The Precautionary Principle

I’ve been asked to talk about the precautionary principle, and I thought I’d begin by explaining what it is. It’s a principle, not a statute, so there’s no canonical statement, but I think most people who have thought about it would accept the so-called Wingspread definition, named after the place where the meeting was held :

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. In this context the proponent of an activity, rather than the public, should bear the burden of proof.

In a similar vein, the European Commission has written:

The Precautionary Principle applies “where preliminary objective scientific evaluation indicates that there are reasonable grounds for concern …”

That all sounds reasonable enough, and it is. There have, however, been many attacks on the precautionary principle, many of them, not surprisingly, from people and organisations that stand to lose from it.

The most common criticism is that the principle is anti-scientific, whereas, as we can see from the definitions, it is based on science. It doesn’t come into play until there is scientific evidence of danger, and it generally requires that more science be carried out to resolve the issue one way or the other.

We’re told that it’s vacuous because it doesn’t lead to definite decisions. It’s too weak to bother with because it does nothing that standard risk assessment doesn’t do. Others tell us that it is so strong that it would stop all progress in its tracks. We’re told it’s merely a cover for protectionism, which is a charge laid against all forms of regulation by those who find them inconvenient, and we’re told the issues are better dealt with by the courts, when in fact the precautionary principle ought to be applied in the courts - it’s not an alternative.

The precautionary principle does not lead directly to decisions because it is not an algorithm for taking decisions. It’s about the burden of proof. In the criminal court, the defendant does not have to prove he is innocent. It is for the prosecution to prove he is guilty. They do not, of course, have to prove it in the sense of a mathematical theorem. The jury are to convict if they are convinced, not absolutely, which is pretty much impossible, but “beyond reasonable doubt.” The judge will instruct them about the burden of proof, but it is still for the jury to weigh up the evidence and decide.

In the same way, applying the precautionary principle puts the onus on the proponent to prove that new technology is safe. But this has to be done beyond reasonable doubt, not in some unattainable absolute sense, and society or its representatives still have to weigh up the evidence and decide. And the burden of proof is on the proponents of new technology for the same reason it is on the prosecution in a criminal court: it’s because the consequences of being wrong in one direction are so much greater than in the other.

A simple example shows how the precautionary principle could work. If it had been current in Elizabethan times, it would not have stopped Sir Walter Raleigh from introducing tobacco to Europe. There was no scientific evidence that it is harmful

It was only after the Second War that the evidence began to appear, and then the precautionary principle would have made a difference. For instead of having to carry out long series of experiments until there was overwhelming evidence of harm, the government could have started to act much sooner. And a lot of lives would have been saved.

There are many real examples I could cite, but time is short, so here is just one. Asbestos was first mined in Canada in 1879. By 1898, Lucy Deane, one of the first Women Inspector of Factories in the UK reported that not only was it hazardous to those who worked with it, but that it was worse than other kinds of dust because the particles are sharp.

In 1917, despite more evidence from the UK and France, and experiments on rats, the Factory Department found there was insufficient evidence to call for action. Yet in the very next year, North American insurance companies were refusing to cover asbestos workers because of the dangers.

Over the next 50 years, the evidence accumulated that asbestos is dangerous not only to those in the industry but also to their relatives, to people living close to asbestos factories, and to people in buildings where asbestos is used. Despite that, while there were regulations about factories, it continued to be used well into the 1970s. It wasn’t until 1998, the centenary of Lucy Deane’s report, that the UK finally banned its use altogether.

Notice that throughout this whole story, the authorities were applying what we might call the antiprecautionary principle. A new product must be considered to be safe unless society can prove that it is not. That’s why it took a century to get rid of asbestos, why it has taken so long for there to be effective action against tobacco, and why the World Trade Organisation is backing the Americans’ attempts to force BST beef more or less literally down our throats.

Opponents of the precautionary principle often argue that it costs money to apply. So it often does, though they generally over-estimate the amount, chiefly by assuming that there is no substitute. It generally turns out that we can find substitutes if we set our minds to it, as we certainly have with asbestos.

Besides, health and safety generally do cost. It’s a cost we pay because we judge that not taking those measures will cost us even more. No one will ever know the price we have paid for our refusal to apply the precautionary principle in the case of asbestos. Even if we had the data, we can’t be certain when it would have been banned - that would have depended on the judgement of people at the time.

We can, however, get an idea of the magnitude of the cost from an analysis by the Dutch Ministry of Health and Social Security. They estimated that if asbestos had been banned in the Netherlands in 1965 when the mesothelioma evidence had been widely accepted, instead of in 1993, the country would have saved 34,000 victims and €18 billion in building and compensation costs. That's just for one rather small country and assuming action was taken well after the precautionary principle would have demanded it, but it indicates the vast scale of the problem.

And it’s not over yet. Because mesothelioma takes so long to develop, it is estimated that even now, when asbestos is no longer used in Europe, there are about another quarter of a million deaths to come. And that’s not counting deaths from lung cancer and asbestosis.

I’m not going to talk about current problems because I have no time and the other speakers will deal with them. But there’s no shortage of them, for example GM crops,
BST, nuclear energy, climate change (will it be abrupt?), nanotechnology and so on.

I’ll just finish by summing up my responses to the criticisms I mentioned earlier:

The precautionary principle is not an algorithm for taking decisions, but a statement that the burden of proof lies with the innovator, not the rest of society.

It requires more sound science, not less, because an innovator must provide solid evidence of safety, not vague assurances.

It is neither so weak as to be vacuous, nor so strong that it would halt progress. It is not an alternative to legal proceedings; it should be a part of them.

It may impose a cost, but the cost of putting things right afterwards can be orders of magnitude greater.

The cost is often overestimated because alternatives may exist or can be developed if resources are allocated for the purpose.

Further reading:

The precautionary principle is science based. Peter Saunders & Mae-Wan Ho. Science in Society No. 18 (Spring 2003) pp35-36.

Late lessons from early warnings - the precautionary principle 1896-2000 (ed P Harramoës et al) 209pp European Environment Agency ISBN 92-9167-323-4
Free: order from www.eea.eu.int (Very informative, and a surprisingly good read.)