If you don't look before crossing roads you will, before long, be hit by a truck. Even if you do look but neglect to account for bends in the road, sooner or later you will still be hit by a truck.
Precaution is a simple lifesaving logic that some governments have attempted to codify as "the precautionary principle" to reduce the probability that new technologies and industrial products will later bite back. It is a logic, however, that many industries, scientists and (notably) the US government, vehemently oppose. Last month, however, came a salutary lesson in the value of looking at the science before one leaps.
In May, the journal Reproductive Toxicology published a paper that showed Canadian women now routinely have GM pesticides – called Bt toxins – present in their blood streams. So, too, do 80% of their unborn babies. Presumably, they acquired the toxins by eating GM corn or from livestock fed on it. By itself, this result does not prove that any harm has occurred – though it is hardly reassuring. And, as the first experiment of its kind, it needs repeating. What it does definitively prove, however, is equally important: the remarkable complacency of the global safety regulators of GM crops who have argued that this was impossible.
Finding Bt toxins in human blood highlights two particular regulatory decisions that illustrate this complacency. Regulators first decided to rely on a simple model of a human gut (a test tube, actually) to convince themselves that Bt toxins are completely broken down during digestion. They then chose to assume that this proved Bt toxins would never reach human organs. This led to their second crucial decision, which was not to require meaningful toxicological tests on Bt toxins.
The flaws in their reasoning stand starkly exposed by the Reproductive Toxicology paper. Not only have simple blood tests apparently demolished what regulators have long argued – that Bt toxins are not absorbed – but the paper also illustrates how easy it would have been for regulators to check their test-tube assumptions.
These failings are not merely a question of regulators disregarding the precautionary principle. What is rarely appreciated is that the precautionary principle is also good science. As Karl Popper first argued in the 1930s, science proceeds largely by testing hypotheses against data. Refusing to test a hypothesis was not entertained by Popper as a serious scientific choice. To to actively choose ignorance – which is effectively what these regulators have been doing – is neither good science nor good regulation.
Because regulators made the decisions they did, this new data raises potentially very serious and presently unanswerable safety questions about GM crops. No one, for example, can say if Bt toxins are likely to harm humans, since there is so little toxicological data available. Second, the findings open up a safety issue that regulators have always considered closed: that meat from animals fed GM plants might take up novel GM components and thus be different or unsafe. Third, the results may apply to other GM pesticides. If they do, human exposure to these toxins may be much higher than the authors observed. It is hard to imagine a finding more damaging to the credibility of GM regulatory institutions.
That regulators should repeatedly choose ignorance over both precaution (which is required by EU law in this case) and good science will probably surprise most readers. That is because there is a myth, much promoted by governments and industry, that scientific risk assessments typically fail only when events or science take an unforeseeable turn.
Detailed examinations of historical failures consistently show something very different. From thalidomide to BSE, oil rigs and nuclear power stations, failures result when regulators specifically and repeatedly choose not to know crucial facts. In the simplest possible terms, they cross roads regardless.
Jonathan Latham, Guardian