Good science, bad science?

Martin Pigeon, of Corporate Europe Observatory, has written a response to this article

Across the EU, there are increasing calls for regulation and policy decisions to be based on scientific evidence. But recently a controversial debate has developed regarding the sources of that evidence – specifically, when it is linked to industry. 

Industry spends billions of euros each year on research, much of which goes on generating evidence to demonstrate to regulators that new products and technologies are safe. However, industry is increasingly criticised for paying for science that furthers its own economic interests, and policy makers and regulators are finding themselves under pressure to rely instead on so-called ‘independent science’.

The notion that science becomes biased if it is funded or influenced by industry is deeply offensive to industry scientists and the academics they work with. But more importantly, the idea that ‘independent science’, which in many cases does not comply with internationally accepted quality standards or experimental protocols, should necessarily represent a more reliable and legitimate source of evidence is alarming and potentially harmful.

As an example, last year the European Food Safety Authority (EFSA) developed a bee testing guidance document and published an opinion on risks associated with neonicotinoid seed treatments, which ultimately led to the commission imposing a ban on these treatments. However, in drafting those documents, and following its own policy on independent evidence, the EFSA systematically excluded dialogue not only with industry experts, but also public domain experts who had worked with industry, or had sat on expert panels alongside industry experts. And with their exclusion came the wholesale rejection of scientific evidence from honey bee field studies. 

The notion that science becomes biased if it is funded or influenced by industry is deeply offensive

Those studies are essential to understand how pesticides affect bee populations, but the experts who were consulted, through no fault of their own, had no apparent practical experience or expertise in conducting these technically and logistically challenging investigations. I find it hard to understand how the vital insights of experts I know, trust and admire were excluded because of perceived conflicts of interest, and I regret that we could end up with an experimental guidance document that is considered unusable by the very industry experts who will be required to comply with it.

This approach to safeguarding independence is analogous to excluding experts from Mercedes Benz or BMW from investigations into car safety. Or barring anyone who had ever worked for or with Boeing and Airbus from discussions on improving aircraft safety. Why should excluding the chemical industry be any different?

Beyond an obvious frustration, this situation also presents a worrying dilemma for industry. Should companies avoid working with academia, lest the association blemish academics and bar them from contributing to policy decisions and experimental guidance documents? Such separation would not only be regrettable, it would in many cases be incompatible with funding requirements for academics to collaborate with industry partners, or for public domain scientists to establish public–private partnerships. 

Industry and regulators’ reliance on quality assurance standards such as good laboratory practice (GLP) has also been criticised as a barrier preventing regulators from using independent science. The associated debate about good versus bad science is both regrettable and divisive, pitting private sector innovators, regulators and academic interests against each other.  

This debate is paradoxical for those involved in the late 1980s when GLP compliance was imposed on industry. Many of us perceived it as burdensome and restrictive, but we had to recognise it was being imposed for a legitimate reason. Once the dust had settled – and after significant investment in protocols, standard operating procedures and quality assurance systems – we found that our work continued and that it was better documented than before. We even became proud of what we had achieved, knowing it provided the best available basis for our science to be used in regulatory decision making.

Clearly, regulators and policy makers cannot allow their decisions to be biased by conflicts of interest. But exclusion is not the answer, especially considering the various potential conflicts of interest that may exist. 

For instance, the European non-governmental organisation (NGO) community tends to position itself as representing civil society, but in reality their business models are often heavily reliant on donations from foundations or funding from national and EU governmental bodies (the EU spent €4.49 billion (£3.5 billion) funding NGOs in 2013). This may be justifiable, even desirable, but the notion that NGOs represent society and that government-funded scientists are independent, whereas industrial science is biased, is itself biased and inconsistent. 

The truth is that scientists tend to be paid by somebody, whether it is a company, a government, a research council, a consultancy or a foundation. We must move beyond superficial notions of independence to a more pragmatic, constructive and balanced approach. If we openly acknowledge all potential conflicts of interest, decision makers can evaluate the available evidence, rather than impairing their agency by denying themselves the wealth of public and private sector experience.

Two things should happen: first, robust criteria should be agreed for generating and assuring the quality of scientific evidence used in regulation or policy making. Second, rules for transparency should be established and applied equally to those who generate or use scientific evidence with the purpose of influencing policy, irrespective of their affiliation.

In this way, policy makers and regulators could have the highest confidence in the best available scientific evidence and expertise, with a clear understanding of potential vested interests, from wherever they originate.

Paul Leonard is head of innovation and technology policy at BASF

Martin Pigeon, of Corporate Europe Observatory, has written a response to this article