In a new essay for the Mercatus Bridge, I ask, “How Many Lives Are Lost Due to the Precautionary Principle?” The essay builds on two recent case studies of how the precautionary principle can result in unnecessary suffering and deaths. The first case study involves the Japanese government’s decision in 2011 to entirely abandon nuclear energy following the Fukushima Daiichi nuclear accident. The second involves Golden Rice, a form of rice that was genetically engineered to contain beta-carotene, which helps combat vitamin A deficiency. Anti-GMO resistance among environmental activists and regulatory officials held up the diffusion of this miracle food. New reports and books now document how these precautionary decisions diminished human welfare instead of improving it. I encourage you to jump over to the Bridge and read the entire story.
I concluded the essay by noting that, “It is time to reject the simplistic logic of the precautionary principle and move toward a more rational, balanced approach to the governance of technologies. Our lives and well-being depend upon it.” Some read that as a complete rejection of all preemptive regulation. I certainly was not arguing that, so let me clarify a few things.
There are, of course, “hard” and “soft” variants of the precautionary principle (PP). In my new essay, I am mostly focused on the very hardest variety (of a prohibitionary nature). They are the most concerning because they completely foreclose all future experimentation with new and better ways of doing things. In a section of my last book entitled, “When Does Precaution Make Sense?” I noted that outright bans on new goods and services are justified when the risk being evaluated can be shown to be highly probably, tangible, immediate, irreversible, and potentially catastrophic in nature. [See this essay for more on this point, including that entire section of my book reprinted as an appendix.]
However, “existential” risks are open to interpretation and far rarer than some suggest. Governments justly restrict the possession of uranium and bazookas and such grounds, but it would be imprudent to ban the development of all new AI technologies on the theory that one day we might get a Terminator scenario if we don’t.
Softer PP varieties of a permitting nature (such as FAA and FDA permitting regimes) are somewhat easier to justify because they at least leave the door open for some innovation, albeit after significant delay. It is impossible in advance to determine exactly how many lives are saved or lost because of long regulatory review processes, but some new products (such as large aircraft or pharmaceuticals) obviously deserve greater scrutiny because of the potential for adverse and catastrophic outcomes without some degree of initial oversight.
However, taken to the extreme and applied in too rigid of a fashion, even softer varieties of the PP can result in unnecessary suffering and deaths. Slowing experiments with potentially new and better ways of doing things means we are stuck with a status quo that can be sub-optimal, even deadly in its own right.
All roads lead back to improved benefit-cost analysis, better risk modeling, constant retrospective review, and stepped-up risk education/communication efforts. But the over-zealous and unthinking application of the PP shuts down that process almost entirely and forecloses any sort of policy or market experimentation. Flexibility, adaptability, and humility in policymaking are crucial to avoid policy errors.
Toward that end, as I noted in my last law review article, newer “soft law” governance tools offer us the chance to craft superior governance frameworks for existing and emerging technologies. Multistakeholder processes, agency guidances, collaborative best practices, and various other informal governance mechanisms are often better suited to address fast-moving sectors and technologies. In my next book, I argue that this is even true for many “existential risk” scenarios that people fear today. Preemptive controls – including some of a precautionary nature – will still be needed in many circumstances. (Genetic editing will be one such candidate). But we must still guard against overreaction and excessive control of technologies that have the potential to fundamentally improve human well-being.
In sum, trial-and-error is valuable both in the marketplace and in government policymaking settings. The fundamental problem with the precautionary principle is that is ends all such trial-and-error experimentation, including within regulatory regimes themselves! Greater flexibility is needed to ensure that public policy can more accurately balance risk and benefits and improve human well-being as a result. But the precautionary principle will almost never achieve that. We need more open, adaptive, and entrepreneurial governance mechanisms to achieve superior public health outcomes.
My next book, due out in April 2020, does a deeper dive into these issues. Stay tuned for more.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.