The precautionary principle
"Genetically modified crops raise more questions than they answer. Insufficient
laboratory tests have been done on the effect of GM crops before going to
field-scale trials. The precautionary principle means that, if you are unsure
of what the result will be, especially if it is going to be so serious, it
would be wiser not to do it until more evidence is available."
RSPB spokesperson. Source: Belfast Telegraph, 4 June 2003
If you want to see a good idea become misunderstood and abused, give it a clear
and simple name. Very often, this is the cue for people to start mistaking the
clarity and simplicity of the nomenclature for clarity and simplicity in the
Such is the fate of "the precautionary principle". You can see it
being appealed to in all sorts of contexts, as though it were a straightforward,
obvious idea we can all understand and use. In fact, the principle is used in
a myriad of varying, vague and misleading ways.
In the example above, the principle is formulated as: Do not do anything where
the outcome is uncertain and potentially serious until more evidence is available.
The obvious problem here is that the outcome of almost all actions is uncertain,
and often potentially serious. Even a simple decision like driving a long distance
has an uncertain, potentially fatal outcome. (In the UK, road accidents are
the most common cause of death among the under 35s.) Are we then not to drive
"until more evidence is available"?
Obviously this is absurd (and not what was intended). The point is to have
enough evidence, not more. We do have enough evidence to make an informed choice
about driving and most of us judge that the risks are worth taking. In contrast,
the RSPB obviously thinks we don’t have enough evidence about GM crops to make
a decision about field trials. The principle we are then following should read:
Do not do anything where the outcome is uncertain and potentially serious until
enough evidence is available to make an informed choice.
The key question now becomes, "what is enough evidence?" "Enough
to know what the risks are" isn’t a particularly informative answer, since
it doesn’t add much to our concept of "an informed choice" to say
it is one where the risks are known. This answer can mean something substantive,
however, if it means "that which is sufficient to conduct a rigorous risk
assessment". Risk assessment is about making decisions based on the desirability
of outcomes weighted against the risks involved in the actions required to achieve
these outcomes. Of course, it is for expert risk assessors to determine just
what is enough evidence to conduct a risk analysis. But this is not a problem.
It is just another example of the division of intellectual labour, an acknowledgement
that a layperson is in no position to judge what precisely what constitutes
We can now state the precautionary principle in a coherent way: do not proceed
in a course of action with potentially serious consequences until you have sufficient
evidence to conduct a proper risk analysis. This captures the spirit of what
the RSPB spokesperson was trying to say.
This makes it clear that the precautionary principle needs to run in tandem
with risk assessment. It is not a stand-alone principle. Nor does it amount
to the claim that the existence of risk or uncertainty in themselves provide
reasons for not acting. Risk analysis is all about dealing with uncertainty
– where there is certainty there is no risk. The caution that is urged by the
principle is thus against acting on the basis of inadequate evidence. It is
not a kind of blanket policy of risk-aversion. This is just as well, because
if the precautionary principle just amounted to saying "do not take unnecessarily
large risks" it would be a mere platitude.
It is unfortunate, therefore, that the RSPB spokesperson chose to define the
principle in terms of being "unsure of what the result will be." For,
by doing so, a popular misconception about the principle was reinforced: that
it is about being risk-averse where there is uncertainty. Consider this pair
of definitions of the precautionary principle:
"You should not adopt any new technology unless you are certain it is
safe. (Rocky Mountain News, CO, 21 May 2003)
"Nothing should ever be done until it has been proved safe." (Toronto
Star, Canada, 19 May 2003)
The key words here are "certain" and "proved". These demands
are too high. It takes years to be sure that any new drug, for example, is totally
safe beyond all reasonable doubt. And because this version of the precautionary
principle asks us to prove a negative – that no harm is done – it puts absolute
certitude beyond us. Yet the principle demands certitude or else it says we
should not go ahead.
Look at how the idea of the precautionary principle is actually used and more
often than not, I would wager, it is used in this defective sense. The principle
is not used together with risk analysis but as a substitute for it. If this
were just a question of rhetorical shorthand, there would be little problem.
(Except in those cases where such shorthand disguised the fact that a proper
risk analysis has the opposite conclusion to the one claimed.) But the misuse
of the principle is objectionable mostly because it encourages people to ignore
the fact that we have to make decisions based on incomplete knowledge and judgement
of risk, and instead offers the illusion that we can do nothing that isn’t certain
to be safe. The popularisation of certain versions of the principle is thus
contributing to the unrealistic and ignorant attitude that society increasingly
has towards risk.