The 2018 conference of the European Food Safety Authority (EFSA), taking place in Parma, Italy, opened on September 18, 2018, with a plenary session “Where science meets society: Putting risk assessment in context.” The session, chaired by Barbara Gallani of EFSA, addressed the complex “interplay between science, risk assessment and policy” and the concern that in a world where “values are becoming more influential than facts in shaping public opinion,” risk assessment science “is increasingly mistrusted.” Therefore, the session aimed to “explore approaches that can contribute to restoring the credibility of and trust in risk assessment, by placing it in a societal context.”
Silvio Funtowicz of the University of Bergen, Norway, explored “the role of facts and values (an instance of which is risk assessment and risk management)” in the ongoing crisis between ‘science’ and ‘society’ on issues related to “health, the environment, diet and lifestyles.” These issues are of particular importance to ‘society’ because of “their intimate relation to the human condition, and also the societal and personal stakes invested in them,” Funtowicz explained. He pointed out that, currently, “when science is deployed in the policy context, we are aware of the possibility that facts are uncertain, values in dispute, stakes high and decisions urgent.”
Kevin Elliott of the Michigan State University, U.S., argued that risk assessment science (and science more generally) should strive for a “value-management ideal,” rather than a “value-free ideal.” Elliott discussed numerous value judgements that have to be made in risk assessments (FPF reported). “Try[ing] to prevent ethical and social value considerations from influencing these judgements” would be “unrealistic,” he said, because, even if we respect the evidence at hand, there are still choices to be made while interpreting it. Moreover, efforts to remove value judgements could in fact “prevent needed reflection about values.” This, in turn, could “result in conflict and distrust.” Therefore, Elliott suggested instead, that “values in science and risk assessment should be managed.” This can be done by promoting “reflection, openness and engagement.” There should be more discussion about “how to communicate about value judgements,” and “how to make these value judgements responsibly.”
Philip Macdonald of the Canadian Food Inspection Agency addressed the issue of problem formulation and the role it plays in risk assessment as well as in the communication between risk assessors and risk managers. Problem formulation, more commonly called hazard identification, is important for establishing the “parameters of risk assessments,” as it “defines your assumptions and hypotheses.” In this regard, he argued that many risk assessors “are not completely aware of themselves and their values.” Furthermore, often a clear policy guidance is lacking, and “without policy guidance, risk assessors are left to make a series of small policy decisions throughout the risk assessment.” Further influenced by “the push towards big data and the capacity to generate prodigious amounts of data,” risk assessors could find themselves “driven by wrong curiosity” and going “down unproductive roads.” This in a sense that “more data does not necessarily mean more precision,” because what it really can do is “to lead you to a place where you have more questions.” Thus, it is important that risk assessors learn to separate the “need to know” from “nice to know.” Macdonald concluded that a failure to arrive at a clear problem formulation could compromise the whole risk assessment, as “you may fail to come up with a constructive package for risk managers.” He further called on the risk assessors “to avoid the ongoing ‘value-driven’ risk assessment” and said that “discussion between risk assessor and risk manager is critical to ensure risk assessments provide value to decision makers.”
The subsequent panel discussion included Klaus Berend of the European Commission (EC), Julie Girling of the European People’s Party, Bjorn Hansen of the European Chemicals Agency (ECHA), Bernhard Url of EFSA, and two representatives of non-governmental organizations, namely Monique Goyens of The European Consumer Organization (BEUC) and Jorgo Riss of Greenpeace. The discussion was moderated by Frederic Simon of news provider Euractiv.
The first discussion question focused on the factors that panelists see as the most important for the fit-for-purpose risk assessment.
Berend said that a clear problem formulation is important because vague formulations like “no harmful effects” are difficult to translate into concrete actions. He also called for providing more transparency regarding the results of risk assessments and said that it is important to clearly identify the potential benefits of technology in order to compare it to the risks.
Girling said that, as a politician, she is the person who “uses and misuses the information to make points,” but she is always transparently communicating her values in this process. She thinks that reliable science is very important, but has to admit that science, and especially the way the science is presented to her, is “not always reliable.” She also agrees that it is important to weigh risks against benefits. With regard to involving society, she sees a potential problem of bias there, because “you will only involve the people that do want to be involved.”
Goyens said that she would choose elements providing for more openness in society, thus calling for more transparency, but also for “more engagement and inclusiveness.”
Jorgo Riss outlined four criteria, namely the need to (1) use publicly available and peer-reviewed studies and refrain from using “secret science,” (2) recognize uncertainties and be explicit about data gaps, (3) ensure that there is no conflict of interest among the people making the assessments, and (4) ensure that all the breadth of evidence is covered, i.e., not only the studies provided by the applicant, but also from independent science. Following these four rules would allow organizations to “get more trust and establish more trustworthiness,” Riss concluded.
Hansen called for more clarity from the side of policy makers in communicating to scientists and risk assessors about “what questions do they actually want answered,” because it is “a big difference when the question is ‘Is there no risk?’ or ‘Could you find risk?’” He also commented that, while no one would consciously want to base their decision on ‘nonreliable science,’ the current situation is that “reliable science is a rarity.” With this he did not mean that studies themselves are unreliable, but that “the science we get is normally not sufficient to make a reliable judgement.” Therefore, being clear about what you know and what you do not know becomes important.
Url said that for him, important things include science that encompasses all the available evidence, involvement of society, and communication. He sees EFSA’s mission in providing “trusted science for safe food” to “enable decision makers to take decisions.”
The moderator summarized that “reliable science framed with clear policy goals” seems to be the most important component in “fit-for-purpose risk assessment.” The next question, therefore, addressed what these policy goals should be.
Goyens emphasized that discussion about policy goals should involve all stakeholders, and that there should be “courage to ask ‘nasty’ questions.”
Riss said that, in his opinion, policy formulations should be discussed on the political level, i.e., by the EC, while EFSA should make a conscious effort to refrain from any judgements in the assessments it provides. Url countered that “there are many issues that are value-based and should not be camouflaged as pure value-free science.”
Hansen touched upon the issue of two common types of chemical regulations, which might result in a confusion. In one approach, chemicals are regulated with thresholds where certain levels are determined to be safe, while in another approach, it is determined that a chemical should not be present at any levels. It is a policy decision that criteria are sometimes level-based and sometimes presence-based. These two approaches should be “clearly differentiated and articulated.” To this, Goyens commented that it is not always easy to determine safe levels for a chemical because of concerns about mixture effects, and Riss added that “chemicals often turn out where they are not expected to be.”
The next question addressed whether the recent transparency proposal for the general food law (FPF reported) is fit-for-purpose. Girling did not want to answer this question in order not to “preempt the decisions.” However, she commented that, while no one would object to more transparency, there could also be commercial arguments to “keep something less transparent.” Both Goyens and Riss welcomed the “proposal for EFSA to stop using secret studies.”
EFSA (2018). “Where science meets society: Putting risk assessment in context.” Plenary session outline
EFSA (September 18, 2018). “EFSA conference opens with call to harness the ‘power of collaboration.’”