750,000 a year killed by Chinese pollution

There is Richard MacGregor's article titled 750,000 a year killed by Chinese pollution accompanied by a Financial Times editorial titled  China must come clean about its poisonous environment.  Here is Richard MacGregor's report:

Beijing engineered the removal of nearly a third of a World Bank report on pollution in China because of concerns that findings on premature deaths could provoke “social unrest”.  The report, produced in co-operation with Chinese government ministries over several years, found about 750,000 people die prematurely in China each year, mainly from air pollution in large cities.

China’s State Environment Protection Agency (Sepa) and health ministry asked the World Bank to cut the calculations of premature deaths from the report when a draft was finished last year, according to Bank advisers and Chinese officials.  Advisers to the research team said ministries told them this information, including a detailed map showing which parts of the country suffered the most deaths, was too sensitive. 

“The World Bank was told that it could not publish this information. It was too sensitive and could cause social unrest,” one adviser to the study told the Financial Times.  Sixteen of the world’s 20 most polluted cities are in China, according to previous World Bank research.

Guo Xiaomin, a retired Sepa official who co-ordinated the Chinese research team, said some material was omitted from the pollution report because of concerns that the methodology was unreliable. But he also said such information on premature deaths “could cause misunderstanding”.  “We did not announce these figures. We did not want to make this report too thick,” he said in an interview.

The pared-down report, “Cost of Pollution in China”, has yet to be officially launched but a version, which can be downloaded from the internet was released at a conference in Beijing in March.

Missing from this report are the research project’s findings that high air-pollution levels in Chinese cities is leading to the premature deaths of 350,000-400,000 people each year. A further 300,000 people die prematurely each year from exposure to poor air indoors, according to advisers, but little discussion of this issue survived in the report because it was outside the ambit of the Chinese ministries which sponsored the research.  Another 60,000-odd premature deaths were attributable to poor-quality water, largely in the countryside, from severe diarrhoea, and stomach, liver and bladder cancers. 

The mortality information was “reluctantly” excised by the World Bank from the published report, according to advisers to the research project.

Sepa and the health ministry declined to comment. The World Bank said that the findings of the report were still being discussed with the government.  A spokesperson said: “The conference version of the report did not include some of the issues still under discussion.” She said the findings of the report were due to be released as a series of papers soon.

I look at this sentence: "Sepa and the health ministry declined to comment" and I think these people must be brain dead.  Silence is likely to be taken to be admission of guilt that will generate a wave of criticisms.  Why not deal with the issue head on?

First, consider the World Bank study.  Actually, I have no information about this study itself, but I take it to be a top-quality study like similar studies on the same subject.  The problem here is that it is hard to estimate the number of pre-mature deaths due to pollution.  If you want to estimate the number of excess deaths due to the war in Iraq, you compare the mortality rates before and after the invasion and you get the number of excess deaths (e.g. the Lancet study).  If you want to estimate the number of pre-mature deaths due to tobacco, you can compare the age-specific-adjusted mortality rates among smokers and non-smokers (e.g. American Medical Association study).  In these cases, you always qualify your results by stating that there can be unaccounted for causes (and that was how the tobacco lobby managed to stall anti-tobacco legislation in the United States for several decades).

But if you want to estimate the number of pre-mature deaths due to pollution, you do what?  You cannot compare the mortality rates of the present China against a China without pollution.  The latter does not exist objectively.  Therefore, you have to make some assumptions (e.g. compare mortality rates in high pollution cities versus low pollution ones), but these assumptions are tenuous.  For example, the high pollution cities are the economically developed urban agglomerations while the lower pollution cities are economically undeveloped cities in the backlands.  Are you sure that they would have the same mortality rates if pollution does not exist?  Have you thought about the impact due to differences in income, diet, nutrition, health care services, etc?  The technical explanations may elude most readers, and that may be the reason why Sepa and health ministry do not want to comment as it may only make matters worse.  But their silence is making things worse now.

For illustration about why the technical description is hard, here is a statement made at a US Congressional hearing:

There are two ways of studying the health effects of particulate matter: time-series studies and cohort studies. Time series studies track people over short pollution episodes, correlating morbidity (illness) and mortality with daily pollution levels. As of 1997, numerous time-series studies had reported associations between PM and daily mortality and morbidity. Landmark studies include Dockery and Pope (1994), Schwartz (1994), Katsouyanni, et al. (1997). These studies were criticized because they were largely conducted in single locations chosen for unspecified reasons and were analyzed with different statistical approaches.

In 2000, a Health Effects Institute study used explicit criteria to select cities from a well-defined sampling frame and analyzed them in a consistent fashion. The results from this 90-city study corroborated previous results, including the Katsouyanni 15-city study and a recent meta-analysis of 29 studies in 23 locations in Europe and North and South America (Levy, et al. 2000).

Cohort studies follow initially healthy people over longer periods to see how they develop disease or die. Landmark cohort studies include the Harvard Six Cities Study (Dockery, et al. 1993) and the American Cancer Society Study (Pope, et al. 1995). These studies followed large numbers of individuals over many years and observed their rates of mortality. They found that long-term average mortality rates were 17 percent to 26 percent higher in those living in communities with higher levels of PM2E even after accounting for the effects of other risk factors. The results have been used to demonstrate lifespan reduction attributable to exposure to PM2E pollution. Other researchers have independently confirmed the findings of both of these studies.

A recent study reanalyzed the American Cancer Society (ACS) data (Pope, Thurston and Krewski 2002). This study tracked people over a longer time and controlled more extensively for individual risk factors. They compared data on particulate and gaseous air pollution with data on the cause of death among 500,000 people followed for 16 years by the ACS. After compensating for risk factors, as well as possible regional differences, the researchers found that every 10-microgram increase in fine particles per cubic meter of air produces a 6 percent increase in the risk of death by cardiopulmonary disease, and 8 percent for lung cancer. This is similar to the risk faced by those with long-term exposure to second-hand smoke.

Here is the statement of the limitations of methodology from the preliminary February 2007 World Bank report titled Cost of Pollution in China (see page 28):

Limitation of study areas

The meta-analysis by Aunan and Pan (2004) was primarily based on the results from several large cities and not on middle and small cities.  However, the characteristics of air pollution in middle and small cities are often different from those in large ones, and the age structure and susceptibility to air pollution of the local population may also vary with city size.  So the extrapolation of the exposure-response functions to the other cities should be considered carefully.

Limitations of methodology

Large-sample epidemological cohort studeis, similar to those carried out in the U.S. to study the effects of long-term exposure to PM on mortality, have not been undertaken in China.  Most cross-sectional studies in China have been ecological studies, in which no detailed information at the individual level is collected.  This implies that the studies in different locations may not be comparable due to site-specific characteristics.  More importantly, the studies do not control for confounding factors that may affect mortality (such as socioeconomic status), which may also be correlated with air pollution.  

What is the general public going to think about this type of description?  They don't know, they don't want to know about methodological issues and they only want a count of the number of dead people due to air pollution

Secondly, how would you interpret the figure 750,000?  If you line the corpses in a row down the road, then it is a horrible sight.  You would like it to be zero, but that is not happening anywhere in the world.  So what is a realistic goal for China to aim for?  Or to be even simpler, what is the number if China were just to meet the global average?  Here is a global estimate from BBC:

The World Health Organization (WHO) says 3 million people are killed worldwide by outdoor air pollution annually from vehicles and industrial emissions, and 1.6 million indoors through using solid fuel. Most are in poor countries.

So 3.0 + 1.6 million = 4.6 million people were killed worldwide by air pollution.  China has about 20% of the population of the world.  20% of 4.6 million people is 920,000.  This is for indoor/outdoor air pollution only.  The World Bank estimate of 750,000 is for all pre-mature deaths in China, most of which are due to air pollution.  Now what?  It would seem that China has even fewer pollution-related deaths compared to the rest of the world! 

Unfortunately, it does not work that way.  I have no idea what the methodology in either the World Health Organization or the World Bank studies are.  Therefore, I have no idea what this comparison means.  If I am going to have an informed opinion, I better be informed first.

Related Links:

China cancels environmental report  Mitchell Landsberg, Los Angeles Times.  July 24, 2007.

From a public relations standpoint, it didn't look good. In the space of less than a month, China had quashed two potentially embarrassing environmental reports that would have said what most people already know: This is a country facing a costly and increasingly deadly environmental crisis.

First, in early July, reports surfaced that China had successfully lobbied the World Bank to redact portions of an environmental assessment that calculated how many people were likely to die prematurely as a result of air pollution.

Then, late last week, the government announced that it was canceling plans to publish a "green GDP" report that would have calculated the cost of pollution to China's rapidly growing economy, as measured by its gross domestic product.

The decisions, on their face, appeared to suggest reluctance at the top of China's government to acknowledge the seriousness of environmental degradation that has caused the worst air pollution in the world, and water pollution that has left millions of people without local sources of potable water.

Chinese and Western experts, however, said Monday that authorities might have acted for reasons not readily apparent to casual observers. They said the reluctance to publicize the country's environmental woes might have had more to do with political relations between the central government and provincial leaders than with a fear of airing dirty laundry.

"As soon as you develop a system like this, then you can do a ranking of environmental performance of local governments," said Andres Liebenthal, the environmental coordinator for the World Bank office in Beijing, who worked with China's environmental protection agency on both of the reports. "And so the ones that are highly ranked are fine, and the ones that are ranked low are not happy with it, so there's a pushback."

On paper, many environmentalists agree, China has some of the strongest pollution control policies in the world. Its effort to calculate the environmental toll on its GDP was bold by international standards. Environmentalists have failed to persuade many developed nations, including the United States, to undertake such an accounting.

China released a "green GDP" report for the first time in September and was preparing a second annual report when the decision was made to spike it. The report last year calculated the cost of pollution at $67.7 billion, or just over 3% of China's gross domestic product

Despite relatively strong laws, enforcement of China's environmental policies is patchy at best, largely left to provincial governments that have a stake in local economic growth, regardless of the environmental cost. And it is those officials, some experts believe, who may have put the brakes on the recent reports.

Wang Jinnan, a senior expert at the Chinese Academy for Environmental Planning who played a lead role in the GDP report, was quoted Monday in the Beijing News as saying the effort had drawn fierce opposition from local officials eager to maintain growth.  Taking into account "the costs of environmental damage would lead to a huge fall in the quality of economic growth in some areas," Wang said. "At present, many areas still place GDP above all else, and when such thinking dominates, the size of resistance to a green GDP can well be imagined."

Christian Averous, who recently wrote a report on China's environment for the Paris-based Organization for Economic Cooperation and Development, agreed that the central government appears to have limited control over how environmental policies are carried out.  "There is an implementation gap concerning environmental policies in China," he said. "A number of laws and policies are not implemented, and this is partly due to territorial governments' attitudes."

Averous added that his report was written at the request of the Chinese government and that he faced no pressure to censor or withhold his findings. The report included estimates from earlier reports that 600,000 people a year could die prematurely by 2020 because of urban air pollution and offered 51 specific recommendations for cleaning up the nation's environment.

The Chinese government has pledged to adopt the recommendations, Averous said. They include holding local leaders more accountable to the central government and giving China's environmental protection agency the status of a full government ministry.

Liebenthal said the Chinese government objected to the inclusion of mortality figures in the recent World Bank report because they were based on new and controversial methodology. According to an article in the Financial Times newspaper, the report would have said that about 750,000 people in China die prematurely each year, mainly because of urban air pollution.  "The statistics are problematic," Liebenthal said. "There are legitimate scientific arguments about the appropriateness and accuracy of the methodology."

But Liebenthal said he believed that it would be in China's best interests to publish the data, particularly on the environmental costs to the economy.  "On the whole, I think this kind of research is important for China to do," he said. "China is an important country, pollution is important in China, and it is important for the country to understand the extent to which its economic progress is maybe less than it seems because of some environmental side effects that need to be taken into account."