As long as I’ve been involved in federal regulatory issues (since 1990), regulators have used junk epidemiology to justify overregulation. For the first time ever, we have an administration that is committed to stopping overregulation. So JunkScience.com petitioned the Trump administration today to stop the misuse and abuse of epidemiology by issuing epidemiologic standards for federal agencies.
The petition is below.
###
February 20, 2018
President Donald J. Trump
The White House
1600 Pennsylvania Ave., N.W.
Washington, D.C. 20500
Re: Petition for Federal Standards to Stop Overregulation Based on Junk Epidemiology
Dear President Trump,
I am submitting this petition under the First Amendment right to petition the federal government to redress grievances. I request that you issue Executive branch-wide standards for the use of epidemiology studies by regulatory agencies.
An alternative request is that you direct regulatory agencies to issue their own such standards via public notice and comment. Pending the issuance of such standards, regulatory agencies should be ordered to suspend all use of epidemiology studies pending review under the new standards.
This petition is consistent with your initiative to reduce overregulation that hurts the economy without providing commensurate or even any benefit.
Just one example of the significance of the problem of junk epidemiology is President Obama’s key war-on-coal regulations issued by the U.S. Environmental Protection Agency (EPA). As you know, these rules were responsible for destroying about 94% of the market value of the coal industry and killing many thousands of coal industry jobs during the period 2011-2016 without providing any health, environmental or economic benefits whatsoever. The rules in question were “justified” on the basis of about $600 million worth of EPA-funded epidemiologic studies. These studies relied on secret data, and were either poorly or even fraudulently conducted and reviewed.
You justifiably complain about “fake news.” This petition would go a long way toward preventing the “fake science” that has been unjustifiably harming our economy and standard of living for decades.
Background
Epidemiology is the statistical study of the incidence of disease in human populations. Importantly, epidemiology is merely a branch of statistics; it is not science. Epidemiology does not provide biological or medical explanations (i.e., physical plausibility) for its purported results.
Epidemiology’s statistical nature is most useful when looking for high rates of rare disease in a population. The classic examples of properly applied epidemiology are food poisoning incidents and the link between heavy smoking and lung cancer.
Unfortunately, however, overzealous regulatory agencies have been disregarding the limitations of epidemiology for almost 30 years. They often pretend that epidemiology is a complete science, not merely statistics. They often improperly use epidemiology to study low rates of common diseases.
The data used in epidemiology studies is often of such poor quality that epidemiologists refuse to share their data with independent researchers for purposes of replicating and verifying results, a tradition fundamental to the scientific method. In the case of EPA’s war-on-coal rules, EPA-funded researchers have been hiding data from public review for more than 20 years ⎯ even defying the request of EPA’s own statutorily mandated science advisory board and Congressional subpoena for the data.
The abuse of epidemiology by federal regulatory agencies can be exemplified to laymen by comparing the number of deaths attributed to smoking against the number of deaths attributed to blue-sky clean air.
The Department of Health and Human Services claims that smoking kills about 440,000 people per year. But the Obama EPA claimed that fine particulate matter (soot and dust called “PM2.5”) in everyday blue-sky outdoor air kills 570,000 per year. So, smoking kills 440,000 while blue-sky outdoor air kills almost 30 percent more on an annual basis? One can easily understand why the EPA-funded epidemiologists have been hiding their data for 20-plus years.
Current Epidemiologic Standards in the Federal Government
The first effort to issue standards for interpreting epidemiology studies was articulated by famed British epidemiologist Sir Austin Bradford Hill in 1965. Hill almost uncannily foresaw the most common abuse of epidemiology we see today ⎯ i.e., inappropriate reliance on weak statistical correlations (also called “weak associations”) that likely reflect only poor data quality or chance, versus meaningful results.
The adage “correlation is not causation” should come to mind here. Not only is the adage true, but also weak correlations (or weak associations) never portend causation. Weak associations are just meaningless, statistical noise. There is not a single example in the scientific literature of a weak association epidemiology study whose reported association turned out to be scientifically valid.
The Obama EPA used this statistical noise to unjustifiably wreak havoc on the coal industry.
While Hill’s criteria do appear in some agency guidance documents concerning the use and interpretation of epidemiology, they uniformly omit Hill’s warning about the unreliability of weak associations. As a consequence, regulatory-happy federal agencies often disregard Hill’s standards and misinterpret statistical noise as cause-and-effect relationships in order to justify their (over)regulatory agendas.
Though the federal courts have received some guidance on the interpretation of epidemiology from the National Academy of Sciences and an international standards group (Grading of Recommendations Assessment, Development and Evaluation or “GRADE”) has issued some standards for interpreting epidemiology studies, federal regulatory agencies have remained oblivious and their misuse and abuse of epidemiology is ongoing.
Congress has also tried to rein in the abuse of epidemiology. The House-passed HONEST Act would require that epidemiologic data relied on by EPA be made available to the public for purposes of verification and study replication. Although the bill has passed the past three House sessions, it has been stranded in a Senate that requires 60 votes to pass a bill.
The Lack of Epidemiology Standards Threatens Efforts to Reduce Overregulation
It is a safe bet that virtually all epidemiology-based federal regulatory efforts over the past 25 years or so may be considered as “fake science” or “junk science.” This is because federal agencies, especially the EPA, have taken actions or issued warnings or regulations based on the statistical noise that is weak association epidemiology. This “fake science” should be held up to new robust federal epidemiology standards, and then validated or discarded based on its actual merits. Otherwise any deregulatory agenda is at severe risk of failure or rollback.
Consider the EPA’s proposed repeal of the Obama war-on-coal rule known as the Clean Power Plan (CPP). Although the CPP is ostensibly a rule addressing greenhouse gas emissions, the Obama EPA actually justified the rule on the basis that reduced coal plant greenhouse gas emissions would necessarily mean reduced emissions of the afore-mentioned PM2.5 from coal plants.
As the Obama EPA had determined (by secret science-based weak association epidemiology) that PM2.5 was associated with thousands of premature deaths annually (each valued by EPA via junk economics at about $9 million), the CPP was “determined” by the Obama EPA to provide billions of dollars in benefits annually ⎯ an imaginary amount of benefits that far exceeded the actual multi-billion estimated compliance costs of the CPP.
The Trump EPA has proposed to repeal the CPP the basis that PM2.5 causes no deaths at current levels ⎯ essentially ignoring the fake science of previous EPAs on PM2.5. This more realistic view of PM2.5 reduced the CPP’s estimated and imaginary benefits to well below its actual compliance costs.
Reducing the overregulation of all the PM2.5-dependent war-on-coal rules ⎯ including the Cross-State Air Pollution Rule and Mercury Air Transport Standard (MATS) ⎯ requires a review of the PM2.5 epidemiology under new standards. The Obama EPA’s onerous and benefit-less ozone air quality standards also depend on the PM2.5 fake science. It would be possible to reduce that rule’s expensive and pointless overregulation by reviewing its underlying science under sound principles and standards for epidemiology.
Conclusion
I have enclosed with this petition a copy of my recent book, “Scare Pollution: Why and How to Fix the EPA.” Please note that Sen. Jim Inhofe and Dr. George Wolff, a former chairman of the EPA’s Clean Air Act Scientific Advisory Committee, have both endorsed “Scare Pollution.” The book explains in more detail much of what is mentioned in this letter.
Epidemiology has been grossly abused by regulators and university researchers for so long, the vast majority of epidemiologists no longer care whether their work is charitably described as “garbage-in, garbage-out.”
That situation may be fine for agenda-driven regulators and their grant-hungry university epidemiologists, but it is a terribly destructive situation for the economy, taxpayers and science.
I am happy to answer any questions you may have.
Sincerely,
/s/
Steve Milloy, MHS, JD, LLM
Publisher
Trump EPA Transition Team member
Enclosure: Scare Pollution: Why and How to Fix the EPA