The Precautionary Principle

On 5 March, Tony Blair gave a speech in his Sedgefield constituency in which he sought to justify his actions in Iraq by emphasising the unprecedented threat that global terrorism poses to the civilised world. He called this threat ‘real and existential’, and argued that politicians had no choice but to confront it ‘whatever the political cost’. This is because the alternative – the possibility that terrorists might get their hands on weapons of mass destruction – was too awful to contemplate. In the days that followed, this speech, like everything else the prime minister says and does with reference to Iraq, was picked over by wave after wave of journalists and commentators. Those who had supported the war concluded that it was a passionate and heartfelt defence of what had been a brave and justified decision. Those who opposed the war found it strong on rhetoric but short on substance, and wondered whether all the passion might not, as so often with Blair, be concealing baser political motives. But what almost no one bothered to ask was whether the central claim in the speech was true. Is it true that the threat of global terrorism has altered ‘the balance of risk’, as Blair called it, so that actions like the one against Iraq can be justified by considering the worst-case scenario if action is not taken? Should worst-case scenarios, if they are sufficiently terrible, trump all other considerations when politicians have to decide what to do?

There can be no doubt that these are weighty questions. Indeed, it is hard to think of any that carry more weight, given the exotic array of worst-case scenarios we are now faced with. In his book Our Final Century, Martin Rees puts the chances of the human race surviving the next hundred years at 50:50.* The list of things that could go horribly wrong ranges from the highly unlikely (a giant asteroid strike) to the frankly bizarre (rogue scientific experiments rolling the cosmos up into a tiny ball) to the all too familiar – global warming, viral mutation, and nuclear or biological terrorism. In the light of this, it is hard to dispute Blair’s claim that considerations of how to deal with catastrophic risks are far more important than any of the questions that are still being asked about who said what to whom in the immediate run up to war. But that is just why it has proved so hard to take him at his word. For Blair’s opponents, it’s impossible to concede the importance of these large questions about risk without also seeming to agree to his plea that they should move on from questioning him endlessly about the details of his own conduct before the war. For Blair’s supporters, it is not necessary to concede their importance, because they have already decided that, largely speaking, Blair did the right thing. The issue has been prejudged both ways – either the war was wrong, in which case these can’t be the most important questions, or the war was right, in which case these are questions to which we already know the answer. The problem is that Blair has raised the issue of risk in order to defend his own course of conduct, which immediately trivialises it. But that doesn’t mean that it’s impossible to think seriously about the broader ramifications of what he was saying. It just means that it’s necessary to divorce what was said from the devious and somewhat desperate politician who was saying it.

Blair’s basic argument is easily confused with what is often called the doctrine of pre-emption, which states that in the war on terrorism governments cannot and should not wait to be attacked before fighting back. But Blair’s case doesn’t have to be put in such incendiary terms, which serve only to collapse arguments about risk into a political ideology – neo-conservatism – and reduce them to a form of warmongering. Blair’s position can just as well be expressed in the more neutral language of precaution. What lawyers, bureaucrats and even some philosophers like to call the precautionary principle states that when faced with risks with uncertain and potentially catastrophic downsides, it’s always better to err on the side of caution. In such circumstances, the burden of proof is said to lie with those who downplay the risk of disaster, rather than with those who argue that the risks are real, even if they might be quite small. This appears to be Blair’s current position on the war in Iraq. As he conceded in his Sedgefield speech, he understands how ‘sensible people’ might have come to opposite conclusions about the threat posed by Saddam:

Their argument is one I understand totally. It is that Iraq posed no direct, immediate threat to Britain; and that Iraq’s WMD [programme] . . . was not serious enough to warrant war, certainly not without a specific UN resolution mandating military action. And they argue: Saddam could, in any event, be contained.

But what this stance does not take into account, according to Blair, is precisely ‘the balance of risk’, meaning it does not take seriously enough the downside of getting things wrong. Blair can accept that his opponents might be right, and he acknowledges that their case has been strengthened by the failure to find any evidence of Iraq’s WMD in the year since the war ended. But they might also be wrong, and the consequences of their being wrong, of containment not working, were potentially much more serious than the consequences of his being wrong, and Saddam not having any weapons. This is why Blair is able to argue that the failure to find WMD in Iraq is not, ultimately, the issue. ‘The key point’, he says, ‘is that it is the threat that is the issue.’

The precautionary principle is sometimes summed up by the familiar proverb ‘better safe than sorry’. Like most proverbs, this one doesn’t help much if you stop to think about it: of course it is better to be safe than sorry (safe is good, sorry is bad, so it’s a bit of a no-brainer). What the precautionary principle states is that if there is a chance you might be sorry, it is better to be sorry but safe. This is the crux of Blair’s argument, though he can’t quite bring himself to spell it out in these terms, because he is still hoping the WMD will be found, so that he won’t even have to say he is sorry. But if he does have to apologise for getting it wrong, this line of defence has the advantage that it is not the exclusive preserve of neo-conservative warmongers. The precautionary principle is championed by all sorts of people who were not at all keen on the war in Iraq. For example, it’s often used to urge much stronger interventionist action to deal with the threat of global warming. Even if some of the science is uncertain, it is argued, the balance of risk requires acting as though the gloomiest predictions about global warming were the most accurate, because getting that wrong is less dangerous than acting on the basis of more sanguine predictions and getting that wrong. Some risks, in other words, are just not worth taking. Why, Blair might say, should we be willing to think this about one threat to our way of life, but not to think it about another?

The tempting response is to point out that in the case of Iraq, taking precautions meant dropping bombs on innocent civilians, whereas in the case of global warming, it simply means insisting on more responsible use of energy resources. But this doesn’t work, because it assumes that restricting the practices that cause global warming is without serious cost. In reality, limiting the practices that produce global warming would inevitably inhibit economic growth, including in those parts of the world where economic growth is desperately needed to increase standards of living and life expectancy. If it turns out that the gloomy science is wrong, lives would have been sacrificed for little or no gain. Therefore, the precautionary principle may require the needless sacrifice of innocent life whether it is applied to war in Iraq or to the emission of carbon dioxide. In both cases, the argument must be the same: it’s worth taking a chance on the needless sacrifice of innocent lives only because the risks of not taking that chance are so much greater.

It’s not possible to argue that the precautionary principle only makes sense when applied to nice environmental issues and not to nasty military ones. But it is possible to argue that it doesn’t make sense in either case. Indeed, this is what its application to the war in Iraq brings out – how can something be called precautionary if it involves a readiness to throw away lives on a supposition? As Cass Sunstein recently pointed out in his Seeley Lectures in Cambridge, the precautionary principle is flawed however it is used – whether the issue is the environment, food safety, terrorism or war – because it is self-contradictory: it can always be used to argue both that we should be more careful and that we should not be too careful. Blair captured this double standard perfectly in his Sedgefield speech. ‘This is not a time to err on the side of caution,’ he said, ‘not a time to weigh the risks to an infinite balance; not a time for the cynicism of the worldly wise who favour playing it long.’ And yet his speech also argues exactly the opposite – that what matters is taking precautions against future disaster, seeing the big picture, weighing the overall balance of risks. In the very next paragraph, he remarks: ‘It is monstrously premature to think that the threat has passed. The risk remains in the balance here and abroad.’ This, then, is not a time to err on the side of caution and not a time to err on the side of incaution. Such an argument can be used to justify anything.

The trouble with the precautionary principle is that it purports to be a way of evaluating risk, yet it insists that some risks are simply not worth weighing in the balance. This could only make sense if it were true that some risks are simply off the scale of our everyday experience of danger. Presumably, this is what Blair was getting at when he said that global terrorism posed an ‘existential’ threat. But existential is a slippery word, in politics as well as philosophy. If global terrorism posed a threat to the existence of all human life on earth, in the way that a ten-mile-wide asteroid heading towards us would, then it might make sense to place it in a different category from all other risks, even if the chances of disaster were relatively slight. But it’s extremely implausible that it does pose a threat of this kind, or at least that it poses any more of a threat than all sorts of other things, including a war between states that are permitted to hold onto their large nuclear stockpiles. If terrorism poses an existential threat, it is not to our existence, but to our way of existence: it threatens our prosperity, our security, our ability to live as we choose, our peace of mind. But while all this is true, it’s not clear that it makes for a qualitative difference between terrorism and the other sorts of risks that we face. Threats of disruption to our way of life, even of the massive disruption that would be caused by a large-scale terrorist attack, can still be compared with the threatened disruption that would be caused trying to prevent them. Yet the precautionary principle implies that in the end there is no comparison. In his Sedgefield speech, Blair discussed his decision to place armed guards and barricades around Heathrow, in response to an intelligence warning of a terrorist attack that never came. He asked us, as ever, to put ourselves in his shoes:

Sit in my seat. Here is the intelligence. Here is the advice. Do you ignore it? But, of course, intelligence is precisely that: intelligence. It is not hard fact. It has its limitations. On each occasion, the most careful judgment has to be made taking account of everything we know and advice available. But in making that judgment, would you prefer us to act, even if it turns out to be wrong? Or not to act and hope it’s OK? And suppose we don’t act and the intelligence turns out to be right, how forgiving will people be?

This passage captures the essence of the precautionary principle: risk assessment will only take you so far, at which point you have to start thinking about the worst that could happen. It also captures something else about the existential threat posed by global terrorism. A large-scale terrorist attack could have cataclysmic consequences for the political existence of whichever politicians happen to be in charge when it takes place. ‘How forgiving will people be?’ Blair asks. The truth is that no one really knows. The evidence of 11 September and of the bombings in Madrid cuts both ways. Many people, undoubtedly, would blame the government. But many others might accept that no government can act on all the intelligence it receives without destroying the way of life it is trying to protect; and even when it does act, it cannot always be sure that its actions will work. What people seem to mind most, if the Spanish elections are anything to go by, is a rush to judgment after the event rather than a failure to exercise the correct judgment before it. The example of 11 September also suggests that following a really traumatic incident the public are as likely to unite around whoever happens to be in charge as they are to seek out someone to blame. Still, because no one can be sure, the precautionary principle comes into play here too. The worst-case scenario is that a major terrorist attack, about which there were plausible intelligence warnings, might destroy not only the careers of individual politicians but perhaps the existence of some political parties, maybe even the viability of an entire system of government. Why take a chance? So much, then, for Blair’s claim that politicians should act ‘whatever the political cost’.

It cannot be argued that terrorism confronts us with risks that are somehow off the scale. An attack on Heathrow, though bad, is by no means the worst thing that could happen. But it could be argued that terrorism confronts us with risks that cannot be measured on any reliable scale, because the evidence is always so uncertain. In this respect, terrorism does pose different sorts of problem from, say, global warming. Although we may not know which scientific accounts of climate change are the most accurate, the different theories are at least scientific ones, and therefore provide the basis for various kinds of risk assessment. But there is no comparable social or political science of terrorism. It is indeed, as Blair says, a question of ‘intelligence’ rather than ‘hard facts’. Moreover, even if the intelligence is as reliable as it can be, the predictable outcomes are far from certain. Were Saddam Hussein to have acquired nuclear weapons, would that have increased the likelihood of their falling into the hands of people willing to use them by a factor of 1, or of 10, or of 100? In the face of such radical uncertainty, it seems to make sense to fear the worst. It also seems to make sense to suppose that any terrorist organisation that is seeking to get hold of nuclear weapons would be willing to use them.

It would be a mistake, however, to assume that because certain risks are unquantifiable in their own terms, no comparative judgments can be made with respect to them. We cannot begin to know what the real likelihood is of al-Qaida acquiring weapons of mass destruction, not even if we had an accurate record of who else has such weapons, which we don’t. But we can compare the likely costs and benefits of trying to deal with this unquantifiable threat at different times and in different places. Indeed, the best case that might be made for seeking to disarm Saddam by force is one that relies on this kind of comparative risk assessment, rather than depending on the precautionary principle. Faced with a wholly unknowable threat, the rational thing to do is to compare those outcomes we can predict, and ignore those we cannot. We might not know, in the end, whether al-Qaida was likelier to acquire nuclear weapons from Iraq, from North Korea or from Pakistan. But we could know with a reasonable degree of certainty that a military confrontation with Iraq carried far fewer risks than a military confrontation with either North Korea or Pakistan. Equally, if Iraq was to be attacked, another old proverb looks quite plausible: the sooner the better. Going to war now, when Saddam was relatively weak, carried fewer risks than waiting for five years, or for ten, when he might be much stronger. So, the best available answers to the questions most often posed before the war – why Iraq and why now – were in both cases the same: because other places and other times carried too many risks. Blair seems to have recognised this, at least in private. In Peter Stothard’s 30 Days, which tells the story of the war from the point of view of those inside Downing Street, we hear the response the prime minister liked to give behind closed doors to those who asked why Saddam: ‘Because we can.’

If there was such a case to be made in terms of comparative rather than absolute risk, and if Blair was willing to allude to it in private, why has he not been more willing to pursue it in public? The answer is because he can’t. It’s not a defensible position, or at least it isn’t for Blair’s government. One problem is that any justification of the war in these terms depends on being able to demonstrate that all the risks of an invasion were taken seriously, including a range of worst-case scenarios for its aftermath. But if any of the champions of war seriously considered that things would turn out as they have – no WMD, no democracy, no real prospect of either – then they are not letting on. There is just not enough visible evidence of long-term planning for this war to be plausibly defended in cost-benefit terms.† Equally, once it becomes a question of relative risk, it’s no longer possible for politicians to rule out of bounds questions about whether going to war increases or decreases the risk of terrorist reprisals. If the government were warned, as we know they were, that an invasion of Iraq was likely to increase the threat posed by al-Qaida to British targets, at least in the short-term, then the case has to be made that these losses are a price worth paying for the possibility of long-term benefits. It is not enough to argue, as Jack Straw did following the Madrid bombings, that ‘al-Qaida will go on and would have gone on irrespective of the war in Iraq, until they are firmly stopped.’ Nor is it enough to say, as he also said: ‘We did it for the best of motives.’ This is the very weakest version of the precautionary argument: we had to do something; Iraq was something; Iraq was justified. It assumes that, faced with a threat like the one posed by al-Qaida, nothing can be weighed in the balance until the threat is finally vanquished. Yet if the Iraq war is to be justified in cost-benefit terms, it can only be because the real costs, recognised as such, are outweighed by the benefits.

This, though, points to the deeper difficulty. A rational risk assessment of war, or of anything else, is an incredibly hard sell for any politician. It carries too many political risks of its own. Politicians have little to gain, and much to lose, by seeking to present their arguments in strict cost-benefit terms. What they have to gain is that they can claim to be behaving rationally. What they have to lose is that behaving rationally about risk leaves even the most sympathetic politicians sounding cynical, and heartless, and lacking in conviction. It means admitting in advance that you might be wrong, because being wrong is one of the risks that have to be weighed in the balance. It also means putting a price on human life, and measuring its loss against the alternatives. Politicians, like insurance agents, do this all the time, but unlike insurance agents they don’t like to be seen doing it. Nor can politicians leave the business of cold calculation up to the technocrats, and then simply argue that they are acting on the advice of the experts. To do so would be to suggest that the politicians are no longer in control. In an email released at the Hutton Inquiry, Jonathan Powell, the prime minister’s chief of staff, wrote to Alastair Campbell and David Manning, Blair’s foreign policy adviser, of an early draft of the September dossier on Iraq’s WMD programme: ‘I think it is worth explicitly stating what TB keeps saying: this is the advice from the Joint Intelligence Committee. On the basis of this advice, what other action could he as PM take?’ Blair might have been saying this to Powell, but what he eventually told the British public in the final version of the dossier was: ‘I believe that faced with the information available to me, the UK government has been right to support the demands that this issue be confronted and dealt with.’ In other words, what matters is what the prime minister believes to be right, given the information available. He had no choice but to confront the issue. But it was important for the British public to know that the decision to confront the issue was still his choice.

The precautionary principle fits neatly into this way of thinking. Because it can be used to justify both precaution and incaution, it can also be used to plead either necessity (we have no choice) or discretion (we have to make a choice), or both, depending on the circumstances. The ability to dress up choice as necessity and necessity as choice was an indispensable part of the prime minister’s armoury in the run up to the war, and he deployed it to devastating effect. Looking back now over the four set-piece debates on Iraq in the House of Commons (24 September and 25 November 2002, and 26 February and 18 March 2003), I am struck by how in the first three the government made it clear that there was no need yet to take a final decision about war, because all the options had to be kept open and the course was not finally set. But in the fourth and decisive debate, Blair made his position clear at the outset: ‘This is a tough choice indeed, but it is also a stark one: to stand British troops down now and turn back, or to hold firm to the course that we have set.’ Whatever else you might think about this, it is brilliant politics, and it worked: the government survived the first three debates by offering the prospect of a decisive vote down the line, and it survived the decisive vote because the earlier debates had allowed it to pursue its policy to the point of no return. A rational risk assessment would have meant laying bare the real necessities and the real contingencies at each stage of the decision-making process. It would also have laid bare the really stark choice: either politicians do what the risk assessors tell them, in which case they have no discretion, or they make up their own minds, in which case they have full discretion. The September dossier, which provided the subject for the 24 September debate, was an attempt to blur this distinction. The job of the JIC is to weigh the risks associated with different pieces of intelligence, including the risk that some of them might be wrong, so that politicians can exercise their judgment. But the dossier, which was said to reflect the view of the JIC and not the politicians, made no mention of risk. It merely judged the intelligence to be true. Which allowed Blair to judge that it had to be acted on. Which allowed him to persuade Parliament of the same. Now that it turns out that the intelligence was not so reliable as once thought, Blair has ordered an inquiry into how it was gathered and assessed, but not into the political decisions that followed, because the ultimate decision was provided by the ‘democratic’ judgment of Parliament. Yet Parliament was not provided with the risk assessments on which to make a democratic judgment. The road from the precautionary principle leads by a long and winding route to the terms of reference of the Butler Inquiry. Both are founded on the same, unjustified supposition: that in the face of an existential threat, some risks are just not worth considering.