LL2-09 digest — Ch9 The pesticide DBCP and male infertility#

EEA 2013, report pp. 203–214 (PDF 205–216). Authors: Eula Bingham, who as US Assistant Secretary of Labor for Occupational Safety and Health in 1977 headed OSHA during the rulemaking she describes (she discloses the role but does not reflect on it), and Celeste Monforton. Box 9.2 is by Maria Albin. There are no panels and no industry response.

Core story#

By 1958, Shell and Dow had rat data showing that the nematicide DBCP was absorbed through skin and lungs and shrank testes in a dose-dependent way (p. 204). Their consultant, Charles Hine, recommended keeping workplace air below 1 ppm and impermeable clothing where skin contact was likely. In discussions with the FDA, “the Shell representative” called these recommendations “impractical” (p. 205). The 1961 published study (Torkelson, Hine et al.) also said control should rely on instruments, not smell (p. 205).

DBCP was registered in 1961 with a recommended 1 ppm limit, set below the lowest dose tested (5 ppm). The label carried no reproductive warning and there was no medical surveillance. Regulators voiced doubts, but the USDA accepted Shell’s argument that DBCP’s smell (1.7 ppm, above the limit itself) was warning enough (p. 205).

In July 1977 workers compared notes at lunch, and their union arranged sperm counts (p. 204). NIOSH then found 9 of 13 production workers had no sperm (the other 4 had very low counts) at a plant with air levels of 0.29–0.43 ppm. At a Dow plant with air levels of 0.04–0.4 ppm, half of 106 workers were affected (p. 206). OSHA issued an emergency standard within about two months and a 1 ppb final rule in 1978. EPA suspended distribution in 1977 and made the prohibition permanent in 1979, except for Hawaiian pineapples until 1985 (pp. 206–207).

Export continued. Shell and Amvac kept selling abroad. When Dow moved to halt shipments, Standard Fruit (later Dole) threatened a breach-of-contract claim and settled by indemnifying Dow (p. 209). Plantation use is documented until at least 1985 in Central America and about 1986 in the Philippines, with English-only labels and inadequate protection (pp. 207–209). Compensation came decades later, varied widely (a 1997 settlement averaged USD 1,500 per worker; a 1992 Costa Rica one about USD 20,000 each) and remains contested (pp. 209–210). Box 9.2 recounts Dole’s 2009 suit against a Swedish documentary, later ruled a SLAPP (p. 208). DBCP persists in California groundwater, where it was found in 254 of 1,312 wells in 2010 (p. 210).

Authors’ lessons (pp. 210–211)#

• Exposure below the lowest dose tested was “mistakenly assumed to be safe”.
• Air monitoring missed skin uptake.
• Animal evidence was discounted.
• Independent public expertise (NIOSH) was needed.
• Human evidence came from “the lay and local knowledge of the workers and their wives”.
• Safety rested on “authoritative assertion but without evidence”.
• Labels were generic and not translated.
• Nothing was done about “effects” in animals until they became “adverse effects” in people.
• Animal findings led to no surveillance of exposed men until the workers themselves noticed.
• Warnings stayed in specialist journals and internal company communications.
• Manufacturer knowledge did not reach user companies.
• National standards were not carried into international ones.
• Early warnings about persistence in soil and water were not acted on for many years.
• Information emerged mainly through slow, costly litigation.

Recommendation: closer integration of scientific disciplines, government action and “precautionary approaches” (p. 211). The link the chapter draws to endocrine disruptors is mostly analogy about how warnings are handled; where the summary treats DBCP as evidence “supporting” concern about endocrine disruptors (p. 203), it goes further than the case shows.

Main mechanisms#

• The producers held the early knowledge, and the company’s representative set aside its consultant’s protective advice in dealings with the regulator.
• Ignorance was produced by omission: no low-dose studies, no surveillance, no measure of skin dose.
• Regulators deferred to reassurance.
• A non-significant result at 5 ppm may have been read as safety (my inference; the chapter does not make this link).
• Harm was detected by those harmed. Once a legible endpoint was measured, action came fast.
• A jurisdictional gap and contracts moved the hazard abroad.
• People who depended on DBCP recast its harm; the Peach Council said sterility “may be bad, it is not necessarily so” (p. 206).
• Law served as disclosure, compensation and suppression.
• Persistence created a liability lasting generations.

Transferable insights (strength)#

1. Producers know first. Whether that knowledge becomes protection depends on how the firm acts on it before regulators (pp. 204–205). Strong / suggestive.
2. A limit below the lowest tested dose is an assumption. Here human harm appeared in plants with air levels 12–125 times below the lowest animal dose, plus unmeasured skin uptake (pp. 205–206). Strong.
3. Non-significant results in small studies can be read as “no effect” (pp. 204–205). Suggestive (my inference).
4. Monitoring what is easy to measure, not how exposure actually happens, misses harm. The warning signal sat above its own threshold (pp. 204–206). Strong.
5. Without surveillance, the harmed find the harm, late (pp. 204–205, 211–212). Strong.
6. A legible endpoint, organised affected people and independent experts together can bring rapid action (pp. 204, 206). Strong for this case; suggestive as a rule.
7. Banning use at home but not manufacture and sale for use abroad leaves the hazard in place elsewhere, here for at least 8–9 more years (pp. 207–209, 211). Strong.
8. Indemnity deals shift liability, not hazard (p. 209). Moderate to suggestive (one undated, unsourced episode).
9. Those dependent on a product minimise its harms (pp. 206, 211). Suggestive.
10. Confident assertion without evidence, accepted by doubting regulators (pp. 205, 211). Strong.
11. Emergency or interim powers allow action under uncertainty (pp. 206–207). Moderate.
12. Accountability after the fact is slow, uneven and contested (pp. 208–211). Moderate.
13. Persistence creates long-tail liability (p. 210). Strong for presence; moderate for health significance.

Main caveats#

• Insider authorship.
• Partly adversarial sources. Some come from litigation or advocacy, including DBCP: The Legacy, lawyers’ reports and EWG. Several key passages carry no citation at all: the FDA–Shell–Dow discussions, the cancer bioassays, the Dow–Standard Fruit contract and every settlement figure.
• Scale of harm among plantation workers. “Allegedly” tens of thousands. The best field data are 25 of 28 Panamanian workers, with no comparison group or selection method described.
• Litigation context omitted (background, to verify). Fraud findings in 2009 and non-enforcement of Nicaraguan judgments are not mentioned. Box 9.2 gives Dole one clause.
• Uncited claim. The claim that high doses predict low-dose hazards (p. 207) is uncited and contested.
• Contested priority claim. “First clear example” (p. 203) may be contestable (background, unverified: Kepone workers, 1975).
• Missing analysis. No discussion of benefits, substitutes or costs.
• Internal inconsistencies.
• Registration year: 1961 vs 1964.
• First production: 1955 vs 1956.
• Final OSHA standard: 11 vs 17 March 1978.
• Sprayer studies from 1979–80 are cited as informing a 1977 decision.
• The November 1979 EPA action is a “permanent” prohibition in the text but a “Suspension order and notice of intent to cancel” in the references.
• The 1958 Lykken document is an “internal memorandum” in the text but a letter to an outside laboratory in the references.
• “No medical surveillance” (p. 205) vs a 1961 plant medical examination that skipped testicular function (Table 9.1).
• Case selection. DBCP is an unusually clean case, so it may overstate how easy it is to act on early warnings.