Change Is in the Air?
Construction sites in China are ubiquitous, and the fences that guard the sites almost always feature a computer-generated image of what the site will look like upon completion. These images feature verdant shades of green for grass and trees, shiny new buildings, and a background that fades into the smog. I was struck by the smog in one of these images during my first month in China, and since have noted the rarity with which one finds an image that doesn’t feature smog. My Chinese friends generally mention something about being realistic when I ask them, but I am struck by the question: is a smoggy skyline built into the software that generates these images, or does nearly every graphical artist opt to shade the surrounding buildings in a semi-opaque grey? One thing is certain—smog and bad air quality are omnipresent in China.
Last November I wrote an article that tried to come to terms with China’s air pollution, and to investigate the claim that the pollution in Beijing was as bad as smoking as many as a pack or two of cigarettes per day. In short, based on a few of the most prominent articles about the effects of air pollution, and particularly pm2.5 particles, on human health, I found that at its worst, the air pollution in China amounts to smoking just over a cigarette per day. This is based on the assumption that all pm2.5s behave the same, and that this is true no matter how you inhale them—over the course of a couple minutes for a cigarette, or over the course of the whole day for air pollution. At the very end of the article I suggested that the graph of relative risks shows that air pollution carries a higher risk of getting sick than cigarettes do for the (relatively) low concentration rates for which it has been measured.
In this article I hope to follow up on this conclusion using information from several studies that attempt to directly measure the effects of air pollution on Chinese residents in terms of life expectancy. I also explore the conclusions of a study about the accuracy of air pollution reporting by Chinese government agencies.
First, I hope to correct an error in the first article, which may explain much of the difference in the impact of inhaling similar amounts of particulates from cigarettes vs air pollution. The original article showed a higher relative risk across the board for illnesses (like lung cancer and heart disease) for inhaling between .3 and .5 mg of PM2.5 of air pollution than for inhaling the same amount of PM2.5s of cigarette smoke. It turns out that this is because not all PM2.5s are created equal—and while how those PM2.5s interact with our bodies still isn’t very well understood, it does appear that what the particles are coated in is important. What makes air pollution particles especially bad is that they tend to be coated in more reactive substances than their cigarette-smoke counterparts, increasing the impact they have when they enter our systems.
In general, smokers inhale much higher quantities of particulates than those who breathe polluted air, so their studies tend to focus on a much higher value for particulate inhalation than air pollution studies. This also makes the two hard to compare. In the previous air pollution article I described the difficulties of attributing illness or death directly to air pollution because we can’t study a world where everything is the same except for pollution, so using data from smoking studies is the best we can do. Fortunately for us, a few researchers have attempted to tackle this question with some interesting research designs.
In 2013 a team of researchers conducted a study that used the Huai River policy to observe the effects of additional air pollution on the population. The Huai River policy lasted from 1950 to 1980 and provided free central heating to homes north of the Huai River, mostly in the form of coal-powered heating units. As many of these units continue to heat northern homes, the policy has had lasting effects on air pollution, and thus on respiratory and cardiovascular illnesses. Furthermore, China’s hukou policy heavily restricts migration, generally preventing those living in the south from moving to the north and vice-versa, reducing the uncertainty in the study. Ultimately, this study found that the life expectancy for people living in the north was 5.5 years shorter than for those living in the south, probably caused by an increase of 184 μg/m3 of total suspended particulates (all particulate matter, not just PM2.5s). While not perfect, this study does a pretty good job comparing similar places that have different levels of air pollution, and using that to estimate the health effects of that pollution. The study goes on to conclude that exposure to an additional 100 μg/m3 of airborne particulates over the course of one’s life is associated with a 3-year decrease in life expectancy. For comparison, the US Center for Disease Control claims that smoking is associated with a 10-year average decrease in lifespan.
A study last year by the Beijing Center for Disease Prevention and Control paints an even gloomier picture: that an 18-year-old male can expect to spend nearly 20 of their remaining (average of) 62 years in ill health. Even worse, an 18-year-old female can expect to spend nearly 28 of her remaining years in ill health. The study uses the World Health Organization’s measure of Health Adjusted Life Expectancy (HALE), or how much of their remaining life an individual can expect to spend in good health. While this study doesn’t aim to distill the effects of air pollution, it seems likely that the cause of at least a large portion of the city’s woes is air pollution. Another peer-reviewed publication on the topic of the effects of Beijing’s air pollution on the health of its residents supports this claim (but uses a measure, Years of Life Lost, that would take too much space to explain here).
With the effects of air pollution almost certainly clear to policymakers, and mostly available to anyone willing to put in some research (except for the Bloomberg piece, all of the sources linked here are available inside the great firewall), it seems hard to believe that the Chinese government is doing nothing. Indeed, last year Prime Minister Li KeQiang declared a war on coal, and in 2014 coal consumption decreased for the first time in 15 years. Additionally, in a system that bases promotion of officials on a set of performance indicators, environmental conditions have been made part of the evaluation process. This includes air pollution, and in particular, the number of “blue sky days”, or days in which the overall air quality index, or AQI is lower than 100. This term has made its way into the common lexicon—following a string of uncharacteristically clear days in Beijing during the 2014 APEC summit netizens began using the term “APEC blue” to describe a seemingly good thing that won’t last.
Last fall, Dalia Ghanem and Junjie Zhang of UC Davis and UC San Diego published a study that used statistical techniques to demonstrate that many of China’s most polluted cities were manipulating their air quality data to artificially increase the number of “blue sky days” they could report. While air pollution over the course of a year tends to be normally distributed, or fall under a “bell curve”, Ghanem and Zhang found that the number of days reported with just below an AQI of 100 far exceeded the number of days with just above an AQI of 100. In other words, officials in polluted cities were trying to improve their chances at promotion by taking days where the AQI was just above the threshold for a blue sky day and reporting it as just below the threshold.
This chart, taken from the Ghanem and Zhang article, shows the distribution of daily concentrations of PM10s in Chongqing. Note the discontinuity at about .15—a value which if it is exceeded results in an AQI of above 100, and thus a non-blue sky day. For those seeking to manipulate the data to increase the number of blue sky days in the future while avoiding detection, I recommend scaling the whole curve based on the previous year’s data. Email me if you’re interested. 😉
Why is this important? Well, for one, it is difficult to formulate an effective response to a technical problem if you have bad data about what the problem actually is. Furthermore, it likely contributes to negative health consequences for people who rely on air pollution data to plan activities like outdoor exercise, or even when to wear a pollution mask.
While air pollution is a huge and growing problem in China, it is important to understand that the news isn’t all bad. For example, the silver lining of data manipulation is that air pollution carries enough weight in cadre evaluations to create incentives to fake the data. With stricter enforcement, and especially with promotion incentives weighted more toward cleaner air, probably a number of officials would be willing to bite off the costs of forcing polluters to clean up.
Less than two days ago one of China’s most famous investigative journalists released a self-funded documentary about China’s air pollution. By the evening of the day after its release, the video had garnered over 20 million views on Youku, and at least double that across platforms, according to a Chinese website called The Paper. The documentary blends the story of her daughter, diagnosed with a benign tumor at birth (which she strongly implies is the result of air pollution), who she struggles to protect from Beijing’s worsening air quality. A New York Times article further chronicles the film, and if you have the Chinese ability, I strongly suggest you watch the video itself:
Here it is on Youku for viewers in China.
And a version with English Subtitles (a work in progress)
The entire video is pretty hard-hitting. It goes from comparing China’s air pollution to similar problems that have been largely resolved in the US and Great Britain, to interviewing officials who likely face political repercussions for their positions in the video. However, probably the most interesting part is when Chai Jing tells viewers that only by acting together can we create change. This is exactly the type of rhetoric that is most targeted by censors, and presents somewhat of a mystery about why this video was allowed to be published in the first place. As of now, rumors abound online of some parts already having been censored (the video lends itself to this theory, with a few awkwardly brief cuts). The new Minister of Environmental Protection, Chen Jining, thanked Ms. Chai on Weibo for her work in producing the video. The following day, a propaganda directive was released telling media not to hype the new film, according to the China Digital Times website.
Perhaps the best indication of the current state of affairs of the “War on Air Pollution” is a new environmental law that came into effect on January 1, 2015. The law is by far the most strict attempt to protect China’s environment that has ever been implemented, but according to the journal Nature, still likely faces the significant challenges of 1) it can be superseded by other laws that are likely to favor industry over the environment 2) fragmented oversight in the form of many agencies that already tend to pass the buck when it comes to enforcement 3) lack of a basic right to a clean environment and 4) conflicts of interest at the local level for officials that tend to favor economic growth above all else.
The war on air pollution that started in 2014 is already beginning to have positive effects on air quality. Chengdu’s air quality this year is already much better than the same period last year, and for the first time since the US consulate started collecting data in both cities, Beijing’s air has been better than Chengdu’s for January and February. Now, the question of whether China’s future is an APEC blue or more like the pictures on construction site fences likely depends on the degree to which officials are willing to sacrifice environmental progress to prop-up economic growth in the face of slowing growth numbers. The answer to this million-dollar question might be worth a billion lives.
Updated March 2: Video with English subtitles, propaganda directive added.