Enduring TSCA myths: Absence of evidence of harm = evidence of absence of harm

Richard Denison, Ph.D., is a Senior Scientist.

A subcommittee of the Senate Environment and Public Works Committee held a hearing on reform of the Toxic Substances Control Act (TSCA) last week.  It’s good to have attention drawn to the issue so early in the new Congress.

But – especially if you missed Senator Lautenberg’s opening statement and the testimony of EPA Assistant Administrator Steve Owens – you may well have come away from the hearing with the impression that TSCA is basically working quite well and really only needs a few tweaks, or what the chemical industry loves to call “modernization.”

Nothing could be further from the truth, of course.  This 35-year-old law is not only outdated, it’s proven ineffectual in myriad ways.  The false impression that all TSCA needs is a little polishing-up was, unfortunately, bolstered by the invoking of two particularly pernicious myths about TSCA by some of the other hearing witnesses.

The first myth, which I’ll deal with in this post, is that EPA’s new chemicals program has worked very well – so often repeated by some former EPA officials that it has become virtual dogma.

The second myth, which I’ll address in a subsequent post, is that we can confidently rely on existing information to identify all of the chemicals of concern on the market today, and safely set aside the rest. 

Myth #1:  The New Chemicals Program is working well

I can only explain the perpetuation of this myth as a rather desperate need to point to something pursuant to TSCA that has at least sustained a high level of activity in the toxics office of the Agency.

I have blogged extensively about the major shortcomings of the new chemicals authorities granted EPA under TSCA.  And a whole series of reports by the Government  Accountability Office and, most recently, a report by EPA’s own Office of the Inspector General (OIG), have also documented extensive problems.

I won’t repeat all of this extensive analysis again (see this post and this post for details), but will address one argument made by those claiming the new chemicals program has worked:  namely, that none of the new chemicals reviewed by EPA has later been shown to be a problem.

First, this argument begs the question:  How would we know if a problem arose later?  A bedrock problem of TSCA is that there is no mandate to assess the risks of chemicals once they are on the market.  Nor is there any mandate to monitor for them in workplaces, in the environment or in our own bodies.  Nor even to identify their presence in products.

Second, new chemicals are heavily cloaked in a mantle of secrecy, with the identities of most of them hidden from public view in almost all settings.  Most of them are not even specifically listed on the public version of the TSCA InventoryAs I’ve noted before, their identities are even hidden in the context of industry submissions of health and safety data, a problem EPA has only recently begun to seek to remedy.

This secrecy stymies efforts to uncover and link problems to new chemicals.  For example, one cannot reliably cross-reference lists of carcinogens or lists of chemicals of concern identified by various governments against a list of new chemicals under TSCA.  That’s because the vast majority of them are not listed by standard identifiers on the public version of the Inventory, only by generic name and a mystery for-EPA-internal-use accession number.

Third, for the vast majority of new chemicals EPA has reviewed, TSCA provides EPA with no means to “look back,” that is, no way even to consider, let alone investigate, whether such a chemical later poses any risks.

In a very small number of cases, EPA issues a Significant New Use Rule, or SNUR, for a new chemical; this has been done for at most 6% of all new chemicals that have been added to the TSCA Inventory since TSCA was adopted (see page 4 here).   Where a SNUR exists, a manufacturer of a chemical that plans to produce or import it in a manner inconsistent with the conditions of the SNUR must first notify EPA.

Hence, for the small subset of new chemicals subject to SNURs, they may provide a limited basis for a “look-back” – but only where the manufacturer plans to deviate from the conditions set forth in the SNUR.

But remember:  Even in these rare cases, mere notice is required.  No health or environmental data need be submitted by the manufacturer, so EPA’s ability to detect a problem is exceedingly limited – just as limited as it is with the original premanufacture notifications (PMNs) submitted prior to manufacture, as I’ve described at length in my earlier posts.

Fourth, about 30 times a year EPA does receive data for new chemicals that companies have developed or obtained and that the companies themselves believe “reasonably supports the conclusion that such substance or mixture presents a substantial risk of injury to health or the environment.”  (see page 13 here)

These so-called “substantial risk” notices are mandated by TSCA’s Section 8(e).  As I noted earlier, however, in almost all cases the identities of the chemicals in question have been masked by the submitter as confidential business information, which stymies any public effort to investigate these concerns further.

The fact that such notices are submitted for new chemicals even this often is really quite astounding, given that there is no routine requirement for companies to test their chemicals, and that, in order to require such testing, EPA must first show evidence of at least potential risk.

One can only imagine how much affirmative evidence of substantial risk posed by new chemicals after they enter the market might surface if there were a requirement under TSCA to actually look for such evidence.

Finally, as I’ve noted before, there is not a basis for presuming new chemicals are safe.  While the results of EPA’s new chemical reviews are held as top secret information, the European Union reports that its review of new chemicals over the years before REACH found that about 70% of them possessed at least one dangerous property (page 27).  That finding is based on the receipt by EU officials of actual test data for those new chemicals.  The EU and virtually all of the rest of the developed world require such data for new chemicals as a condition for market entry – in marked contrast to the U.S. under TSCA.

In summary, so much for our ability to have any confidence in the argument that none of the new chemicals reviewed by EPA has been shown later to be a problem.

Let me repeat here two telling conclusions of EPA’s own Inspector General on this matter:

“EPA’s assurance that new chemicals introduced into commerce do not pose unreasonable risks to workers, consumers, or the environment is not supported by data or actual testing.”


“In cases where full information does not exist or analyses are limited, EPA reports the new chemicals as not having risk.”  [And remember, 85% of new chemical notices contain no toxicity data whatsoever; see page 6 here.]

EPA’s new chemicals program directly embodies the larger problem inherent in TSCA:  It forces EPA falsely to equate the absence of evidence of harm with evidence of absence of harm.

What ever happened to evidence-based decision-making?

In my next post, I’ll take on that second pernicious myth:  that we can confidently rely on existing information to identify all of the chemicals of concern on the market today, and safely set aside the rest.

This entry was posted in Health Policy, Regulation and tagged , , . Bookmark the permalink. Both comments and trackbacks are currently closed.