Science, Society, Sustainability
The ISIS website is archived by the British Library as UK national documentary heritage ISIS members area log in ISIS facebook page ISIS twitter page ISIS youtube channel ISIS vimeo channel

ISIS Report 4/4/03

The Precautionary Principle is Science-based

The precautionary principle is based on good science. Prof. Peter Saunders and Dr. Mae-Wan Ho look at a few of the many examples where scientific evidence has made a compelling case for the application of the precautionary principle.

A fully referenced version of this article is posted on ISIS Members’ website. Details here.


The precautionary principle is simply a statement that we should not go ahead with a new technology, or persist with an old one, unless we are convinced it is safe. This sounds such an obviously sensible idea that we might expect it to be accepted by almost everyone and without question. Yet many objections have been raised against it.

We are told it is nothing more than a statement that we should be careful, and so says nothing that’s not already accepted, while at the same time others argue precisely the opposite: that it is so powerful that applying it would stop progress dead in its tracks. We are told that it sanctifies unscientific prejudice when in fact it requires scientific evidence before it is applied and demands that good science be used in place of sweeping and unjustified assurances of safety. We are even told these matters should be left to the courts as if that were an alternative, whereas it is the courts themselves that should be applying the precautionary principle.

Statement of the precautionary principle

Most of those who support the precautionary principle would accept that it is well expressed by the Wingspread statement [1, 2]:

"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."

This immediately deals with two of the common objections raised. First, the principle does not support unscientific prejudice. To say that the potential hazards do not have to be fully established scientifically makes it clear that the principle is about cases where there is scientific evidence. The European Commission states this explicitly in its Communication on the Precautionary Principle [3], writing that it applies "where preliminary objective scientific evaluation indicates that there are reasonable grounds for concern …"

Second, the principle is about the burden of proof [4, 5]. It is not an algorithm for making decisions that dispenses with judgement any more than the legal principle that the burden of proof in a criminal trial lies with the prosecution makes it unnecessary to have a jury to consider the evidence and come to a decision. It is a part of decision-making, not a substitute for it.

Moreover, like the legal principle, the precautionary principle does not demand absolute proof. A jury is not supposed to convict only on the balance of probabilities - the standard used in civil actions - but it does not need absolute proof that the defendant is guilty. It must only be convinced "beyond reasonable doubt". And what constitutes reasonable doubt in any given situation is also a matter for the jury to decide. The precautionary principle would no more stop all technological progress than the principle of the burden of proof makes it impossible to obtain convictions in the criminal courts [4].

The precautionary principle and the courts

Many opponents of the precautionary principle argue that the issues it is meant to deal with are better decided in the courts. But the precautionary principle and the legal processes are not alternatives. They can, and should be, used together. The recent actions against the tobacco manufacturers succeeded only because of the evidence that the companies were aware of the dangers and did nothing about them. If the precautionary principle had been applied, then the companies’ liability would date from the much earlier time when the scientific evidence was suggesting that there could be a real danger. Many lives might have been saved, or at least many more smokers or their survivors would be eligible for compensation.

It does not make sense to argue that the precautionary principle can operate in regulation but not in the courts because legal liability has an effect similar to that of regulation. Instead of acting within explicit rules set down by governments, companies or individuals are influenced by their judgement of the consequences if things go wrong. Driving without insurance is against the law in most countries, but most of us are even more concerned about the very large damages we might have to pay if we caused a serious accident.

It is interesting to consider how the history of tobacco would have been affected if the precautionary principle had been applied throughout. Despite what its more vehement opponents may say, the principle would not have prevented Sir Walter Raleigh from introducing tobacco into England. It would have had no effect at all until about 60 years ago, because before then, there was no scientific evidence of harm.

Once there was good evidence that people who smoked were far more likely to develop lung cancer, however, the precautionary principle would have made a significant difference. Governments would not have felt obliged to wait until there was a known mechanism linking smoking and cancer – a "smoking gun", so to speak. They would have become involved much earlier, and the restrictions on tobacco advertising, the large increases in excise duties in countries such as the UK, and the bans on smoking in public places might have come into effect many years before they did. More individuals would have been aware of the risk they were taking, and would have given up smoking.

Thus not only are two major objections to the precautionary principle contradictory, they each fail separately. It would not have stopped smoking being introduced, but once there was scientific evidence of risk, application of the precautionary principle would have saved many lives.

The case of BST

In 1997, the European Union banned the import of products from cattle that have been treated with bovine somatotropin (BST), a hormone that, when given to cattle, increases milk yields by about 10%. The USA immediately appealed to the World Trade Organisation (WTO), claiming that the issue was not really one of safety at all. They argued that there was no known example of humans being affected by BST, and that the EU’s action was merely a device to close their markets to imports from the USA.

In its original decision, the WTO gave the EU a year to provide evidence of harm to humans. If they could not do this, the ban would have to be lifted. This is a clear example of how the precautionary principle can make a real difference, because had the principle been invoked, the WTO would have been very unlikely to make such a ruling. In fact, the WTO was applying what we might call the anti-precautionary principle: it is for society to show that something is dangerous, instead of requiring the perpetrator to show it is safe.

Now it is true that there is no known example of humans being affected by BST. But it does not follow that there is no danger. First of all, many harmful effects take a long time to become obvious. The harmful effects of tobacco, for example, became evident only after many years of smoking. Besides, even if BST is harmful to humans, it will be very difficult to establish this because there is no control group. The original work on lung cancer was possible only because there were people who smoked and others who did not, and the researchers knew which were which. That is not possible with something that everyone, apart from vegetarians, consumes.

There are, however, good scientific grounds for suspecting that BST might be harmful, not because it would act as a growth hormone in humans. It is largely destroyed by pasteurisation and doesn’t interact with the appropriate receptors in humans. It does, however, stimulate the production of "insulin-like growth factors", which are identical in cattle and in humans and survive pasteurisation. High levels of these are associated with a greater risk of cancer development, though it is not yet known whether they are a cause or merely a marker [6].

We also have to ask whether BST has other, so far unrecognised, effects. A common feature of hormones is that they are involved in more than their primary role. In particular, they often modify the effects of other hormones.

If the world were starving because of a shortage of milk, then we might weigh up the costs and the benefits and decide that, even using the precautionary principle, the best decision was to allow BST. But there is already a surplus of milk, so much so that, for example, the European Union has developed an elaborate system of quotas to reduce production.

The only benefits from using BST go to the companies that produce the hormone. Given that, the evidence is surely sufficient to convince us that it should not be used, and even more so that the WTO should not force it (more or less literally) down the throats of those that do not want to consume the products of cattle that have been treated with it.

In the event, the WTO backed off, and decided to postpone taking a decision on BST. The result is that the EU is allowed to maintain its ban, but at the same time no precedent has been set; presumably this was the intention. We await further developments.

Let’s look at two currents issues, climate change and genetic modification.

Climate change

Climate change may not appear to come under the scope of the precautionary principle because hardly anyone doubts that our planet is getting warmer, and that the chief cause of this is the burning of fossil fuels. The global climate will change. Areas that are now fertile will become dry. The polar ice caps will melt, the sea level will rise and flood a great deal of land that is now inhabited. Northern Europe may become much colder if the influx of fresh water into the North Atlantic stops the Gulf Stream.

That much is well established. There is, however, a further possibility. The Earth’s climate is a large, complex non-linear dynamical system, and it is well known that when such systems are perturbed, they can undergo changes that are big, abrupt or catastrophic, and, at least in the short term, irreversible [7].

We know such changes have occurred in the past on the Earth. We also know that we are at present perturbing the climate by causing a large increase of carbon dioxide in the atmosphere. This has already led to a significant warming and there seems hardly any doubt that the average global temperature will rise still further, with estimates ranging from 1.5C to 6C over the next century. What we do not know is whether we are about to trigger a much bigger, catastrophic increase.

The precautionary principle tells us that in balancing the damage that may result from global warming against the cost of keeping it under control (it is already too late to counter the effects of our actions in the last century), we should take into account the possibility that the increase in temperature may be considerably greater and more rapid than has been estimated, and if so, it will probably be very difficult to bring the temperature down again even by a drastic reduction in the emission of greenhouse gases.

Genetic modification

The issue of genetically modified organisms (GMOs) is an area that cries out for application of the precautionary principle, if only because so much damage can still be prevented at this stage [8].

The most commonly raised objection to the introduction of genetically modified crops is ecological, that the genes may spread to other species. That is indeed a danger; more than that, it has already happened. In Canada and the United States, the genes that make oil seed rape tolerant to herbicides have spread to crops and weeds, which end up tolerant to multiple herbicides. That makes the herbicides useless and the weeds harder to control than before.

But while the ecological problems are real, and have attracted the most attention, they are by no means the whole story. The technology itself is a cause for concern. To be sure, hardly anyone is likely to die immediately after eating GM food. Apart from acute toxins and allergens, any harmful effects are likely to appear only in the longer term. There is evidence that many of the Bt toxins engineered into GM crops as biopesticides are actual or potential allergens for human beings, and toxic to a wide range of beneficial species [9]. But it will be very hard to identify these and other effects by epidemiological studies because there is no control group.

We are often told that GM foods must be safe because Americans have been eating them for years. But if there have been harmful effects, how would we know? As in the case of BST, there is no control group. If all Americans are eating GM foods, none but the most immediate harmful effects are likely to be recognised.

There is evidence strongly suggesting that GMOs are hazardous. First, transgenic DNA is not, as is so often claimed, "just the same as natural breeding." It is different. For example, when researchers created mutants for herbicide tolerance both by genetic engineering and by conventional mutagenesis, they found that the transgenes were up to 30 times more likely to spread to wild-type plants [10]. The more rapid spreading of transgenes is a potential hazard in itself, but what is crucial here is the demonstration that the transgene was different. Genetic engineering is not merely reproducing what happens in nature, and it is creating new combinations of genes that have never existed.

Transgenic DNA can also be transferred (horizontally) to unrelated species, to bacteria in the soil or in the gut and to cells of all animals including humans [11]. When mice were fed viral or transgenic DNA, not only was the DNA not completely degraded in the gut (as we used to be assured it would be), it passed through the wall of the intestine into the blood stream and even became incorporated in the genome of some mouse cells [12]. When fed to pregnant mice, the foreign DNA was found in some cells of the foetuses and newborn, showing it had gone through the placenta [13].

The researchers raised concerns over the possibility that transgenic DNA integrated into human cells could result in mutations and trigger cancer, as we did [14, 15]. This prediction has sadly become reality in the first cancer cases identified among the handful of ‘successes’ in gene therapy at the end of 2002 [16]. These patients were exposed to transgenic DNA similar in construction to those in GM foods.

The technology by which many GMOs are made is inherently dangerous, also because it often involves the creation, directly or indirectly, of super-viruses [17], which, unlike most natural viruses, are capable of crossing species barriers.

Genetic engineering further relies on the assumption that the piece of DNA that is transferred from one organism into a totally different one – from a fish to a tomato, for example – will have precisely the same effect in the second organism that it did in the first, and no other. This flies in the face of our modern understanding of genetics and of developmental biology. Organisms are a lot more complicated than that. Molecular biologists have long since given up defining a gene in terms of a more or less contiguous stretch of DNA. This alone raises the question of what exactly it is that is transferred.

We have a long way to go before we understand how the genome works, except that it is remarkably fluid and dynamic as it responds to multiple levels of feedback from the environment, to maintain itself constant or to change as appropriate to ecological challenges [18]. That may make it an interesting time to be a biologist, but it also means that in genetic engineering we are playing with a system we do not understand.

What are the benefits? We are often told that we must push ahead with the technology because otherwise millions of people in the developing world will starve. But there is easily enough food to feed everyone, and the best estimates are that using only conventional crops that will remain the case for at least 25 years and probably far into the future as well [19]. If people are starving – and millions are – that is not because there is not enough food but because it is not getting to them.

The problem of hunger is a problem not of production but of distribution. And distribution is not helped if we shift from small scale, local farming, where food is produced by the people who need it, to large agri-business. Yet it is the latter that genetic modification is designed to promote. Monoculture increases susceptibility to disease and pests, whereas smaller scale bio-diverse farming practices can mitigate the problem to the point where there is no need even to consider genetic modification as a solution [20, 21, 22].

Genetic modification may offer the opportunity for improving crops at some future time. The precautionary principle does not rule this out, nor does it exclude properly contained research to develop new varieties. It does, however, require that we should not press ahead with commercial crops until we have carried out the research necessary to establish that the technology we are using is safe.


The precautionary principle is neither so weak that it is empty nor so strong that it would halt all progress in technology. Far from being unscientific, it is based on science and it generally requires that more good science, not less, be undertaken so that sweeping assurances of safety can be replaced by solid evidence. The principle does, however, place more of the responsibility for safety on those who stand to profit if the technology goes ahead, rather than on those who will have to bear the costs if things go wrong.

We have given only a few of the many examples where the precautionary principle based on good science, should be comprehensively applied to protect human health and the environment.

A fully referenced version of this article is posted on ISIS Members’ website. Details here.

Recommended Reading

sitemap | contact ISIS

© 1999-2016 The Institute of Science in Society