Underpinning these stories is a common theme: the increasing public disquiet about a government's ability to assess and manage risk.

The problem is partly one of trust. Opinion polls in the UK, for instance, repeatedly show that the public does not trust the government to take decisions on matters of safety. The pressure groups that campaign for tougher regulations are held in higher esteem than government or industry scientists.

It is also a measure of the difficulties faced by scientists in assessing the possible environmental and health risks of innovations.

Current risk assessment techniques are unable to cope with some complex problems. Moreover, the scientific understanding underpinning certain new technologies may be too crude to lead to confident risk assessments.

These difficulties in assessing risk have given rise to calls for greater use of the 'precautionary principle' to deal with safety hazards. As defined in the 1992 Rio Declaration on Environment and Development, this principle says: "Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation."

This principle has been incorporated in international environmental treaties ranging from climate change to biodiversity conservation. Its use is increasingly being extended beyond environmental issues to other safety concerns.

In April 1999, the EU Council of Ministers adopted a resolution urging the commission "to be in the future even more determined to be guided by the precautionary principle in preparing proposals for legislation and in its other consumer-related activities". But the growing use of the precautionary principle is highly controversial. Its opponents argue that it imposes excessive costs, hinders trade, stifles development of technology, and is insufficiently grounded in science.

Critics argue that the "safety-first" attitude to risk has spawned numerous regulations, which are costly and sometimes counter-productive. "While economic regulation has declined dramatically since the 1980s, 'social regulation' has increased substantially," according to William Stanbury, of the University of British Columbia, in a recent study by the Fraser Institute, a Canadian think-tank. He argues that the current approach to risk is haphazard, irrational, involves insufficient cost benefit analysis, and pays excessive regard to interest groups.

Critics also focus on the risk that the precautionary principle can be misused as a disguised barrier to trade. Some suggest that the long- running EU ban on imports of US beef from hormone-treated cattle is an example of how the precautionary principle can be used as a trade barrier.

The US strongly disputes the EU's argument that the hormones could pose health risks. But the commission argues that it is entitled to take the precautionary principle into account when complying with international agreements on trade. "Far from being a way of evading obligations arising from the World Trade Organisation agreements, the envisaged use of the precautionary principle complies with these obligations," it says.

The argument over hormone-treated beef highlights a crucial issue about the regulation of risks, according to UK scientists working for the Global Environmental

Change Programme, part of the Economic and Science Research Council. Writing in Nature, the international science journal, in November 1999, they argued that instead of requiring the EU to provide evidence to support its doubts about the growth hormones, the onus should be on the producers of the hormones to demonstrate their safety.

"WTO regulations and other international trade rules increasingly assume that new products are safe until proven otherwise: the burden of proof falls heavily on those who are worried about unforeseen or untested safety and environmental issues," they said.

"Assuming that products are safe until proven otherwise may lead to what can be described as 'soft disasters' - large-scale health and environmental problems that emerge slowly but at high cost to society.

Such disasters mostly occur because excessive faith has been placed upon limited data about the safety of a product or process, ignoring many possible eventualities where there is little or no information."

But the idea that the burden of proof about product safety should lie with the producer provokes a strong rebuttal from some commentators, who point out that it is theoretically impossible to prove that an action will cause no harm.

"The precautionary principle will block the development of any technology if there is the slightest theoretical possibility of harm. So it cannot be a valid rule for rational decisions," said Soren Holm and John Harris, of the Institute of Medicine, Law and Bioethics at the University of Manchester, in a letter to Nature in 1999.

The researchers argued that the precautionary principle is fundamentally flawed because it obliges regulators to give greater weight to evidence suggesting risk, over evidence that points in the opposite direction.

But champions of the precautionary principle defend it against the charge that it is unscientific. The perceived dichotomy between science and precaution is misleading, according to Edward Groth, of the Consumers Union of US. "At times, precautionary approaches, with their emphasis on what science does not know as well as what is known, may in fact require more rigorous science than risk assessment, which has been known to brush aside uncertainties in order to answer too narrowly-drawn questions," he says.

Indeed, the credibility of the precautionary principle depends on its being applied within a scientific context. In a recent position paper, the US administration said that although it supported a precautionary approach, "precaution must be exercised as part of a science-based approach to regulation and not as a substitute for such an approach".

As technological change places heavy demands on risk assessors, it is likely that the precautionary principle will be pressed into use more often. But if its application is driven by a political, rather than a scientific, agenda, its credibility will be left in doubt.