In March 2011 a 9.0 magnitude earthquake triggered a tsunami which led to the automatic shutdown of the Fukushima Daiichi nuclear reactor in Japan. While the reactor’s shutdown sequence started well, waves from the tsunami breached the nuclear plant’s seawall to flood and disable backup generators. Without the generators cooling the reactor’s nuclear core there was a meltdown and radioactive leakage. In this extreme series of events, what was the best way forward?
The Japanese government took action and required approximately 154,000 people to evacuate, the seemingly safe choice.
But was seemingly safe the right choice?
This was the application of the Precautionary Principle. The Precautionary Principle has been defined in many related ways but I’ll show just one definition here, from the 1998 Wingspread Statement on the Precautionary Principle.
“When an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically.”
That is, when you see a threat, avoid it even if you don’t know the impact of the avoidance.
Seems like generally good advice. But if you’ve been reading my posts for a while you already know that things are more complicated than that. Let’s look at an example of a study of the Fukushima disaster that claims the Precautionary Principle led to worse outcomes than otherwise.
Damned if you do
The combined earthquake and tsunami killed over ten thousand people, mainly through drowning and blunt trauma. This was a large disaster by any measure. In the midst of this situation, how should we act to minimize deaths, casualties, and other negative effects?
In their paper “Be Cautious with the Precautionary Principle: Evidence from Fukushima Daiichi Nuclear Accident” the authors claim that the effort to avoid the impact from the nuclear meltdown exceeded negative outcomes of the meltdown itself. That is, they claim that the government’s application of the Precautionary Principle led to more deaths than otherwise.
From their abstract:
“After the accident, all nuclear power stations ceased operation and nuclear power was replaced by fossil fuels, causing an exogenous increase in electricity prices. This increase led to a reduction in energy consumption, which caused an increase in mortality during very cold temperatures. We estimate that the increase in mortality from higher electricity prices outnumbers the mortality from the accident itself, suggesting the decision to cease nuclear production has contributed to more deaths than the accident itself.”
After Fukushima there was a 40% increase in electricity prices in parts of Japan more highly dependent on nuclear energy and even a 10% increase in areas with low dependence. The authors then calculated that these higher prices changed consumer behavior and so the “higher electricity prices resulted in at least an additional 1,280 deaths during 2011-2014,” in an area representing just 28% of Japan’s population. (Most Japanese homes use electricity rather than gas or kerosene for both heating and cooling.) The number of deaths from cold then was higher than that from the nuclear disaster.
Further to this we have the following measure of deaths from radiation and evacuation. “No deaths have yet to be directly attributable to radiation exposure, though projections estimate a cumulative 130 deaths. An estimated 1,232 deaths occurred as a result of the evacuation after the accident.”
Note the authors don’t try to project if the deaths from radiation would be higher if there were no evacuations. Also not studied is the longer-term impact in mortality from lower air quality resulting from burning more fossil fuels for energy, an effect seen in the US after the closure of the Three Mile Island nuclear plant after a nuclear meltdown, though one that had no serious effects other than reducing domestic demand for nuclear power.
It’s hard or impossible to know outcomes in many situations, especially when action must come quickly during a crisis. But the Fukushima Daiichi nuclear incident strikes me as one where it seemed safer to do something rather than nothing.
The authors spend little time on why governments apply the Precautionary Principle, noting only that the decisions are “based more on emotions and instincts than on reason and rationality” and “deaths from higher energy prices are largely unnoticed.” This is an area I hope they will develop more and where I differ in evaluating how to apply the principle.
My take on the Precautionary Principle is that when presented with an uncertain future where there is an approaching uncertain risk of large chaotic negative outcomes or one where there is more certain likelihood of stable negative outcomes, choose the latter. In the case of Fukushima, that would mean
- Comparing a unknown, but potential large death rate from the nuclear incident’s radiation, with no evacuation, and keeping nuclear power operational nationally, versus
- Deaths from hurriedly evacuating the infirm but removing (or reducing) deaths from exposure to radiation, followed by a future where nuclear power is limited and imported fossil fuels increase fuel prices and impact mortality.
If the Japanese government was guilty of misjudging this one it is because they failed to issue temporary subsidies for the higher fuel prices, the more predictable and slow-moving of the above threats to life.
The paper’s authors have a point that we should be aware that the Precautionary Principle may fail because of human biases. But perhaps the existence of these biases tells us that we should listen to them rather than reject them.
Precautionary Principle approaches are relevant to other near-term applications, including:
- We often value doing something over doing nothing. That can leave us in Fukushima situations where evacuations cause more deaths than doing nothing.
- The Precautionary Principle does not weigh trade-offs in decisions. This is one of the counterarguments to efforts on reversing climate change. Since reduction in carbon emissions comes at a cost, what other impact will climate change reduction attempts have?
- The Precautionary Principle attempts to weigh severity of outcome in decisions. This is one of arguments for action to reverse climate change. Since the belief is if the world crosses a climate threshold impact would be massive, be willing to incur the large costs.
- Autonomous Vehicle fleet roll outs. I’ve earlier written about the systemic risks in AV deployments and why few are talking about risks in this industry.
- Universal Basic Income. I’ve also written about plans for UBI a couple times. As a top-down government program it seems to be a plan to reduce the risk of automation. But what new risks does UBI create? Few discuss this.
- Who does the deciding and what are their incentives? Does a government that is judged to have done too little to avoid risk find less popular support in the future? So should we see more applications of the Precautionary Principle by democratic governments than autocratic ones?
In a serious incident, doing something which ends up badly is judged as more forgivable than doing nothing which ends up badly. Doing something looks like strength. Doing nothing looks like negligence.
The “rational vs emotional” portrayal by the authors above also doesn’t sit right with me. It tries to make a balanced equation, something like one death today is equal to 100 people having their life expectancy shortened by one year. I don’t think people make decisions in that way or value that style of decision-making.
Are there other reasons that the Precautionary Principle could be a good default approach? I’m sticking with it for now.
Consider
- The Precautionary Principle is pretty broad, as a principle must be. That also means that it can be applied in a range of ways.
- When the timing is near-term (the deciders face the consequences) are people more likely to choose the Precautionary Principle? When the timing is in the far future (the deciders will not be around to face consequences) are people less likely to choose the Precautionary Principle?
