Link to an article.
This is probably where the TV news got the story.
Here's a link to the organization primarily responsible for distributing the information.
Here's a link to the paper written by the firm Abt, which carried out some of the research for CATF.
The analysis that Abt performed was to see what the future death rate from particulate matter (assumed to be primarily from diesel emissions) might be in 2010, and how expensive it will be to care for public health at that time. I read over parts of the paper with special attention to their methodology. What it comes down to is pure statistics and lots of guesses (albeit educated ones). Abt repeatedly states that their work in this matter is quite speculative, and the high number of variables and the assumptions that had to be made call to question the perceived accuracy of the final numbers.
Of course, CATF and the news media don't bother to discuss the (potentially massive) statistical error in the report. By no means am I saying the research is useless or "bad". In fact, Abt's research paper seems quite reasonably written, to me. Problem is, the people at CATF (and your friendly neighborhood news affiliat) took the numerical results of the study and plastered them out as "fact", without so much as a word about the fact that the numbers are not "truth", but rather "statistical possibilities".
The other research that appears to link death numbers directly to diesel exhaust appears to be based on statistics of statistical data from various environmental agencies including CARB and the EPA. If we look at the lung cancer statistics, for example, the EPA has refused to provide any evidence directly linking diesel particulates with lung cancer; only CARB has done so. And CARB simply says "3 in 10000 lung cancer cases are caused by diesel exhaust" - a figure based on an EPA study into the risk of cancer from diesel soot. I didn't take the time to read the EPA study (you could make a career reading all these papers).
In my *opinion*, this study suffers from the effects of compounding error. CATF has taken statistical data, with all its inherent error, and performed statistical analysis on it - complete with their own assumptions and introduced error. It appears people spent a lot of time on this research, and found exactly what they wanted to find in the end.
This is why you never never never not ever believe any statistic anyone quotes you unless you've taken the time to learn what's been assumed in the study.
Do I think diesel particulate matter is toxic? Yes. Do I think you can quantify its effects across the entire population of the united States? No.
Just my opinions. Flame away!
-Ryan
