I agree with the director of the European Food Safety Agency (EFSA) that the historical mistake of mandating industry to test their own products should not be repeated (B. Url Nature 553, 381; 2018). The agency’s own record badly shook public trust in 2010, for example, when it overrode 800 or so academic studies on the chemical bisphenol A and its risk to health and the environment — relying instead on the results of four industry-sponsored protocol studies (see go.nature.com/2fi1vcs).
Government-funded academic studies have no vested interest in manipulating data. The best take into account factors such as the cumulative risk of exposure (for example, to dozens of pesticide residues at the same time), do not assume safe levels of carcinogens, test experimentally rather than dismiss low-dose effects of chemicals, and incorporate epidemiology into their evaluations.
On the weedkiller glyphosate, the ‘Monsanto papers’ (see go.nature.com/2tfpbwy) indicate that the main problem has been the manipulation of data in industry-sponsored studies. In my view, the only role for industry is to contribute to the cost of safety-evaluation studies.
A full reform of EFSA panels would help to restore public trust in the agency’s work. For example, a stricter conflict-of-interest policy would promote independence and objectivity, and including more academic scientists would boost scientific insight and raise scientific standards.
Nature 555, 443 (2018)
Competing Financial Interests
H.M. is a member of staff at the Pesticide Action Network.