Dedicated to Putting an End to a Rogue State Agency

Join Our CARB Awareness Email List
For Email Marketing you can trust
Eve's Corner
Make Your Signs
Make Your Own T-Shirts
Tips for Truckers
The Gov
CARB Payroll
Who is CARB
Hien Tran Fraud
Reading Room
Links of Interest

The Science is Certainly Not Settled


"Oh My God! I am going to die of something"

On February 25, 2010 there was a symposium held to look at the integrity of the report titled: "Methodology for Estimating Premature Deaths Associated with Long-term Exposures to Fine Airborne Particulate Matter in California", upon which the Heavy Duty Truck and Bus regulations are based. These regulations have already hurt the trucking industry in California, and as additional regulations are phased in, promise more destruction to the industry--and to the economy of California. For those who are reading this information for the first time, A quick review of the events that lead to this point might be in order.

  • May 2008: The report was released listing "Dr." Hien Tran as the lead author.
  • June 23, 2008: Dr. John Dale Dunn comments on the report's inadequacy.
  • July 11, 2008: Dr. James Enstrom comments on errors contained in the Tran report.
  • July 7, 2008: Dr. S. Stanly Young writes to the Governor with concerns about the report.
  • November 4, 2008 CARB's reply to Dr. Young assuring him that Hein Tran is a qualified Ph.D.
  • December 10, 2008: Hien Tran confesses that he does not hold a valid PH.D.
  • December 11, 2008: Mary Nichols, Dr. John Balmes and another board members are informed of Tran's fraud and keep quiet.
  • December 11, 2008: Board Votes to pass the heavy duty truck and bus regulation.
  • December 18, 2008: Chris Reed of San Diego Tribune breaks the story of Tran fraud. Story picked up by others through 2009
  • April 2009: Tran suspended for 60 days and demoted but keeps his job at CARB.
  • September 24, 2009: Public comments at Board meeting are first time remaining board members hear of fraud.
  • November 19, 2009: Dr. Telles requests that the truck rule be set aside--it is not
  • November 19, 2009: Board votes to redo the Tran report--Mary Nichols promises an open process with everyone at the table.
  • February 26, 2010: Symposium is held.
The report in question is actually a compilation of studies by various entities.  There were two problems. First, Tran did not have the data upon which the studies were based and second, he decided which studies to include in the report. Some studies, which did not show results that CARB wanted were eliminated.  Those studies that agreed with what CARB wanted were included.  Thus the Board saw a report prepared by an unqualified lier that justified their vote. The peer review included some of the people whose studies were included in the Tran report.

Click video for full size
Now, at the symposium, many of the scientists who did the studies and the peer review were present to defend their work. Also at the table, were scientists whose work was ignored in the original report along with others who had concerns about the methods used in the included studies. Also in attendance at the Symposium were members of the public hurt by the CARB rules.  There were four Board members present: Mary Nichols, Ken Yeager, John Balmes & Sandra Berg.  Board staff included James Goldstine, Ellen Peter & Bart Croes. During the presentations by those scientists holding opposing views to the report, many of the CARB group played with their cellphones and were not paying attention.

Part I


Slides for the following presentations are available by clicking on [SLIDES].  More material is also available from the CARB website.

Dan Greenbaum gave a brief history of PM2.5 Studies. Parts 1 and 2 last 14 minutes.  [SLIDES]

Next, Mary Ross of EPA gave a presentation titled "Integrated Science Assessment for Particulate Mater lasting 15 Minutes.  [SLIDES]

Arden Pope III discussed his study which was one of the studies in the Tran report. Parts 1 and 2 last 15 minutes. [SLIDES]

Another study included in the Tran report was by Dan Krewski he talked about his study for 10 minutes. [SLIDES]

Aaron Cohen spoke on his project "The Global Burden of Disease from Air Pollution" he spoke on his method of risk assessment. His presentation lasted 20 minutes. It is interesting to note that in his second slide he shows that air pollution is much less risky than things like high blood pressure, tobacco, high cholesterol, diet, poor sanitation and even iron deficiency.  Yet the Pope study shows that air pollution is nearly as risky as the items at the top of the list. More on that in later discussion. [SLIDES]

There was a quick summery by Jonathan Samet on what has been discussed so far.

Following a brief break, a series of speakers had opposing views to the studies and the methods used to reach conclusions.  Dr. James Enstrom spoke first. His presentation spoke to the lack of qualified scientists at CARB.  [SLIDES]

Fred Lipfert discussed PM2.5 particle collection and how the particles differ from place to place.  He pointed out that no single source emits PM2.5, but rather PM2.5 particles come from many sources which can not be identified directly. [SLIDES]
Robert Phalen discussed why using the mass of a particle to determine risk is not a very reliable method.  He pointed out that attempting to eliminate a risk entirely can have side effects.  For example, he pointed out that regulations to eliminate PM2.5 particles, would hurt the economy in such a way that many people are pushed beneath the poverty level, which is a much greater risk than breathing PM2.5 particles.  He concluded that the science is not strong enough to make a connection between PM2.5 and mortality and certainly not strong enough to make regulations that will hurt the economy. 
The last presentation before the panel discussion was by Michael Jerrett. He discussed his study funded by CARB and two others which focused on California. [SLIDES]

Panel Discussion:

James Enstrom questioned Michael Jerrett's conclusions since his results nearly matched the results of the Enstrom study. Jerrett explained that his model did not look at total mortality but rather cause specific deaths. The discussion than turned to whether it was appropriate to eliminate some of the data and only pick data where the cause of death was related to pollution.
Tom Hesterburg sited a study that showed that risks at levels of 1.5 to 2 could be the results of other variables.  He also pointed that there are co-pollutants that are not even measured and this adds to the uncertainty of the results. Dan Krewski answered.
Suresh Moolgavkr agreed with Tom's point and pointed out flaws in the HEI study that have not been addressed.  He called the flaws an embarrassment of the PM2.5 study. Both Dan Krewski and Michael Jerrett answered which was followed by a summary by Jonathan Samet for the audience.  He said that the point of the discussion is how does one know which PM2.5 particles are the actual cause of mortality since PM2.5 is made up of many different particles.
The availability of the data became the issue for Rodger McClellan. Most of the studies were based on data owned by the American Cancer Society but ACS will not release the data to anyone to do their own study.  A discussion on data access and collection between Rodger and Dan Krewski followed.
Fred Lipfert had some comments for Dan Krewski about the study underway by one of his students (the study came up in the discussion between Suresh and Dan earlier). Fred suggests that unless she is working at local level, her results are doomed to failure since SO2 is a local pollutant.  He then questioned if Michael Jerrett's study might suffer from "Funding Bias".
Roger McClellan discussed the problem with focusing to closely on any one cause of death.  In the end he said, everyone dies of something and with so many variables it is impossible to extract any cause from all the noise. In any given life there are positive factors and negative factors that contribute to lifespan.  One of the most positive factor is that one has a job.
Suresh Moolgavkr commented that a particular model used for most of these studies is not adequate yet it us used everywhere. He was supported by Fref Lipfert who gave specific instances where the model fails.
Dr. Enstrom made a comment in support of Roger McClellan about the availability of data from the American Cancer Society which is used by HEI. He asked Dr. Greenbaum of HEI to pressure ACS to make the data available. Dr. Greenbaum responded that no one has asked for the data, but even if they did they did not have authority to release that data which is owned by ACS.
Arden Pope commented on the use of the ACS data.  He said that since the participants were aging, follow ups were showing less results.  This is because as people age, old age becomes more a factor in the cause of death.  He pointed out that even cigarette smoking becomes relatively less risky once you reach age 50. He admitted that he could not tell which parts of PM2.5 are the actual cause of the risk but there is risk associated with PM2.5.
Roger McClellan had a remark about relative risk following Arden Pope's comments. He asked if there is not a better way to present the data instead so that the model does not become the message. Michael Jerrett responded with his thoughts on smoking versus air pollution saying there is less effect on the general population by the high risk of smoking since few people smoke, on the other hand he said, more people breathe polluted air and while less risky more people are affected. Rodger said Michael was missing the point which annoyed Michael.  He finished up by reading a letter from the American Cancer Society on why the data would not be released.

The discussions between the scientists show that the science is not settled. Air pollution is certainly not a good thing, but is it worse than a large number of people in the population being unemployed? Comments from the panel members on both sides of the issue agree that falling below the poverty level has a greater affect on one's life span than air pollution. So far, the discussion has focused on the methods used to determine the risk of PM2.5 exposure and there is no agreement on what constitutes PM2.5 or its sources. It is also clear that Tran's cherry picking of the studies to be used in his report, presented the members of CARB with a biased view of the dangers of PM2.5. Of course some of the board members were willing to accept that biased report because it fit with their agenda.

Part II

The afternoon session started with focus on the topic "Which Studies are appropriate to use to estimate PM2.5 Mortality in California". The controversy here is because studies that show low risk from PM2.5 were and still are ignored by CARB, while studies that show moderate risk or very high risk are used to justify CARB's regulations.

Arden Pope's seven minute afternoon presentation concluded that there is not much difference between various locations studied, and that estimates from one part of the country hold true all over the country.  This view is not held by all present at the symposium.  [SLIDES]

(note the slides for this presentation follow the slides for Arden Popes first presentation)
Fred Lipfert showed that it is important to study the individual constituents in PM2.5 because PM2.5 composition can vary from place to place and is dependent of the method by which it is gathered.  Three studies showed inconsistent results using particle mass to determine risk.  [SLIDES]
Jonathan Samet took a moment to clarify some points in Fred Lipfert's presentation before starting the panel discussion on the topic "Which Studies are appropriate to use to estimate PM2.5 Mortality in California".
Roger McClellan comments on Arden Pope's presentation taking exception to his view that PM2.5 is the same everywhere. Arden Pope answered that his point was that people and emissions from traffic are the same across the country. He admitted that the risk from traffic related pollution is very small compared to other things but it was traffic pollution on which his study was focused.  He saw an equally small relative risk across the country and did not understand what components of PM2.5 was causing the slightly elevated risk.
Suresh Moolgavkar comments that the earlier presentation by Arden Pope and his slide that showed, as expected, smoking to be a much greater risk than PM2.5.  However in two studies (Ostro & Miller) PM2.5 is shown to be a greater risk than smoking and should be discarded because they clearly show bias.  Earlier Michael Jarrett suggested that these two California specific studies were valid. Arden Pope admitted that he agreed that some studies showed the PM2.5 risk to be higher than what they should be.
James Enstrom also disagrees with Pope's summery of the California data. Citing Jarrett's work, the California Teachers Cohort (Ostro) and the Adventists Study, which were anomalous, were also very small studies. Other larger studies show overwhelmingly that there is no current risk from PM2.5 in California. Jonathan Sammet recommended that CARB do a detailed analysis of the studies.
Tom Hesterberg points out that in one particular study which was acceptable to Arden Pope, the dose of PM2.5 was too high, the method of exposure was wrong and the conclusion reached wrong.  Arden Pope admitted that he was not a toxicologist and had to defer to Tom as the expert.  Arden concluded by saying that his point is not that one study is better than another but rather CARB should not use their own data alone and should include data from many sources.
Melanie Marty responds to Tom Hesterberg about the large number of studies that show correlation between heart and lung disease and PM pollution.  Tom responds that the dosage is always too high in those studies to which Melanie referred.
Roger Mcclellan did not want people listening to the presentations and discussions to think that air pollution in California is the same as every place else. He showed disappointment with Dan Krewski, Arden Pope and Jonathan Samet for not admitting that they had read the same studies that show that pollution in California is very different. Sue Paulson said that nobody believes that air pollution is the same everywhere but it has the same effect everywhere. Robert Phalen pointed out that the fine dust in California makes up much of what is considered pollution in California.
Fred Lipfirt commented that CARB needs to study traffic instead of wasting time on constituent studies.

George Thurston agreed with Pulson saying that the particles associated with health issues are from fossil fuel consumption. Roger Mcclellan had a strong opposing view and maintained his position that California is different from areas covered by the ACS report.
Jonathan Samet summarized the preceding round-table discussion and suggested that while there are plenty of studies that focus on California there needs to be more investigation by CARB to understand why they do not all agree.

The next topic in the afternoon session was titled: "Risk Assessment: Specification of the C-R function for long-term PM2.5 exposure-related mortality and the treatment of uncertainty".  Here the presentations focus on how models used to access risk are really based on estimates and that there is plenty of uncertainty in the conclusions. The round-table discussion shows that there is not agreement in some of the assumptions used in various studies. Those who use those assumptions, defend their work.

"Risk Assessment: Specification of the C-R function for long-term PM2.5 exposure-related mortality and the treatment of uncertainty"
Zachary Pekar's presentation "Risk Assessment: Specification of the C-R function for long-term PM2.5 exposure-related mortality and the treatment of uncertainty"
Suresh Moolgavkar opened his "Risk Assessment: Specification of the C-R function for long-term PM2.5 exposure-related mortality and the treatment of uncertainty" presentation by saying if you start your study with the assumption that PM2.5 is killing people, that is what you will find. His focus was on the uncertainties in the various studies that lead researchers to the wrong conclusion about PM2.5.  [SLIDES]     [TABLE OF RISKS]
"Risk Assessment: Specification of the C-R function for long-term PM2.5 exposure-related mortality and the treatment of uncertainty"
Round-Table Discussion
Following opening comments by Jonathan Samet, Fred Lipfert commented on the uncertainty introduced by the selection of the model. Michael Lipsett followed with more comments, defending the use of models that give biased results or that used bad data.
Dan Krewski and Michael Jerrett defend their studies despite the fact their studies show much greater risks than many other studies with larger groups and more data.
Suresh Moolgavkar noted that a paper co-authored by Dan Krewski which shows that the studies showing much greater risk from PM2.5 show that risk because of a non-linear concentration/response. He pointed out that for some reason, that paper is now being ignored.
Dan Krewski's quick presentation on "Expert Elicitation", a process by which guesswork is passed off as science.
George Thurston's presentation: "What studies are appropriate to use to estimate health impacts of specific sources such as diesel PM".    [SLIDES]
Tom Hesterberg's presentation: "What studies are appropriate to use to estimate health impacts of specific sources such as diesel PM".    [SLIDES]
"What studies are appropriate to use to estimate health impacts of specific sources such as diesel PM"
Round-Table Discussion
Fred Lipfert comments that conditions in Europe are much different than in the United States.  For example he points out that diesel pipes in Europe are near the ground unlike in the U.S. where the pipes are up in the air.  In Europe, the houses are not set back like they are here.  The bottom line is that there are a lot of apples and oranges to sort out before using studies from locations away from here.
James Enstrom suggested that CARB make changes to the designation of PM2.5 in light of the new information presented.

Roger Mcclellan comments that the regulators must "look to the future" and not be stuck in the past.  There has been so much improvement in air quality over the past years, but no one has answered the question "what is considered success".
Suresh Moolgavkar forced Arden Pope to admit that he does not understand why the numbers show active smoking is less dangerous than burning fossil fuel when it is clearly not so, he was cut off by John Samet.
Fred Lipfert comments on the 'circular argument' by Arden Pope.


Questions from the public and a final wrap-up by Dan Greenbaum.

Following the presentations and discussions, Mary Nichols attempts to explain why CARB is regulating PM2.5. She spent 8 minutes to say that PM2.5 is regulated by CARB because it is mandated by the federal EPA rather than health reasons.  She said that since PM2.5 was regulated from all other sources, diesel was the last place to regulate.  She continued to remark that it is out of CARB's control and that these regulations are all mandated elsewhere.  (see Tom Hesterberg's comments below which are a response to Mary Nichols' )
The following questions were answered or dismissed by Jonathan Samet: Have any of these studies demonstrated a risk of 2.0 or higher which is the standard used by federal courts for admissibility? What is a clear definition of premature mortality?  How is the carbon from tire tread dust separated from the carbon from diesel exhaust? Are sulfates toxic?  Can one death be shown to be the result of diesel particle emissions in California?
As engine technologies improve and particle emissions decline, will health effects decline into insignificance, and if so when? What are the different effects between old diesel fuel, upon which the data in the studies is based, and new diesel of today?  Tom Hesterberg, Dan Greenbaum, Roger McClellen and Melanie Marty comment.
Shouldn't we be monitoring more pollutants and which ones when these data are used? Robert Phalen, Melanie Marty, Jonathan Samet, Roger McClellan, George Thurston and Michael Jerrett comment.  There was a general feeling the the current method of monitoring was not sufficient.
Tom Hesterberg comment on Mary Nichols comments. He pointed out that the new regulations should be delayed because new studies are showing that the risk is not as great as when the regulations were made. Since most of the current diesel vehicles will be replaced with in a 10 year period using the current regulations and the replacement vehicles are much cleaner, the accelerated replacement in the new regulations are too much of an economic burden.
How do we know the relationship between exposure and risk?  Roger McClellen points out that there is no relationship between the models and reality.
A final summery of the PM2.5 symposium was given by Dan Greenbaulm