I. From medical risk to environmental, product-related, terrorism-related, and climate-related risk

In the 1980s, in medical school, I first became aware of the concept of risk regulation. Of course, virtually all regular medical procedures involve risk. In medicine, risk regulation was predominantly self-regulation that proceeded through professional standards, which have to be learned by new entrants to the profession. The concept of iatrogenic risk required a specific approach, since this kind of risk was not necessarily self-evident to physicians, as their well-intended acts were the cause. There was also government regulation of risks associated with the unauthorized practice of medicine, ethical issues such as abortion, and several other areas. Practice guidelines emerged to improve health care and reduce cost. These guidelines created risks for individual patients, however, as some could not receive the best available therapy on cost grounds.

As I had ventured into law in the evenings and weekends and had obtained a law degree before I passed the general physician exams, I was a natural for medical law. At the time, the Dutch government was reviewing the regulation of clinical trials and medical experiments. This topic was of great interest to me, since I had witnessed firsthand how medical doctors, health care institutions and the industry struggled with balancing the various interests involved. I conducted a research project for the Dutch government on this subject, and did my doctoral thesis on the regulation of medical experiments and liability for harm resulting therefrom.Footnote ¹ In this context, risk management was a matter of, on the one hand, ensuring the risks were proportional to the potential benefits, and, on the other, requiring that the human subjects were provided with adequate risk-related information so that they could decide whether to give informed consent. As a member of an institutional review board (or medical-ethical review committee), I learned how risk-related issues present themselves in actual clinical trials in many different guises. I realized that theory and practice need each other; a good theory is practical, but good practice informs or even induces the theory.

When I entered the practice of law, the focus of my work shifted from the medical area to environment, health and safety. A main difference with medical risk was that decisions about exposure are often made collectively, not individually. In medical practice and clinical trials, each individual gets to decide whether the risks are acceptable; in environmental, health and safety regulation, exposure decisions are made collectively, and it is hard, if not impossible, for any individual to reduce exposure, except, in some cases, by moving out of an area or out of a job. This is not always the case, however. In some situations, including lifestyle-related risks such as smoking or eating habits, collective decision-making about exposure is not the preferred way and resort is had to instruments like informational regulation or, to use a more popular term, “nudging”.Footnote ²

Environmental risk has never been addressed only by regulation, however. Liability for environmental harm was seen as a supplement and, in some cases, as an alternative to regulation. As I first learned at Yale Law School, if markets work perfectly and corporate assets cover the full amount of potential liabilities, liability can make regulation redundant. Liability can be created through administrative or civil law. Indeed, this choice was the chief political struggle in connection with the conception of the EU Environmental Liability Directive (ELD).Footnote ³ The Commission’s initial thinking was focused heavily on civil liability, but the EU finally settled on an administrative liability regime. While imposing liability for environmental damage broadly defined, the ELD additionally established a prevention regime. Through its deterrent financial effects, liability itself also has preventive effects, as it causes companies to take cost-effective prevention measures. This preventive effect of liability should be kept in mind when risks are considered for regulation.

As EU regulation moved away from a focus on risks relating to manufacturing to tackle systematically product-related environmental and health risks, my focus changed too. Inspired by the German “Toepfer-Verordnung”, the EU Packaging and Packaging Waste DirectiveFootnote ⁴ first introduced the concept of the “circular economy” into EU law. The concept sounded simple: get rid of the linear model of “make-use-dispose” and implement take back and reuse/recycling programs, but the reality was complex. Through life-cycle analyses (LCAs) regulators and companies tried to figure out how to reduce the environmental impact of packaging. This proved to be a battlefield, however, as the outcomes of LCAs depend very much on the assumptions that are made and the boundaries of the model.

In the area of EU chemical risk regulation (and, maybe, beyond as well), the REACH RegulationFootnote ⁵ is the crown jewel. Advertised as the world’s most ambitious chemical safety regulation, REACH is a comprehensive program that includes a series of loosely connected regulatory programs supplemented by a separate regulation on chemical classification and labeling. Although REACH established the European Chemicals Agency known as “ECHA”, it does not grant it regulatory authority; the Commission and the Member States exercise the regulatory powers under REACH. The combination of, by and large, stand-alone regulatory programs and Member State powers proved to be a recipe for political manoeuvering. If a Member State is not satisfied with the result of an assessment of a particular chemical risk it feels strongly about, it can attempt to launch another EU procedure to go after the same risk. The choice of regulatory instrument under REACH, i.e. authorization or restriction, continues to be a problem as of today, although a newly-created “risk management option” analysis tries to create more predictability around this issue.Footnote ⁶

In addition, I have been involved in areas of risk regulation that are not (or not substantially or directly) science-based. Of course, the introduction of the precautionary principle in the European treaties has redefined the role of science in risk regulation in a more fundamental way (see further under III., below). But here I refer to areas such as the risks to privacy as a result of personal data collection and processingFootnote ⁷ and terrorism-related risk. In a project comparing precaution in EU and US risk regulation led by Jonathan Wiener and Michael Rogers, I first had an opportunity to think systematically about these issues.Footnote ⁸ Terrorism-related risk was the subject of a study that we did for the European Commission.Footnote ⁹ The regulation of these kinds of risks are not predominantly a function of what science tells us about risks, but rather of judgments about the relative weight of other interests, “the kind of society we want to live in,” and values, emotions, and sentiments. In areas of risk regulation we consider science-based, we should not underestimate the role similar considerations may play. Biotechnology (GMOs) is a case in point; although entirely safe, environmental groups have been able to create an anti-biotechnology regulatory climate in Europe, raising their profile and funding in the course of doing so.

Possibly the most problematic area of risk regulation is climate change, the “mother of all risks”. Climate regulation appears to be science-based. Climate science is unique, however, in its strong and direct links with policy-making, which have resulted in the establishment of the IPCC and driven the science in a particular direction.Footnote ¹⁰ To borrow Curry’s words, “[t]he ‘policy cart’ was clearly leading the scientific ‘horse’”.Footnote ¹¹ There seems to be broad scientific consensus among climate scientists (the notorious “97%”), but upon closer examination this claim becomes blurry.Footnote ¹² In the area of climate change, the precautionary principle permeates the entire debate. On and off, I had been involved in specific aspects of climate policy-making for a decade or so, but took a strong interest in it when a Dutch court ordered the government to step up the fight against climate change.Footnote ¹³ The risks of politicization of the science and of the judiciary became focus points of my research in this area. In a series of publications, I expressed concerns about both climate science catering to policy makers and courts dictating ostensibly science-based policies to governments. In my view, these problems are not limited to climate change; an issue like endocrine disrupting chemicals shows similar characteristics (politicized science, high levels of NGO and media interest, and indecisive policy-makers).

II. Is there something we could call “risk regulation” and, if so, what are the boundaries of this area?

About five years ago, I learned about the concept of the “risk society”.Footnote ¹⁴ Intrigued by it, I read several books explaining and discussing the concept. As I learned more, however, I became more skeptical of its basic tenets. In an article published in the Journal of Risk Research, I explain why the concept of risk society is not an acceptable model for risk regulation.Footnote ¹⁵ The reasons I set out are that “risk society” discards the tools of rational risk regulation, and replaces them with post-modern devices, such as the equation of actual and perceived risks, an anti-industrial environmental protection ethos, and an activist attitude to regulation driven by the precautionary principle. Risk society thus creates risks of its own, including the risks associated with politicized science and decentralized decision-making in a direct democracy of sub-politics.

In the course of this work, I realized also that any policy issue can be framed as a risk. One could speak of the risks of poverty, of income inequality, illiteracy, migration, unemployment, discrimination, etc. Because both the active imposition of a risk and the absence of positive quality can be said to involve some risk, virtually any problem could be framed as a risk and presented as an issue of risk regulation; and there almost always is science available to support the plea. On the one hand, it could be useful to view the world through the prism of risk regulation,Footnote ¹⁶ but, on the other hand, it might deprive the term risk regulation of much of its usefulness.

The term “risk regulation” should be reserved for law- and policy-making on physical risks that can be studied by the natural sciences (biological and physical sciences). In other words, this area involves science-based regulation of physical risks to human health, safety, and the environment. Under this definition, psychological risks, e.g. the long-term mental problems associated with child abuse, would be included insofar as they also give rise to physical risks. Socio-economic risks, such as those arising from race discrimination or poverty, would generally not be covered, nor would the area of banking regulation, which is aimed at reducing financial risk. That does not mean that science is not relevant at all to these problems; science is relevant, but the requisite science is not natural science, and it generally plays a lesser, less definitive, or more explicitly political role. A program of increased government spending, for instance, could be defended with reference to Keynesian economic theory, but it generally understood that this kind of economics is tied to a particular political position.

Even under this somewhat restrictive definition, risk regulation research is a very broad field of enquiry. It can, and maybe should, be multi- and inter-disciplinary. It can include research on risk perception, risk communication, risk politics, risk management, and risk regulation, and address general (horizontal) issues, such as how science is used in regulatory processes and on what grounds paternalism in risk regulation can be justified, and specific (vertical) issues, such as the regulation of environmental exposures, aviation safety, foodstuffs, chemicals, tobacco control, and other life-style risks.

III. The future of risk regulation, and how facts, values and preferences can play their legitimate roles

The future of risk regulation is uncertain. The “low hanging fruit” has been harvested. As we start to regulate ever smaller, more remote risks, the benefits of regulation decrease, while the costs increase. Also in a post-trust, post-truth society in search of ever more security and safety, sound risk regulation hinges critically on reliable science, reasonable interest groups, robust procedures, and disciplined regulators. In my opinion, risk regulation itself will increasingly pose risks. The five main risks include inadequate prioritization, overregulation, “smorgasbordism” (i.e. the tendency to deploy several regulatory instruments to address the same risk), precautionary hazard-based regulation, and pseudo-science-based regulation.

While I do not think we live in a post-trust or post-truth society, trust and truth are always issues. This is no less so in risk regulation. Trust and truthfulness are properties of individuals, but they are also a function of the functioning of institutions and the structure of the system. If the institutions and the systems perform well, they should be able to generate both trust and truth. If this is true, we should focus on improving our risk regulatory systems. Sound science-based risk regulation, however, requires not only well-functioning regulators. It requires that the whole chain of institutional arrangements performs well, from the production of science, to scientific advice, to the making of rules. Rule-making includes processes such as transparency, risk assessment, cost-benefit analysis, and public consultation.Footnote ¹⁷ Risk (regulation) research can play an important role in improving the various links in the regulatory chain.

With media hypes and scares dominating the risk regulatory arena, “hyped”, but relatively insignificant risks may draw regulators’ attention and resources away from more significant risks. It is hard to control for this risk; better training of journalists might help, but the demand for sensational stories will likely continue to be met. Environmental groups, of course, know this, and play on these sentiments. The result often is inadequate risk prioritization, and overall higher risk levels (or less risk reduction for the resources invested in safety). The same over-reaction causes overregulation of some risks or “smorgasbordism”. Safe GMOs are subject to onerous prior risk assessment and approval regimes, as well as reporting and labelling requirements. Nudging, i.e. anything that influences personal choice on issues such as drinking soda or alcohol and eating fast food, does not necessarily require regulation, but often needs some sort of supporting regulatory framework. The risk here is that the nudge is simply added to a pre-existing, heavy direct command and control regulatory regime. In that case, nudging is just another incentive, and does not result in libertarian paternalism.

Science and precaution require special attention. The precautionary principle is intended to allow regulation to proceed where the science is uncertain, often boiling down to a requirement that science exclude the presence of risk. Under the generally accepted falsification approach, or any other model of science requiring the use of proper, testable hypotheses, however, science cannot exclude risk in general; it can merely confirm or deny the existence of a specific verifiable risk. Precautionary dogmas, such as the “Linear-Non-Threshold” (LNT) assumption for carcinogens, and the “phase-out-substitution” paradigm for hazardous substances, aggravate the situation. As a practical matter, the discussion often shifts to hazard, as opposed to risk. If risk is hazard plus exposure, hazard is theoretical risk. This opens the door to regulation in the absence of a showing of actual risk. Unfortunately, to make things worse, this tendency spills back into the practice of science, which will investigate hazard, rather than risk, a task it is better at in any event.

Through the precautionary vision of risk society, the role of science in regulation has substantially increased. Science is power, and thus can serve as an instrument for social control, raising the specter of politicized science. Where science is entangled in the “iron triangle” of “policy-making-science-valorization”,Footnote ¹⁸ it becomes impervious to critique and revision. The “iron triangle” involves the interaction between funding of science by politicians for policy ends, the production of often evocative science by scientists, and the valorization of this science by vested interests (either commercial interests or media, advocates, and spin doctors making alarmist declarations), often followed by more political initiatives, more funding, more science, and more valorization. In contentious areas such as climate change and endocrine disruption, the risk is that the scientists will cater to policy-makers’ demand for certain kinds of outcomes. I referred to this phenomenon as “scientific capture”.Footnote ¹⁹ In some cases, it is in and through battles in science that political battles are played out.Footnote ²⁰

It is easy to see why this creates a huge problem for society. The sole reason why science benefits from a privileged status, is that science produces objective knowledge. If it no longer produces such knowledge due to its inclusion in the “iron triangle”, it will lose its special status, and risks becoming just another opinion.Footnote ²¹ This is why we need to protect science against politicization and ideological bias. In doing so, we should not assume that the threats come only from outside science; there are strong indications that the ethos of science has changed; under the “counter-norms” of science, scientists are able to introduce ideological bias and advocacy into science.Footnote ²²

Risk (regulation) research’s most valuable use should be to expose “disguise and hide” strategies at all levels. Such strategic behaviour may occur in science, where scientists hide their value and subjective judgments behind ostensibly objective knowledge claims, at the science-policy advice stage, where advisors present value-laden advice as science-based logic, and at the political level, where politicians hide behind the science to pursue politically motivated programs (e.g., an ostensibly science-based climate program to achieve social justice objectives). Research and analysis can identify and unveil these devices.

In addition, risk (regulation) research is most useful if it also focuses on the application of regulations in the real world. There is a tendency to identify a problem, think up a regulatory solution to it, and then assume that this solution will somehow be perfectly and friction-free implemented, with no unintended side effects or intentional distortions. The reality of risk regulation is different and characterized by problems such as bureaucratic malfunctioning, mission creep, selective enforcement, and a host of other issues. To be able to overcome these challenges, risk regulation research itself must be free from ideological, political, and other forms of bias. For risk research, this will be just as difficult to achieve as for other areas of science. It is the only way, however, for risk regulation research to be useful to society. If risk research is able to simultaneously produce objective, policy-relevant knowledge based on analysis of the real world and avoid the threat of ideological bias and politicization, it may help to advance reasonable, balanced, and predictable regulation.

Finally, regulators should think twice before regulating the risks associated with new technologies. Innovation is important to any society, and innovation requires risk-taking. More generally, the often-heard argument that risk regulation stimulates innovation is self-serving, and places undue emphasis on industrial, “man-made” risks. Whenever a risk is proposed for regulation, the question should be asked whether we are better off without the proposed restrictions, or with a more modest regime. Not all risks can and should be regulated; some risks are well worth taking. We should always remember that risk regulation in the real world involves risks.