The damage done, Part 1: A post-mortem on the Trump EPA’s assault on TSCA’s new chemicals program

With the announcement by EPA that it intends to reverse two of the most damaging policy changes the Trump EPA made to EPA’s reviews of new chemicals under the Toxic Substances Control Act (TSCA), there is hope that going forward EPA’s reviews will once again conform to TSCA’s requirements and better protect workers, consumers, the public and the environment.

Predictably, the chemical industry and its phalanx of law firms – who demanded and embraced the Trump EPA’s policy reversals – have been howling loudly, doing their best impressions of Chicken Little.  They predict huge backlogs and economic calamity of all sorts, including an end to American innovation, and their lawyers are already threatening legal action – a clever way to drum up business, no doubt.

But facts are stubborn things.

In this first post I’ll look at a few reasons why the industry’s new round of fear-mongering is not based in fact.  And in a second post I’ll look at the decisions on new chemicals made under the Trump EPA to shed more light on the real reason why industry is upset:  It just may have lost the inside track that yielded such high dividends in the form of flawed approvals of hundreds of new chemicals.  Or, as one prominent industry attorney bluntly said recently in a related context, “the good days are over, quite frankly.”

Debunking the return-of-the-backlog myth

Back in 2018, then-Administrator Scott Pruitt used the excuse of the backlog in new chemical reviews to upend EPA’s then-succeeding implementation of the TSCA amendments governing new chemicals.  Yes, a temporary backlog had developed because the TSCA amendments made major improvements that had to be immediately applied by EPA, without allowing any time for a transition.  But by the time Pruitt announced his illegal policy reversals and claimed they were necessary to address the backlog, EPA’s own data showed that the backlog had already nearly been eliminated by EPA career staff who were diligently implementing the program in a manner that conformed to the law.  Those same data showed that immediately following Pruitt’s imposition of changes, the backlog began to grow again.

EPA’s earlier success in addressing the initial backlog while legally implementing the law – before the Trump EPA upended things – bodes well for the Biden EPA to return to that state of affairs.  EPA’s announcement last week is a strong start to that process.

Debunking the stifling-innovation myth

This industry argument has been one of its favorite go-to’s for many years (see, e.g., here and here).  It goes like this:  “If EPA doesn’t quickly allow virtually every new chemical it reviews onto the market without restriction, American innovation will suffer.”  And industry goes further, to assert that TSCA itself calls on EPA to promote innovation and hence, that conducting rigorous new chemical reviews is somehow at odds with Congress’ intent under TSCA.

This spurious claim relies on the only reference to innovation in all of TSCA, in a list of the law’s policy intentions (section 2(b)(3)), same language in both the old and new law).  Industry players love to distort this provision, so let me quote it in its entirety (emphasis added):

(b) POLICY.—It is the policy of the United States that—

…

(3) authority over chemical substances and mixtures should be exercised in such a manner as not to impede unduly or create unnecessary economic barriers to technological innovation while fulfilling the primary purpose of this Act to assure that such innovation and commerce in such chemical substances and mixtures do not present an unreasonable risk of injury to health or the environment.

When citing this provision, industry typically does one of two things.  It either deceptively paraphrases the provision as saying that, under TSCA, EPA shouldn’t act in a manner that impedes innovation.  Or it selectively quotes the provision – deceptively lopping off the second half of it (the part I added emphasis to above).

But Congress made clear that TSCA’s primary purpose is to ensure that any innovations tied to chemicals do not present unreasonable risks – quite the opposite of industry’s assertion.

Given that the development and application of new chemicals are a clear source of innovation, how else did Congress intend for EPA to provide an assurance that innovation and commerce in chemicals do not present unreasonable risk other than through robust scrutiny of new chemicals prior to their commercialization?

Another reason to question the industry’s playing of the innovation card

You might think based on the industry’s and the last administration’s obsession with streamlining new chemical reviews and emphasizing speed to market over everything else that it really was motivated by its desire to get new chemicals into use as fast as possible.  The facts simply don’t bear that out.

I looked at chemicals that were cleared for market entry by either receiving a “not likely to present an unreasonable risk” determination by EPA, or being subject to a consent order signed by the chemical’s manufacturer and EPA.  I then asked this question:  For how many of these approved chemicals had their manufacturers submitted a Notice of Commencement (NOC), which is required under TSCA to be submitted within 30 days of a company starting to make or import a new chemical for commercial purposes?

If companies’ demands for fast EPA approvals were based on their need to get their new chemicals quickly onto the market, you would expect the they would have promptly filed a NOC for the vast majority of such chemicals.  Yet here are the facts:

• Since TSCA was reformed in 2016, fewer than half (45%) of new chemicals that got a “not likely” determination have been NOC’d. Perhaps counterintuitively, a larger fraction (60%) of new chemicals subject to consent orders have been NOC’d.
• If one looks only at the period since mid-2018 when the Trump EPA’s weakening policy changes took hold, nearly the same proportion (43%) of “not likely” chemicals have been NOC’d. An even smaller fraction of chemicals subject to consent orders (28%) have been NOC’d.

This phenomenon of many approved but un-NOC’d new chemicals has historically been the case even before TSCA reform.  The makers of a large fraction, on the order of half, of all new chemicals EPA reviews and approves for market entry decide for whatever reason not to start commercial manufacture.

Some have argued that the relatively low cost that companies incur to send their chemicals through the review process leads them to seek approval even of chemicals they have little likelihood of commercializing.  While those costs have gone up since TSCA reform because EPA raised the fees it charges (from $2,500 to $16,000) so that they come closer to reflecting the agency’s costs to review new chemicals, the fees are still relatively low; moreover, both in its 2018 fee rule, and again in the proposed revisions now pending, the Trump EPA stated it wanted to keep those fees low so as to “minimize the potential impact on innovation and competitive standing” (2018) or not “stifle economic development in the chemical industry” (2021).

The fact is that EPA spends scarce resources reviewing hundreds of new chemicals every year that their manufacturers are not serious about – and often not in any hurry about – commercializing.  And industry then uses any delays in those reviews to argue that the review process is too rigorous and demand that it be scaled back.

The announcement by EPA about improvements it is making to EPA’s reviews of new chemicals under the Toxic Substances Control Act (TSCA) indicated it will begin by reversing two of the most damaging policy changes the Trump EPA made to the program:

• EPA will cease avoiding issuance of the binding orders TSCA requires to address potential risk or insufficient information:
  “EPA will stop issuing determinations of ‘not likely to present an unreasonable risk’ based on the existence of proposed SNURs [Significant New Use Rules]. Rather than excluding reasonably foreseen conditions of use from EPA’s review of a new substance by means of a SNUR, Congress anticipated that EPA would review all conditions of use when making determinations on new chemicals and, where appropriate, issue orders to address potential risks. Going forward, when EPA’s review leads to a conclusion that one or more uses may present an unreasonable risk, or when EPA lacks the information needed to make a safety finding, the agency will issue an order to address those potential risks.”

• EPA will cease assuming workers are adequately protected from chemical exposures absent binding requirements on employers:
  “EPA now intends to ensure necessary protections for workers identified in its review of new chemicals through regulatory means. Where EPA identifies a potential unreasonable risk to workers that could be addressed with appropriate personal protective equipment (PPE) and hazard communication, EPA will no longer assume that workers are adequately protected under OSHA’s worker protection standards and updated Safety Data Sheets (SDS). Instead, EPA will identify the absence of worker safeguards as “reasonably foreseen” conditions of use, and mandate necessary protections through a TSCA section 5(e) order, as appropriate.”

If you want the details on what was wrong with these policies – legally, scientifically, and health-wise – see EDF’s comments submitted to the agency last year and a summary of them here.

It’s no accident that these two policies were prioritized for reversal.  As I discuss below, each had massive adverse impact on the rigor and outcome of EPA’s reviews of new chemicals.  The result was that the Trump EPA allowed many hundreds of new chemicals to enter commerce under no or insufficient conditions.  It did this by:  1) illegally restricting its review to only the intended uses of a new chemical selected by its maker, hence failing to follow TSCA’s mandate to identify and assess reasonably foreseen uses of the chemicals; and 2) dismissing significant risks to workers that its own reviews identified, despite TSCA’s heightened mandate to protect workers.

Orders became a rarity, and testing virtually non-existent

I noted in Part 1 of this series that early implementation of the TSCA amendments largely adhered to the law’s new requirements.  As anticipated and for very good reason, that led to EPA issuing many more orders imposing conditions on manufacturers of new chemicals than had been the case under the old law.  A good number of those orders came with testing requirements – something that was also called for under amended TSCA in light of how few notices submitted for new chemicals included health and environmental data; TSCA specifically requires EPA to issue orders where the available information is “insufficient to permit a reasoned evaluation of the health and environmental effects” of the new chemicals (TSCA section 5(b)(3)(B)).

Industry bristled at these changes, to say the least.  They took their complaints to then-Administrator Scott Pruitt and other industry-affiliated political appointees to demand change, and they got it.  While it took some time (until mid-2018) for the changes to show up in individual decisions, the effect was dramatic.  The chart below tracks EPA’s own statistics on the outcomes of its new chemicals reviews from May of 2017 (when EPA first began reporting the numbers) through March of this year.  (EPA updates these statistics periodically, replacing the previous ones; I have captured them over this whole period, however, and used them to create the chart.)

The summer 2018 inflection point is impossible to miss:  The issuance of orders flattened while “not likely” determinations soared – the inverse of what had transpired up to that point in time.

Let’s now zoom in to look in more detail at what happened after those policy changes took hold:

The bottom line:  Nearly three-quarters of all new chemicals reviewed since mid-2018 after Pruitt’s policy changes got “not likely” determinations that cleared them to enter commerce without any conditions.  That amounts to more than 425 new chemicals granted unfettered market access.

As for testing requirements, those would only come through an order; I can’t remember the last time I saw an order with any testing requirements.  In fact, the Trump EPA stopped publicly posting orders altogether a year ago, so even the 50 or so such orders it did issue since then can’t be checked to see what they do and do not require of the companies.  But if those we did have access to up to a year ago are any guide, the hidden ones won’t have testing requirements either.

Finally, the rigor of orders themselves appears to have been compromised by the Trump EPA:  It colluded in secret with industry trade associations and law firms to weaken them, as we discovered through documents we obtained through a Freedom of Information Act request last month.

SNURs didn’t come close to keeping up with “not likely” determinations

Under the Trump EPA’s legally flawed approach, which the Biden EPA is now reversing, EPA sought to rely on its issuance of “Significant New Use Rules” in order to:  limit its review of a new chemical to only those uses intended by the manufacturer; on that basis issue a “not likely” determination; and defer any consideration of reasonably foreseen uses of the chemical to some possible future review divorced from the first.

Initially, EPA leadership insisted that the “not likely” finding would be made only once a final SNUR had been promulgated.  That then slipped to have issuance of the finding coincide with the proposal of the SNUR.  That then slipped further to allow the finding to be issued based on EPA’s mere intent to develop a SNUR.

So how did that go?  Since mid-2018, SNURs have been proposed for fewer than half (48%) of the new chemicals getting “not likely” determinations.  And for only 19% have the SNURs actually been finalized.

As to when the SNURs were issued, for only 12 of 206 “not likely” new chemicals with proposed SNURs was the SNUR proposed before the chemical was approved.  For those with proposed SNURs coming after approval, they didn’t get proposed until, on average, 90 days after the chemical was approved.  And not a single one of the 19% of “not likely” chemicals with a final SNUR had that SNUR in place before the chemical was approved without condition; in fact, those final SNURs lagged approvals by an average of 10 months.

So much for the Trump EPA’s plan to make its “not likely” determinations dependent on SNURs.

EPA cleared hundreds of new chemicals that by its own analysis present risk to workers

One of the most egregious acts by the Trump EPA under TSCA was its utter disregard for the elevated risks workers face from chemical exposures.  Embracing industry’s illegal demands that EPA stay out of workplaces – even though TSCA specifically identifies workers as warranting special protection as a vulnerable subpopulation – the Trump EPA systematically dismissed even very high worker risks its own analysis identified.

To do so, EPA made assumptions about worker protections that 1) were not supported by any actual evidence; 2) grossly distorted the regulations and authorities of the Occupational Safety and Health Administration (OSHA); and 3) are wholly inconsistent with the longstanding Industrial Hygiene Hierarchy of Controls (HOC) embodied in both OSHA policy and that of the National Institute for Occupational Safety and Health.  The HOC emphasizes removing chemical hazards from workplaces as the first and best worker risk management option.  (EPA instead chose to focus on the HOC’s option of last resort: assuming workers would always don fully effective personal protective equipment (PPE) even absent any requirement for their employers to provide it.)

How extensively did EPA rely on this nefarious approach?  Of the 425 “not likely” determinations the Trump EPA issued since Pruitt’s damaging policies took hold, nearly 80% identified some risk to workers – but then dismissed it by assuming workers would and could protect themselves by using PPE.  Ironically, by so doing, EPA eviscerated its only opportunity to actually impose through an order workplace controls sufficient to mitigate the risk.

While we are pleased to see EPA’s decision to reverse this policy, it is vital that EPA adhere to the HOC in addressing workplace risks and not merely make PPE use mandatory.

I blogged last summer about the magnitude of the workplace risks EPA found and dismissed.  In this updated table, I have catalogued the exceedances of the Trump EPA’s own risk benchmarks in its 80 most recent “not likely” decisions, made since June 2020, including those made since my previous post.  Here are the overall findings:

• For 41 of the 80 cases, the risks to workers EPA identified exceeded its own benchmarks for dermal or inhalation exposures, or both.
  • In 36 cases, EPA’s dermal risk benchmark was exceeded.
  • In 16 cases, EPA’s inhalation risk benchmark was exceeded.
  • In 11 cases, EPA’s dermal and inhalation risk benchmarks were both exceeded.
• The magnitude of the exceedances was as follows:
  • The median “fold factor” (i.e., magnitude of exceedance) across the 36 dermal exceedances was 15-fold, ranging from 1.1 to 31,000.
  • The median “fold factor” across the 16 inhalation exceedances was 4-fold, ranging from 1.0 to 20.
• For 53 of the 80 cases, EPA identified additional hazards the chemicals pose but it did not quantify the risk “due to a lack of dose-response for these hazards.”

This extent and magnitude of risks to workers that were routinely dismissed through false assumptions should serve as a staggering indictment of the Trump EPA’s willingness to elevate private industry interests over the agency’s mission.

In sum, a lot of damage was done over the last four years, and it is incumbent on the Biden EPA to act aggressively to ensure the irresponsible policies of the last administration that caused the damage are undone.

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