Issue Analysis 48 – Weird Science: Did CASAC Really Support PM2.5?

Recent efforts by the Environmental Protection Agency (EPA) to implement a new air quality standard for fine particulate matter have come under attack. The shaky scientific underpinnings and potential high cost of the proposed standard have caused many observers, including the Department of Transportation and other federal agencies, to show concern that there may be no justification for spending billions of dollars for a new PM2.5 standard. To refute these claims, the EPA cites its own Clean Air Science Advisory Committee (CASAC) staff report as evidence of scientific consensus on the need for a PM2.5 standard. The staff report, as well as a majority of CASAC’s members, supported a new standard in a 19-2 vote, but a close look at CASAC’s May 16th and 17th discussion of the PM2.5 standard shows that no consensus was reached on the issue of health benefits resulting from a new standard. In fact, only two CASAC members voted for a standard as low as that imposed by EPA Administrator Carol Browner.

During its regular review of the National Ambient Air Quality Standards (NAAQS), the EPA decided to propose a new air quality standard for particulate matter 2.5 microns and less in diameter (PM2.5). The standard would implement an annual and daily standard for PM2.5 at 50 and 15 micrograms per cubic meter, respectively. The daily standard would take an average of the 98th percentile of daily PM2.5 levels over three years, which would effectively allow a community to exceed the standard about six times a year. The annual standard would be based on a three-year average of annual PM2.5 concentrations, allowing only 15 micrograms per cubic meter.

Fine particulate matter includes particles such as common dust and smoke from burning wood. The sources of PM2.5 are numerous and include just about everything that burns or blows in the wind. The three largest sources of PM2.5 are paved and unpaved roads, crop dust and residential wood burning. Other, smaller sources, include emissions from industrial combustion, wildfires, industrial processes, and waste disposal and recycling. Cost estimates for PM2.5 controls on all sources are difficult to calculate because only a few monitors of PM2.5 exist, limiting the EPA’s ability to measure the substance. The EPA claims the costs will be around $6.3 billion annually, but this estimate was only for partial compliance of the standard.1 Some estimates put the cost of full compliance as high as $55 billion a year.2 The Science on Particulate Matter. CASAC members cite many reasons why the science behind the new PM2.5 standard is problematic. Most of these problems stem from the fact that the research on particulate matter is incomplete. A review of the proceedings of two CASAC meetings, make it clear that many panel members are uncertain if particulate matter exposure induces significant negative health effects. The lack of proof that mortality results from particulate matter exposure is due to many problems with the science, but other important issues were raised over and over again in the CASAC meetings.

One is the lack of data on actual human exposure to fine particulate matter. This basic information is the foundation of any health study on particulate matter or any other pollutant. Unfortunately, there are only a handful of monitors in the United States currently measuring the levels of PM2.5. Without this important data, scientists are forced to use rough estimates of human exposure levels to PM2.5, which can be very inaccurate. Using other information such as the EPA’s abundant data on coarse particulate matter (PM 10) levels, scientists have generally estimated how much PM2.5 people actually breathe in many areas of the country. Estimating the exposure to PM2.5 using PM 10 data, however, is difficult and fraught with uncertainty. Approximating the actual human exposure to PM2.5 could easily overestimate the number of deaths due to PM2.5, but without the necessary monitoring and analysis, certainty is impossible. This is why many of the CASAC scientists emphasized the need to build a monitoring network that would generate the necessary data to provide an accurate analysis of human exposure to PM2.5.

Some CASAC members commented that PM2.5 mortality studies might have overestimated the average individual’s exposure to outdoor air. They were concerned that the current science on PM2.5 might not consider that Americans spend much of their time inside, thereby limiting exposure to any polluted outdoor air. (The EPA estimates that we spend about 90 percent of our time indoors.)3 If people spend most of their time inside, research on the effects of indoor air pollution, which has been suggested to have a much more significant health impact, would be very important. The EPA confirms problems with indoor air pollution in its guide to indoor air quality: “[A] growing body of scientific evidence has indicated that the air within homes and other buildings can be more seriously polluted than the outdoor air in even the largest and most industrialized cities.”4 Thus, the possible mortality from bad indoor air must be considered when calculating the health risks from PM2.5. The risks to health for many people may be greater from indoor rather than outdoor air. Further study into outdoor versus indoor air exposure is necessary before adopting a new standard.

Another important piece of evidence missing from the science is proof of a real biological mechanism that is activated by PM2.5 exposure. CASAC scientists were concerned that even in high levels of particulate matter exposure, no proof exists that our bodies react negatively to any of the hundreds of PM2.5 particles. Before the danger of PM2.5 can be known for sure, medical researchers must determine if an actual biological mechanism is activated after human exposure to the pollutant. Only through a detailed study of the biology behind PM2.5 exposure, can the real danger of a pollutant be determined.

It is also unclear whether the mortality from PM2.5 predicted by the EPA is not actually due to other environmental factors. Since PM2.5 is a very broad category made up of hundreds of different substances, it is hard to tell which of these particles could be doing harm. Without knowing exactly which particle is to blame, it is difficult for scientists to prove whether factors, such as weather or other pollutants, contribute to the low risk associated with PM2.5. Further, the problem of finding the particle responsible for possible mortality is exacerbated by the previously mentioned problem with monitoring. Since scientists are not even sure how much particulate matter is reaching human lungs, they cannot easily disprove the theory that another factor is causing the mortality showing up in the EPA’s science.

The comments of CASAC make it clear that without understanding these basic physical and biological facts, we will never know for sure if one of the hundreds of PM2.5 particles is causing human mortality. As long as these facts remain unknown, the EPA will merely be guessing what benefits Americans will receive from the new PM2.5 air quality standard. The EPA probably knew this when it asked for money to fund further research into PM2.5. The president’s FY 1998 budget request acknowledged the dearth of PM2.5 research stating: “The budget proposes $26.4 million for PM research. To reduce the great uncertainty about PM’s health effects, EPA will continue its efforts to identify the [biological] mechanisms by which particles affect human health.” CASAC’s comments only reiterate this EPA statement, saying that extensive data analysis is necessary to determine for sure if a negative health effect from fine PM exposure exists.> In their own words. CASAC scientists have various backgrounds including epidemiology — the study of health risk associated with exposure to various chemicals — medicine, and chemistry. CASAC members have various affiliations ranging from universities and research institutions to industry. They are chosen both on the basis of their standing in their field and their diverse affiliations. Here is what these experts had to say about a few of the main issues regarding a new PM2.5 standard.5 Basic research is lacking to determine if health problems are caused by PM2.5. Some CASAC members clearly believe that the most important priority is to answer, through more research, the numerous questions remaining about PM2.5 Some believe that this evidence will not be found until years after the new standard is implemented.

Dr. Dan Menzel: “I have written to members of Congress, in particular to Senators Boxer and Feinstein of California, regarding my concerns over the lack of basic research being carried out in support of the particulate matter standard, and my grave concern that the support for not only the agency, but other federal research endeavors in the National Institutes of Health, for example, and research in this area is dreadful and results in the uncertainties that we see that exist in the data base regarding the health effects of particulate matter.” (CASAC meeting transcript, p. 4, 5/16/96)

Dr. Petros Koutrakis: “We do not know what the smoking gun is, and I think we should really be open to any kind of new information that is going to come[.]” (p. 117)

Dr. Allan Legge: “[M]y big, big problem with the whole 2.5 discussion is the fact that while there is a lot of data, I think, that is out there, I just do not think it has been analyzed adequately.” (p. 171 & 172)

Dr. Mark Utell: “I guess the important reason, as I see it, to go forward with a new standard is not to reduce the body counts, because I still struggle understanding how this works, but to allow us to really begin to collect data and do the kinds of studies that one needs to have the monitoring and the measurements in place so that some years from now when we are looking back or we are looking forward, we actually have data to address the question more directly” (p. 180)

Dr. Roger McClellen: “[T]o the best of my knowledge, we do not have a single set of data in terms of a large urban population in which we could attempt to test the association between particle number [PM2.5 levels] and measures of morbidity and mortality. We do not have a single data set.” (p. 123)

Dr. Koutrakis: “The question right now is, everybody has the same question, what is the smoking gun? Is it the fine, is it the coarse, is it the ultrafines, the transition metals, the counts? We don’t really have an answer.” (p. 174)

Dr. Joe Mauderly: “I think we have to realize that these values [standard levels] can only be selected with the understanding that we cannot defend them very well, we cannot defend them with confidence, regardless of where we set the level. If they are too high, we can’t defend that we are appropriately protecting health. If they are too low, we can’t defend on the other side that the protection that is conferred is worth the effort, as we understand it. Now, we must also accept that while we can identify associations between ambient particulate material measured by area monitors and adverse health outcomes, we don’t know how those environmental concentrations correspond to the actual personal exposures of individuals who are expressing the effects. We don’t know that, we don’t whether we are talking about the effects of hot spots, or some superimposition of indoor air quality, or some other personal factors. So we really don’t know how closely these are linked.” (p. 195 & 196)

Dr. Menzel: “I mean, we are all unhappy because there isn’t enough data. I’m disappointed to hear that people think that there will never be enough data, unless there is an enforcement hammer hanging over the heads of people. So I would say that I would reflect the concerns of everyone here, that we really do need to have more information regarding the public health consequences of the continued exposure to the particulate load in which we all live. I hope that that extends to the point of chemical analysis and a particle characterization of public health effect.” (p. 199) A national PM standard will not address the pollution problems of differing regions throughout the United States. The differing chemical composition of PM throughout the country would make a national standard ineffective. CASAC scientists showed concern that a “one-size-fits-all” PM2.5 standard will not respond to the varying pollution levels in different regions. Since some areas of the country might have naturally high levels of particulate matter, some communities might find it very difficult to attain the national standard. Another concern is that the actual particles allegedly responsible for PM2.5 mortality might vary from region to region. If implementation of a PM2.5 standard means regulating virtually hundreds of different substances, it will be difficult to ensure that nationally mandated controls address the specific pollution problem in a particular area.

Dr. Jay Jacobson: “Well, I guess I would just refer to the phrase used by Roger McClellen, the one size fits all issue. I think that is a very, very difficult issue, and clearly what we are starting to do now, is starting to try and tailor the standard to fit more than just one … to fit all seasons of the year, all locations, all susceptible populations.” (p. 184)

Dr. Petros Koutrakis: “It is very puzzling, and very confusing, even to me, maybe I’m not too smart, that how a number you get here will reflect exposures in the East and the West. I am not convinced that just throwing 150ug here, 30ug there, it will mean the same thing for a citizen … who lives in California, and a person who lives in Massachusetts.” (p. 176)

Dr. James Price: “There has been a mention, and I think it is a very important point, of the question there are some real problems with the one size fits all.” (p. 186) Many CASAC members stated that the primary purpose of control measures would be for research. Some CASAC members believe that the most important reason to implement a new standard is to collect the necessary data on human exposure to PM2.5. Many members are obviously concerned that the necessary data will be collected only after a new PM2.5 standard is enacted.

Dr. Mark Utell: “Not knowing, frankly, what the health impacts are, I sure would not want to come down at the lower end of the standard, of the range. I think, in fact, we ought to be liberal in terms of coming up with an imaginary number here that tries to allow us to collect the information and to utilize the information, but not to just decide who is and who is not in compliance.” (p. 180)

Dr. Petros Koutrakis: “I think we should really proceed with the standard that is not going to put the whole country out of compliance, a standard that will allow us to go out there and collect data.” (p. 174)

Dr. George Wolff: “I feel embarrassed that I’m up here recommending a new PM2.5 standard for the reason of getting data, but that is what is compelling me to do that.” (p. 200)

Separate, research grade quality measurements will need to be undertaken above and beyond the control measurements. Data collected by a new PM2.5 regulation will not necessarily be useful for understanding whether or not PM2.5 regulation is necessary. According to Dr. McClellen and Dr. Mauderly, setting up the infrastructure to regulate just for the sake of data measurement is questionable because accuracy will be limited by the monitors used for regulatory compliance purposes. The data collected from the implementation of a PM rule may not be accurate enough to assess the exact exposure of populations to PM. Separate, more sophisticated data collection will have to be carried out. The millions invested in monitoring equipment by states and counties for regulatory compliance may be useless to scientists.

Dr. Roger McClellen: “But a very important part of this is breaking away from what I call under the regulatory lamp post. So that we do not get ourselves hung up so that the measurements we are making are those dictated by regulatory compliance, yes, there is an issue of research grade quality measurements.” (p. 123)

Dr. Mauderly: “We need better data than simply measuring PM10 and PM2.5 will generate by themselves. We need better characterization of airborne particles, both in terms of size, [and] chemistry, including ultra-fines. We need better health effects data in these same areas. We are talking about picking model areas in which to do intensive studies that are not conducted, data are not collected for regulatory purposes, but for research purposes, anticipating that we will join this discussion again.” (p. 194 & 195)

The EPA’s claim of scientific consensus on the need for a PM2.5 regulation is dubious. While a majority opinion was reached, with 19 out of the 21 members of the committee voting for the new standard, many of those voting for the proposal had serious reservations about the effectiveness, feasibility, and general validity of regulating fine particulate matter. Moreover, the majority only supported the need for a standard. There was little or no support for the particular standard chosen by the EPA.

Dr. Carl Shy: “I think, you know, the issue of causality is always brought up as something that we have not settled in this whole concern about PM mortality and morbidity effects. There is no doubt [about] a lack of scientific consensus on that issue.” (p. 112)

Dr. Menzel: “We cannot afford to have a major national ambient air quality standard fall into the arena of being scoffed at, because it will mean the demise of so much else that is important in public health. I’m really concerned over the issue. I am disappointed with the consensus opinion here that we will never get enough information about the 2.5 below fraction unless it is mandated by some regulatory activity.” (p. 198)

Dr. Warren White: “That is how it is being interpreted. It is not really a scientific question. I mean, it seems to me the way the standards are supposed to be set is … to minimize the total number of dead bodies at the end of the year without worrying about cost or feasibility. So my recommendation would be that I haven’t yet heard a good justification for a 24- hour average standard.” (p. 164 & 165)

Dr. Kinley Larntz: “So I have difficulty thinking of a reasonable enforceable daily standard. That is just my opinion.” (p. 172).

Dr. Koutrakis: “So my answer is yes, we need a fine particulate matter standard, but not within the numbers I saw there. Again, I’m not a epidemiologist, I cannot give you a number, but I think we should really be very careful. So I think we should really be prudent and make sure we proceed, but without making a mistake.” (p. 175)

Dr. Mauderly: “But I think that assuming linearity [existence of heath effects at low levels of PM exposure] down to background levels really, really defies biological understanding at this point. I think we still have substantial uncertainty regarding causality, although I think that it is reasonable to assume in our research directions and in our control strategies that there is a casual association. I think we have a tremendous amount of uncertainty regarding the mechanisms by which specific sub- classes of PM can be connected to cardiorespiratory deaths.” (p. 192)

Dr. Wolff: “Whether or not we pick 90 or 15 today, it is not going to make a difference what the air quality is going to be 5 years from now.” (p. 202)

What are the costs of PM2.5 regulation? Sadly, cost was the one aspect of the particulate matter issue the committee hardly discussed. The economic consequences of a new PM rule could not be considered by CASAC because the Clean Air Act states that cost shall not be considered when setting NAAQS. Instead, CASAC members were only required to determine if a new PM2.5 standard would provide any health benefits to Americans. Clearly, CASAC was very unsure what benefits Americans would receive from a new PM2.5 standard. If the committee members were forced to consider the price tags of the regulations they proposed, they might have been even less sure about advocating a new standard. Unfortunately, nobody really knows for sure what the exact costs will be for the new PM2.5 rule. The lack of data that made CASAC’s decision to support a new PM standard so difficult also makes it nearly impossible to approximate the cost to consumers. The EPA puts the annual cost of the PM2.5 rule at $6.3 billion. Other estimates put the figure at several times the EPA’s estimate. Whatever the monetary costs, many believe that the weight of implementation of the new standard will be so enormous that our lifestyles will be forced to change. One Department of Transportation official expressed concern on this very issue. Writing to Sally Katzen at the Office of Information and Regulatory Affairs on November 20, 1996, Frank E. Kruesi, Assistant Secretary for Transportation Policy stated: “Control measures needed to meet the standards could have significant economic impacts on industry, including previously unregulated businesses, and require lifestyle changes by a significant part of the U.S. population.” Conclusion. However many lifestyle changes and billions of dollars a PM2.5 regulation will cost, the EPA should stop putting the cart before the horse and slow down the rush to implement a new PM rule. Their own scientists say that more research needs to be done to solve basic, unanswered questions. To answer the fundamental questions about human exposure, mortality risk by region, and proof of a biological mechanism behind PM2.5 mortality, the EPA must first do the following: (1) Build a sophisticated, research-grade monitoring network, such as the type suggested by CASAC scientists, to determine the actual human exposure to PM2.5 in different regions; (2) After collecting the appropriate data on exposure, the EPA must reanalyze the data to determine if a significant risk of mortality is posed by PM2.5, and show the varying risks from region to region; (3) Finally, medical researchers must work to determine if a biological mechanism exists, proving that PM2.5 dangerously alters the body’s chemistry. Once this research is completed, scientists will be able to positively determine if fine particulate matter poses a significant health risk. Then, and only then, will the EPA administrator be able to make an informed decision on whether to regulate PM2.5. Despite claims of scientific consensus, the EPA is moving forward with the new standard based on policy judgments, not scientific research.

Ultimately, Dr. Shy of CASAC is correct in asserting that the issue of PM2.5 regulation must be determined by the social and political costs. Before the EPA moves to regulate, weighing these various costs, it must have scientific proof of health benefits. Knowing that the costs of regulation can be high, the EPA appoints CASAC to get the best possible scientific advice available. Since CASAC showed doubt in the May meetings, the EPA must work to help answer the questions of our nation’s best scientists. If the EPA can prove a causal link between PM2.5 and mortality, it must work to find the particulate “virus” and its “carrier,” or source, which might be causing the mortality. Then, it can move to discover a regulatory “cure.” Unfortunately, instead of looking for a regulatory silver bullet, the EPA wants to take a multi-billion dollar vacuum to our air, hoping it will improve public health. Neither the Clean Air Act nor the recent court order requires the EPA to make NAAQS more stringent. The EPA only has to review the current standards; therefore, the EPA has time to slow down and allow scientists to do the necessary research. A more sensible approach, as some scientists and politicians have recommended, would be to spend several million (not billion) dollars collecting the basic, research quality data needed to determine if a PM2.5 regulation will improve Americans’ health.

1 EPA Regulatory Impact Analysis for PM 2.5, p. 7-25.

2 Comments of Thomas D. Hopkins, Rochester Institute of Technology and Regulatory Analysis Program, Center For Study of Public Choice, George Mason University, March 10, 1997, p. 17.

3 “The Inside Story: A Guide To Indoor Air Quality,”, 3/11/97.

4 Ibid.

5 These quotes are taken from the transcripts of the CASAC meetings in May. The transcripts were made by County Court Reporters,Inc. The comments of the CASAC scientists are not paraphrased, but may not be exact in all cases.