals. But it is a study of a group of individuals that are probably fairly similar doing one task, an important task for one specific task. What the NOAAs wanted to see was the cognitive effect of pollution in a larger population across a diverse array of tasks.

I had recently been exposed to some advertising by Lumosity and suddenly thought, wow, well, there's a company that is claiming to test the cognitive ability of lots of adults across the United States. Like Linos, you may know Lumosity's name from advertising. They've sponsored a lot of radio and podcasts, including ours for a short time in 2014. Lumosity is a so-called brain game app created by Lumos Labs, a company founded in 2005. It now claims more than 100 million users across nearly 200 countries. So, lots of data.

They have something called the Human Cognition Project, where researchers can apply to either use their data or to use their platform to undertake other tests. Linose was able to get hold of data from more than 100,000 users across the U.S. playing a variety of games over a three-year period. So we have games that measure verbal ability, your attention, your flexibility, so how quickly you can shift from one cognitive task to another, your memory, so this is your very short-term working memory, your math ability, your speed, so speed of processing, and then also problem solving. There's a lot of controversy over whether Lumosity and similar products actually improve cognition. In fact, they paid a $2 million fine in 2016 for deceptive advertising.

But that wasn't the question Linoz was interested in. She and her research partner, the economist Edson Severnini, were looking at a different set of questions. They wanted to know whether day-to-day changes in air pollution in a particular place affected the scores of people who played games on Lumosity. The WHO recommends that in a 24-hour period, particulate matter should be below 25 micrograms per cubic meter. The EPA threshold is higher, at 35. Particulate matter is just one of the many pollutants the EPA tracks across the U.S., rolling up the total into a daily measure called the Air Quality Index, or AQI. Edson Severnini, as a researcher interested in air pollution, was already acutely aware

of how much variation there can be day to day. I always go for my morning walk and I always check on my phone what is the air quality index for the day. If it's below 50, you are in a good or green color of the AQI. If it's between 50 and 100, it's yellow, like the moderate pollution. And then above 100 is when I avoid leaving the house because that's where it starts getting a little bit unhealthy to be outside. Severnini happens to teach at Carnegie Mellon University, which happens to be in Pittsburgh, which happens to be historically one of the most polluted places in America.

For decades, it was a cradle of coal, iron, and steel production. When Charles Dickens visited in 1842, he wrote, Pittsburgh is like Birmingham in England. It certainly has a great quantity of smoke hanging over it. Like Gary, Indiana, Pittsburgh was a two-shirt town, and it's still rated as the ninth worst U.S. metro area for particle pollution. That said, not all parts of the Pittsburgh area have the same level of pollution on a given day, and the same goes for all the places that Severnini and Linneaus wanted to measure in their study. Pollution levels are not measured around us, like attached to our bodies.

So that would be the ideal experiment. You are breathing the air, you know exactly how much pollution you have in that air. It's not the case. And so that creates noise in the data, which would underestimate the relationship between cognitive function and pollution. But they did find a way to address that problem. We used the wind direction that brings pollution from other locations, and that makes a uniform

level of pollution for all individuals in an area, independently on whether they are close or slightly farther away from the monitor. Severnini and Linos ran their analysis across more than 4 million Lumosity gameplay observations and measured that against pollution data across the U.S. What they find? The headline result is that there is a cognitive impact for the working age population. In other words, it's not just among test-taking students. We're actually finding that the largest effects are for people under 50. And not just for baseball umpires either. So this is an issue for the working age population, and we expect that to have pretty significant productivity impacts. The second main result that I think is entirely

novel is that it does seem to affect memory ability. And so if we think across occupations, if we think about sectors that rely more on memory ability, we expect to see the productivity impacts in those areas be more significant. So what are we to make of this information? What kind of policy implications does it have? That's coming up after the break. And also... Hi, Stephen. Hey, Angie. Levitt's here too. Hey, Levitt. Hey, how are you doing, Demner? I play some Lumosity games with my Freakonomics friends, Angela Duckworth and Steve Levitt, who are not aware that this is about pollution levels in their respective cities, because I want this to be truly cutthroat.

This is the kind of thing I'm really good at. Like, I would honestly say this is my specialty. I'm probably more competitive than I'm... Oh, f***. And how would the climate change conversation be different if instead we were talking about pollution? Also, don't forget to check out the newest podcast from the Freakonomics Radio Network. It's called Freakonomics MD, hosted by Bapu Jenna, an MD and economist who loves to explore the hidden side of medicine using

piles of data and an uncanny intuition. The so-called brain training games from Lumosity may not make you smarter, but Freakonomics MD will, I promise. Follow Freakonomics MD wherever you get your podcasts. Angela Duckworth and Steve Leavitt, you're two of the smartest people I know. So I thought we could play some brain games today. Are you both feeling relatively sharp?

Are these the kind of games that one of us is going to win and the other two lose? God, the two of you are made for each other because the first thing Levitt said before we started recording was something about, what did you say, Levitt? I said, it's no fun to play with Angie because she's been playing these games her whole life because she's a psychologist.

And I know zero about psychology. I'll let them play these games all day long. But Levitt, you play trivia, at least. You love gaming. I do. You know, the problem for me is that one of the few things I have left is the belief that my brain still works. And if you take that from me, I'm going

to be really upset. Would you say there's any external factor that might contribute to a subpar performance today? Maybe you didn't sleep well last night. Do you want to just pre-register your conditions? I never sleep well. I have way too many kids. That's my standing excuse is that I haven't had a good night's sleep in about 18 years. Angie, anything you want to register? Well, as you know, I'm a pretty sleeper, but improbably I actually slept fine last night or uncharacteristically, I should say. So there's that. I was going to complain about the time of day, but then again, it's more or less the same time of day for all of us. Yeah, we should say it is late in the day. It's a little after 5 p.m. on the East Coast and Levitt's in Chicago. So that's still end of the day. So he's got a one hour advantage on this.

Can I say my air conditioning is broken and it's really hot? Oh, okay. You win then. In case you don't know, Steve Levitt is an economist at the University of Chicago. He is my Freakonomics co-author and he hosts a podcast called People I Mostly Admire. If you don't already listen to that, you should. Angela Duckworth is a psychologist at the University of Pennsylvania in Philadelphia. She's the author of the book Grit and she is my co-host on the No Stupid Questions podcast, which you should also be listening to. Anyway, the three of us set out

to play three games as part of Lumosity's fit test. One game is said to measure mental flexibility, another memory, and the third, called Train of Thought, purported to test our attention by having us guide different colored trains to their respectively colored destinations. If all that sounds super easy, well, you should try it. Okay, here we go. This is so cute. I love trains. Oh, my God. This is hard. It's one of the hardest things I've ever done in my life.

Oh, my God. Let's see. I got 13,500 points, and I scored better than 70%. I don't even want to tell you guys how I did. Come on. Tell me how you did. I was 34 out of 39, 97%. Wow, Levitt. I just said it was the hardest thing I'd ever done.

I didn't say I was messing up. 97%. Levitt, I'm really impressed. I wonder what strategy you used. 97% wasn't Levitt's actual score. It was his percentile ranking for his age group. That's how Lumosity ranks you. After playing all three games, the results were in.

Levitt was the clear champion with an average percentile rank of 92. Pretty impressive. Although I suspect Levitt may have logged in earlier under a pseudonym to practice. He is sneaky like that and very competitive. On the other hand, he's also really smart, so I'm probably wrong. Angela, meanwhile, was very consistent across the three different tasks, but her average was lower, 71st percentile. For what it's worth, my scores were inconsistent,

a high memory score, but really low attention, which probably has something to do with... Wait, I forgot what I was going to say. Anyway, my average was around the same as Angela's, 72nd percentile. So Angie, how do you feel about your performance on these games today? Well, I'm pretty disappointed, Stephen. I like to think of myself as better than a C-minus brain, but maybe I'm, you know, less smart than I thought I was, at least on these games. Levitt, how do you feel about your performance today? You know, I'm relieved because I have the self-image

that these stupid little games are my forte. I have to say, actually, at the particular moment when we were doing it, I felt great. I mean, I don't sleep as much as I would like to, but honestly, no, I felt very sharp today. I cannot think of a single excuse for not doing well in these games. I'm curious if either of you have ever thought about particulate matter pollution in the atmosphere as a potential contributing factor to cognitive ability? What do you mean particulate matter in the atmosphere? No, I guess the answer is I haven't thought about that. I don't even know what you're talking about. On a particular day, Dubner? You

mean like how much is in the air today? Yeah. So what would you say if I told you that a couple economists have analyzed Lumosity gameplay, just like we did, in different places and found that, quote, even when air pollution is below EPA and World Health Organization quality guidelines, cognition is negatively affected across seven different cognitive domains. Furthermore, their identification relies only on short-term changes in pollution exposure within an individual's play history. Would that surprise you? I'm trying to process this. Levin, what do you think? So I can't say I've heard many more theories that would surprise me more if they were true. But what do I know about the world? So let me read you some numbers. The WHO recommended level is 25 micrograms per cubic meter. The EPA's recommended level is higher, 35. This paper finds negative

cognitive effects at just 20 micrograms per cubic meter. Now, here's what's interesting. In the three cities where we are, I'm in New York, Angela's in Philadelphia, Levitt's in Chicago. On average, in 2019, for instance, New York was the lowest of those three at seven micrograms per cubic meter. Philly is at 10.3, and Chicago was the worst at 12.8. How many particulates are there in Chicago today? Okay. So I have some good news and some bad news. The good news is, Levitt, you are suffering very low particulate matter in Chicago today. As of today, Chicago only had 8.7 micrograms per cubic meter. Philadelphia and New York,

we have very high levels today, as it turns out. Do we? Yeah, 23.4 in New York and 24.6 in Philadelphia. Is there that much variation in particulate matter? There is that much variation, not only place to place, but day to day. Wow. That's what's really interesting, the day to day part. I didn't realize that. Levitt, earlier you said that you just felt incredibly sharp and focused when it came time to do the tasks. Do you think that had anything to do with the

relatively low level of particulate matter in the air in Chicago? I wouldn't think so, but maybe I should start tracking it. I could, without knowledge of the particulates, read how I felt each day. So you each sound relatively skeptical of the findings of this paper. Let me just ask for like a confidence level, zero to ten, let's say, that these findings are somewhere in the ballpark of useful and true. I want to rate my own confidence in saying anything about somebody's findings before reading their paper. I would say that would be like a one.

Okay, fair enough. Levitt, do you want to speculate? I would say it feels like a one in terms of likelihood of being true, and if true, a 10 in terms of importance. So what is the likelihood that local, real-time pollution levels can impair cognitive function in the moment? It might help to know the mechanisms by which this could happen. I'm not a medical expert. What I'm going to say now is based on, you know, reviews of this literature.

That, again, is the Carnegie Mellon economist Edson Severnini. There are two ways where air pollution could impair cognition. One is that they go directly to the brain and then it affects the functioning of the neurons. But also they stimulate pro-inflammatory, I think it's called cytokines. And so this is a more indirect route. But everybody who is doing research on this topic, they always see processes that are affected by pollution, oxidative stress, inflammation, some neuron loss. I should note that Angela, Steve, and I played only a few games on one day. The data that Severnini and Linneaus analyzed was much more robust. Still, I asked

whether our scores should be adjusted based on that day's pollution levels in our respective locations. The impact of this exposure to pollution would be to shift someone in that ranking by about six points. So if you were playing on a day that was above the threshold that we sat and you were performing like in the 75th percentile, on average, you would have been on the 80th percentile that day. And that's just on average. So once you account for the local pollution levels in New York and Philadelphia that day, which were high, and Chicago, which was low, Angela and I might be right up there with Leavitt. In any case, Linneau's and Severnini's study is under review for publication. We asked

Michael Greenstone, the pollution and policy veteran, what he thought of their findings. Now, do keep in mind that Edson Severnini was actually a postdoc under Greenstone. You know, it's a very well done paper in a kind of artificial setting. Long run meaning is a little bit hard to suss out. The more challenging thing is, you know, to find instances where there's long run variation. I think in both the health and in the cognition literatures, the holy grail is not to rely on studies that use either day to day or month to month, but to find a setting where there's like a permanent difference in air pollution. It's much harder to come up with those examples, but that is after all what policy is trying to do. It's not trying to

reduce pollution on Tuesday. It's trying to reduce pollution 365 days a year. Greenstone thinks he may have found the Holy Grail. About seven or eight years ago, I stumbled upon an example from China that seemed to mimic this kind of ideal. And that's something called the Huai River winter heating policy. It dates back to when China was much less wealthy, and there just weren't enough resources to provide winter heating for everybody. So they did something quite arbitrary and capricious. They drew a line across the middle of the country, and that line followed the Hawaii River. The Hawaii River, by the way, runs west-east, not north-south. And they said, okay, if you live north of that line, where it's colder,

we're going to install central heating systems, and we're going to give you free coal. So that's in the north. In the south, the policy was, guys, you're out of luck. No heating. So what Greenstone was looking at had nothing to do with whether people sorted themselves into neighborhoods on the east or west side of a city, like Stefan Hidlick looked at in England. This had to do with comparing the health and educational outcomes of people living on the north side of the river, where people were warmer in the winter but exposed to a lot of coal smoke, and the south side, where you might have

been colder but didn't have much coal smoke. And thanks to Chinese government policy, there was almost no migration from one side of the river to the other. Migration was greatly limited. And I thought, wow, this is the thing I've been searching for. Greenstone was able to analyze data that included roughly 40,000 people living in urban areas within a five-degree latitude range north and south of the river. The first outcome he looked at was life expectancy. If you were born just to the north of the river, those people, they were the intended beneficiary of this policy. On average, they're living about three years less than people born just to the south. And that was such a striking finding, at least to me, that I thought,

wow, I hadn't realized quite how devastating air pollution was, even though I've been working on it. In subsequent research, soon to be released, Greenstone looked at the educational outcome of kids born between 1975 and 1982. Here, he's trying to estimate the cognitive effects of coal pollution. Children born just to the north of the Hawaii River completed almost one full year less of education than kids born just to the south. And not just that, we were able to observe them as adults. And on average, they earned about 13% less than children born just to the south. I think this is the first large-scale evidence on the impacts of long-run early childhood exposure at the levels of concentrations that prevail in many parts of Asia and sub-Saharan Africa.

So how bad is this news? Or maybe a better question to ask, just how damaging is the cognitive impact of air pollution? We have probably been understating the losses from air pollution by about 50%. But then, some kind of good news. That would imply that the benefits of reducing air pollution are 50% larger than we realize and would justify more stringent environmental regulations. High-polluting countries, especially China, have been pushing hard to lower air pollution. As recently as 2013, the particulate matter level in Beijing was over 100 micrograms per cubic meter.

Remember, the level over a 24-hour period recommended by the WHO is under 25. But by 2018, the average level in Beijing had fallen to just over 50, and it has continued to fall across the country. And if you take my estimates literally, they imply that a child born in 2018 relative to a child born in 2013 will live 1.4 years longer. Greenstone says China's trajectory is much more dramatic than ours. The United States accomplished nothing like that so quickly after the Clean Air Act. And as an economist who's done a lot of work on environmental policy,

he's been disappointed with the U.S. government's approach to the broader issue of climate change. The Clean Air Act was really focused on reducing pollution locally in parts of the country where pollution concentrations were very high. CO2 is a totally different ball of wax in the sense that it is a global pollutant. The impact of emitting a ton of CO2 in Fresno is exactly the same as emitting a ton in Bangor, Maine. I think the United States is an extraordinary outlier in the international arena in terms of its difficulty in recognizing and developing a coherent strategy for confronting climate change. The United States is the only country in the G7 that does not have a coordinated national climate policy, and that's striking. I've always wondered why the conversation about climate change hasn't been more of a conversation about pollution. I've also wondered if the climate change conversation might not have become so ideological if it were more about pollution.

The evidence for pollution's impact, the longstanding evidence about the physiological damage and what we've been hearing today about the cognitive damage, that evidence is so persuasive that it's hard to imagine any right-minded human not wanting to fight that fight. It isn't just progressives or Democrats who want clean air and water. Some of the strongest-willed naturalists and preservationists have historically been politically conservative. No one wants their babies or grandparents breathing polluted air. Edson Severnini again. I think making the argument that it affects people in their daily lives, you know, it could be their own productivity

or their children's performance in school or their children's performance in sports that they're playing outside. All of this should be really talked about more often is cool, or their children's performance in sports that they're playing outside. All of this should be really talked about more often when it comes to energy and environmental policy. Talked about may be like this. The level of particulate matter in the air today is above the recommended World Health Organization guidelines. You know, it's a matter of making sure people understand the consequences

because it is sometimes not visible. If your child has an important test today or you're giving a big presentation at work, you might want to consider rescheduling. You go for a walk and you don't see it, the pollution. The Supreme Court will be delaying oral arguments until next week because of a high particulate matter count in Washington, D.C. But, you know, in economics, we always say, like, if you have a problem, you tackle that problem directly. It's much more efficient. Thanks to Edson Severnini, Andrea Linneaus, Michael Greenstone, and Stefan Hiblich for telling us about their research today. Thanks to Angela Duckworth and Steve Levitt for playing brain games with us. Most of all, thanks to you for listening.

If you love Freakonomics Radio or any other show in the Freakonomics Radio network, please recommend it to someone you know. We appreciate your spreading the word. We'll be back next week. Until then, take care of yourself, and if you can, someone else too. Freakonomics Radio is produced by Stitcher and Renbud Radio. We can be reached at radio at Freakonomics.com. This episode was produced by Zach Lipinski. Our staff also includes Allison Craiglow, Greg Rippin, Joel Meyer, Trisha Bobita, Mary Deduke, Ryan Kelly, Brent Katz, Emma Terrell, Lyric Bowditch, Jasmine Klinger, and Jacob Clemente.

Our theme song is Mr. Fortune by The Hitchhikers. All the other music was composed by Luis Guerra. You can get the entire archive of Freakonomics Radio on any podcast app. If you want to read a transcript or the show notes, that's at Freakonomics.com. Again, thanks for listening. Oh my God, this is hard. Oops. I already blew that one. I sent those people to their death. Okay, now I feel like so dumb. The Freakonomics Radio Network.

Stitcher.