Risk

The Robert C Barnard Inaugural Lecture

Delivered to the American Association for the Advancement of Science Washington DC, September 1999

This is a somewhat perplexing honour: to be asked - as a politician - to address a body dedicated to the advancement of science. I assume you have taken this risk for a particular reason although it is not at all clear to me why. Science and politics are uneasy bedfellows at the best of times. When I tell you that the subject of my address today is a discussion of the role of science in assessing risks (and in particular, environmental risks) you should be doubly discomforted. For it is in the field of risk assessment that politicians most freely use - and misuse - scientific evidence; it is, symmetrically, the field of public policy into which scientists are most tempted to stray. The attraction is overwhelming and the results are not always fatal, but it is a minefield for those who have not considered carefully the very different types of discourse that are being used and the quite distinct ends for which scientific information is being sought.

So I must be quite clear at the outset that this is a politician's view. As such, it will add little that is new to a field of enquiry that has, in recent times, spawned a truly voluminous literature (to which, I should add, Mr Barnard has been a distinguished contributor). I am not a scientist. I am a layperson. My observations therefore lack the organising rigour that specialised expertise can bring to bear. Any interest that may attach to my comments is, therefore, rooted in two possibilities. The first is that politicians are rarely challenged to crystallise how they incorporate scientific judgements into the politics of responding to risks. Beside the mountain of academic literature that deals with risk, there is little beyond reactive pronouncements from politicians. Hopefully, my comments will be an interesting insight into the political mind, whatever their intrinsic merit. Secondly, my comments will be illustrated with examples drawn from New Zealand. It might seem presumptuous here in America to draw on New Zealand experience in managing risks. After all, the sheer depth and sophistication of the risk management industry in America - and the range of risks that are being addressed - might be grounds for assuming that there is little new that can be added from a jurisdiction so small and so far away as New Zealand's.

I offer two defences; in the first place, that is the only experience I can draw on. Secondly, New Zealand's response is not without interest because we have engaged in this debate late in the day starting from well behind most developed countries. Whether this is a result of isolation and/or the absence of many of the more celebrated industrial and nuclear risks that have been the subject of so much debate over the last 40 years, I do not know. But it is only in the last 10 years that a political consciousness of the complexity of risk management has arisen. As a result, we have been in a position to erect green field site legislation when others have progressed through several iterations. The advantages and shortcomings of being behind the field and then seeking, rapidly, to sprint to the front are instructive in themselves.

With those caveats and explanations in mind, let me now proceed to describe the way I have thought about environmental risk from the standpoint of political office as Minister for the Environment for most of the last decade in New Zealand. With the exception of a few academics and interest groups driven by convictions rooted in principle and/or ideology, most people who have approached me about environmental concerns have been worried about consequences. They may be actual, adverse effects. Or they may be possible effects that are feared. To the extent that the effects or risks that concern people cannot easily be internalised to those who cause them, my job has been either to find mechanisms that compensate affected parties or decide what level of effect or risk it is prudent to say people should have to live with. It is important that I should spell out my role in this way, since it is the key to how I approach my task. It accounts for the restrictions I apply to my enquiries and inevitably reflects my political and philosophical convictions. Depending on your viewpoint, you may find them a help or a hindrance but it is best that I am transparent about them. Since almost the only point of agreement in the literature on risk is the slippery, subjective nature of the assessments that everyone makes, I had better place my value-laden convictions on the table at the outset - if only to provide you with a point of reference for what follows.

I am sceptical about the efficacy of many proposed government interventions but I don't dismiss the possibility of government action a priori. My scepticism about the possibility of government action stems from a first-hand appreciation of how little governments in societies like ours know in comparison with the endless inventiveness and ingenuity of their citizens; and from practical experience of the unintended consequences of government action. In short, I am well disposed to the practical efficacy of the marketplace as a generator of information.

On the other hand, I view the workings of the marketplace as being deeply dependent on rules that will be effective only so long as they command the willing assent of those to whom they apply. Unlike theorists who assert some prior morality for the distribution of individual and property rights, I take the view that those things that are asserted as rights (and in some countries, such as this one, almost engraved in stone) are the products of social and political conventions. It follows that I don't subscribe to any absolute or inalienable set of rights. I may well assert them as a matter of political rhetoric, but I don't believe there is some predetermined natural order, which dictates our constitutional arrangements. (This is the big claim that separates North American constitutionalism from the more pragmatic Anglo/Australasian view of the terms of our political settlement).

Neither do I bring a strongly predetermined view of human agency to the table. The inhabitants of my universe are strongly motivated by self-interest but not exclusively so; political activity can be strongly rent seeking but is not inevitably so. Government's failure can as easily flow from good intentions as it can from selfishly motivated advocacy. In short, I am at one and the same time aware of the inevitability of political action, sceptical of its efficacy but always open to its possibility given our need to keep the peace and reassess existing institutions in the light of new knowledge. For someone of my disposition, the consequences of environmental risks and our response to them is important. I don't enter the field with a fully-fledged environmental ideology in mind. The various brands of political ecologism that seek to erect an ecocentric view of the world seem to me divorced from the human reality of our situation. As a species we are condemned to an anthropocentric view of the world if only because that is the only standpoint we can physically inhabit. A policy thrust that is indifferent to its consequences for us as human beings seems to me hallucinatory. That does not mean natural and human capital are infinitely interchangeable; nor that we can or should ignore the functioning of planetary ecosystems of which we are a small part. It simply means that any enquiry into environmental risks has to be humanly intelligible. Neither do I discharge my responsibilities believing that it is possible to design a perfect set of institutions to deal with environmental risk - or that the best possible institutions will emerge spontaneously. The former sort of hubris has long been discarded. The latter still persists in a version of market triumphalism that comes perilously close to asserting (in Alexander Pope's words) that "Whatever is, is right." I tend to the view that there is no a priori reason why any set of institutions, be they politically or spontaneously generated, should not be capable of disastrous outcomes. We are, after all, dealing with what Kant termed the crooked timbers of humanity. A conservative and sceptical grasp of the possibilities of government coupled with an acceptance of their necessity is the best ethical equipment I can bring to the task of managing environmental risk in a world filled with people like you and me. If as an academy you are embarrassed by this profession of political motivation, I can only say that it is important that you accept that all politicians are driven by philosophical demons - as are scientists. The fact that they may not even be able to identify them is beside the point. Getting them out into the open won't dissolve what are often debates about incommensurable values. But it should illuminate why, from time to time, people take particularly strong stands. In the same way that politicians who select from scientific sources need to be open about the basis for their selectivity, scientists who venture into the realm of policy prescription need to be clear about the basis for their value judgements.

If assessing and responding to consequences is what environmental politics is all about, how have we responded in New Zealand? I propose to describe our experience by discussing my engagement with two quite distinct issues that illuminate different risks and different calls on the scientific community. The first concerns the release of new organisms in New Zealand; the second, the debate over how we should respond to the risk of human induced climate change.

Releasing New and Modified Organisms

While there are many risks that can be handled either by market or government institutions at the local level, there are some environmental risks that implicate an entire national community. These will, frequently, involve a poorly defined group of affected persons. As a result, there is often a much higher chance of risks being unwittingly imposed rather than voluntarily assumed. And there may be elements of ecosystem-wide irreversibility that aren't encountered with local, site specific activities.

In the New Zealand context, this type of risk has been most acutely encountered with respect to the release of organisms into the general environment. A number of factors have served to make the release of organisms a particularly acute risk management issue. New Zealand is made up of two large islands (slightly larger in combined area than the State of Wyoming) and many smaller ones. Its spatial and temporal isolation (1300 miles and 70 million years from Australia) has bequeathed it a unique indigenous ecology that evolved without land-based mammals. That natural endowment has, over three quarters of the country's surface area, been overlain by an imported, northern hemisphere, temperate grassland ecology. The indigenous ecology is still in the throes of a traumatic round of extinctions as a result of the arrival of humans around 800 years ago and a host of four-legged predators subsequently. The introduced agri-business ecology is in a fragile state but for the opposite reason: its vigour relies in no small part on the absence of predators and pathogens that were left behind in large numbers at the time of European colonisation but could arrive at any time. A large amount of energy is thus devoted to keeping things out of New Zealand, both to avoid additional pressures on indigenous biota and to conserve the competitive advantage of wanted introduced species.

To inform the basis on which new introductions should be permitted, the Government has enacted a statute called the Hazardous Substances and New Organisms Act 1996. The part of the Act that deals with hazardous substances is a conventional, effects based statute that seeks to manage risks through assessing likely costs, benefits and uncertainties, and imposing controls if expected risks exceed regulated thresholds. The new organisms provisions of the Act, by contrast, break new ground by embracing not just exotic organisms but genetically modified ones as well.

The primary purpose clause of the statute talks about protecting the environment, and the health and safety of people and communities, by preventing or managing the adverse effects of hazardous substances and new organisms. The outcomes that are sought in the pursuit of that goal are twofold: "the safeguarding of the life supporting capacity of air, water, soil and ecosystems and the maintenance and enhancement of the capacity of people and communities to provide for their own economic, social and cultural wellbeing and for the reasonably foreseeable needs of future generations". The Act also makes specific reference to the need for a precautionary approach "in managing adverse affects where there is scientific and technical uncertainty about those affects." The extreme lengths to which New Zealand is prepared to go in defending its biosecurity and biodiversity is best described by reproducing in their entirety sections 36 and 37 of the Act:

36. Minimum Standards - The Authority shall decline the application, if the new organism is likely to - (a) Cause any significant displacement of any native species within its natural habitat; or (b) Cause any significant deterioration of natural habitats; or (c) Cause any significant adverse effects on human health and safety; or (d) Cause any significant adverse effect to New Zealand's inherent genetic diversity; or (e) Cause disease, be parasitic, or become a vector for human, animal, or plant disease, unless the purpose of that information or release is to import or release an organism to cause disease, be a parasite, or a vector for disease.

37. Additional matters to be considered - The Authority, when making a decision under section 38 of this Act, shall have regard to - (a) The ability of the organism to establish an undesirable self-sustaining population; and (b) The ease with which the organism could be eradicated if it established an undesirable self-sustaining population.

Only if the minimum standards in section 36 are met and the additional matters in section 37 are adequately dealt with and the positive effects of the organism outweighs its adverse effect and there is sufficient evidence to assess those adverse effects, will an importation or release be permitted. Applications are heard by an expert tribunal (unmemorably called the Environmental Risk Management Authority or ERMA) whose decisions are final. Hearings are open to the public and any member of the public may make submissions.

As if that wasn't enough, the Parliamentary Select Committee that scrutinised the legislation imposed an additional requirement on the decision-maker. It required the Environmental Risk Management Authority to submit to the Government for formal ratification, a methodology that the Authority must apply consistently in considering all applications before it. The Select Committee's concern was that, in the absence of a formal road map of how the Authority would consider risks, applicants (and objectors) would have no prior warning of how applications might be treated and the Authority could, unwittingly, end up dealing with similar cases in quite different ways.

The detailed nature of the methodology that has been promulgated serves as a warning of what happens when the multifaceted world of risk assessment collides with the requirements of the legal universe for consistency and the political domain for exhaustive disclosure. The methodology runs to well over 3000 words and is a minor essay in its own right. The sections of greatest interest to this audience are those that relate to uncertainty, the treatment of risk and the aggregation and comparison of risks, costs and benefits. Here is what it says:

Uncertainty

Faced with uncertainty or dispute about scientific or technical information relating to the potential adverse effects of an organism, the authority must determine the materiality and significance of the uncertainty or dispute "taking into account the extent of agreement on the scope and meaning of the scientific evidence." If the Authority is unable to clarify the matters in contention - either in discussions with the applicants and objectors or through its own endeavours - it must take into account the need for caution in considering the possibility of adverse effects should release be approved. Where there is uncertainty about the costs, benefits and risks at stake, the Authority is directed to "establish the range of uncertainty and ... take into account the probability of the costs, benefits and risks being either more or less than the levels given in evidence."

Approach to risk

In its consideration of applications, the Authority is directed to have regard to "the extent to which the following risk characteristics exist:

(a) exposure to the risk is involuntary;
(b) the risk will persist over time;
(c) the risk is subject to uncontrollable spread and is likely to extend its effects beyond the immediate location and incidence; (d) the potential adverse effects are irreversible; and
(e) the risk is not known or understood by the general public and there is little experience or understanding of possible measures for managing the potential adverse effects."

Aggregation and comparison of risks, costs and benefits

When evaluating the combined impact of risks, costs and benefits, the Authority must "as far as possible

(a) combine groups of risks, costs and benefits using common units of measurement, including where applicable, monetary valuation; and

(b) use other techniques where common units are not possible, including, the identification of dominant risks (being risks that may have a deciding influence), and the ranking of risks in order of significance."

It would be hard to imagine much else that could be considered, but over and above all these provisions there is a powerfully risk averse bias in the statute itself when it comes to the treatment of new organisms. For unlike hazardous substances (in respect of which the Authority may impose wide ranging controls), no conditions may be attached to the release of new organisms. It is all or nothing. Any adverse effects that may result from a release must in effect be of a magnitude that the Authority can justify spreading throughout the entire community. The unspoken logic of the statute is that to allow the imposition of controls to minimise unacceptable risks would, in effect, invite the Authority to cross the threshold of release and take a risk where it might otherwise not be taken. Whilst there are special provisions for the importation or introduction of new organisms into containment for research purposes, release into the environment at large is construed to be irrevocable and beyond mitigation should something go amiss.

As I have already noted, this is perhaps not surprising in a country that has witnessed the truly devastating consequences of poorly planned biological releases. A host of animals like deer, goats, rats and possums are munching their way through New Zealand's ancient forests that are largely unchanged since our separation from Gondwanaland. We have extremely high levels of endemism and, as a result, high rates of extinction among our avifauna which has not co-evolved with mammalian or marsupial predators. New Zealand's indigenous biota has been defenceless in the face of successive bio-invasions.

An understanding of the peculiar ecology of our islands has been slow to emerge. One generation of early settlers released rabbits whose numbers, in the absence of their particular predators, exploded to plague proportions threatening to wipe out a new and vigorous pastoral industry (itself burgeoning on the back of widespread habitat destruction). The next generation introduced weasels and stoats and ferrets as early bio-control measures. These creatures didn't do much for the rabbit problem, but they soon took a liking to the local bird life adding to the pressures that loss of habitat had already placed on some bird species. In more recent times ferrets (along with deer) have emerged from the wilds to present new risks as TB vectors to the pastoral sector they were introduced to assist.

With these catastrophes still vividly imprinted on the national consciousness, it is as I say, not surprising that such an extremely risk averse stance should be adopted in legislation. It is one thing to legislate however. It is another entirely to know whether we have reached a publicly acceptable approach to handling these sorts of risks. Let me illustrate this point by outlining to you the story of one of New Zealand's most celebrated and infamous biosecurity lapses - the illegal introduction of the RCD virus (otherwise known as rabbit haemorrhagic disease) in 1997. RCD represents the latest chapter in New Zealand's battle with rabbits. The proposal to introduce it followed on from the decision a decade earlier not to allow the importation of another biological control - myxomatosis. Myxomatosis had generated concern on a number of fronts but the most contentious aspect of it was the way in which rabbits would die. The disfiguring and lengthy (anything from 48 hours to two weeks) process of death led to a significant animal welfare lobby opposed to the release. At the time that myxomatosis was turned down, however, the possibility of the rabbit calicivirus appeared on the horizon and its apparently swift and lethal action appeared a much more palatable option.

The application to release the virus was not processed under the Hazardous Substances and New Organisms Act which had not, at that stage, been brought into effect. But the Ministry of Agriculture handled the application in a more public and arms length way than had ever previously been attempted so that the spirit of full disclosure, transparency of reasoning and the opportunity for public input enshrined in the HSNO Act was to some extent tested out in advance.The application to release RCD was brought forward in unusual circumstances. The New Zealand and Australian Governments undertook extensive trials on an island off the coast of South Australia to seek to understand the mechanism whereby the virus spread and to test native fauna from the two countries for susceptibility to the disease. In the middle of the trials, the virus escaped to continental Australia (probably on a fly or similar insect vector) and was soon sweeping across the landscape albeit with varying results.

The unplanned release, in Australia, raised the prospect that the virus could well arrive in New Zealand anyway (through accidental or deliberate human agency). It was decided that an application to introduce the virus should be brought forward by a consortium of local and central government agencies, and farmers. There was a range of uncertainties and risks raised by the application all of which were investigated in considerable detail. The two most contentious ones were:

• the potential impact on indigenous flora and fauna caused by the risk of rabbit predators such as stoats and weasels switching to native birds in the event of a dramatic decline in rabbit populations;
• risks to human health on account of the instability of RNA viruses like calicivirus and the possibility that an adverse evolution of the virus could see a change in host specificity or in virulence.

The first risk was able to be assessed with a reasonable level of confidence and contingency plans were designed to deploy supplementary pest control measures in areas where predator prey switching was most likely to occur. The second risk was subjected to extensive international review by virologists. The assessment concluded that the risk of an adverse epidemiological event was very low, but not zero; that the probability of the virus producing clinical disease or pathological change in species other than the rabbit was low; and that the likelihood of an adverse change in its host specificity or virulence was very low.

There is not room here to discuss the decision in detail (and in any case I am not scientifically equipped to provide an informed commentary). But it is clear from the decision - and this seems, intuitively, decisive - that since the virus was already present in 40 countries, New Zealanders were clearly exposed to the same level of background risk as people in countries where the virus was already present. The movement of people globally is such that viruses will inevitably cross every border and defeat every public isolation and biosecurity defence we are capable of designing - unless of course we bring international commerce to a complete halt.

Notwithstanding the extensive enquiries made by the Ministry of Agriculture, the decision-maker (a senior MAF officer who held independent and unreviewable powers) turned down the application. None of the risks (including those I have outlined) were nominated as the reasons for his decision. Rather, he remained unconvinced that the benefits advanced by the applicants would actually come to pass given the wide range of uncertainty about the pathogenesis and epidemiology of the disease and in particular how the virus was transmitted in other countries and whether transmission was even possible in New Zealand conditions.

The decision was a surprise to many people given the reasonably reassuring answers on the key risks and the host specificity observed where the virus was present abroad. Surely, many suggested, it was worth giving the virus a try even if the benefits were uncertain. On this point, the decision-maker simply noted that experimentation and trialing was impossible because containment of the virus, once it was released outside of the laboratory, was impossible. However scientifically or ecologically appropriate trials were, they could not be contained so the risks would have to be run in their entirety. Reading very much between the lines, the decision-maker was saying that he did not feel able to conduct a real live experiment, uncontained, in the community at large. And in this, his sentiment foreshadowed the very high level of risk aversion that found its way into the HSNO Act to preclude release under specified conditions.

If the decision was surprising, the sequel was - at least in some quarters - less so: the virus was illegally imported into New Zealand. Its performance, much as predicted, was uneven. The contingency plans for flora and fauna threatened by prey-switching were activated. And to date, the virus has maintained host specificity as it has in every other country. The outcome from a science point of view was unremarkable. But from the political side of the risk management fence, there are several points of interest.

In the first place, despite a thorough analysis of the risks, opponents of the release were unmoved. In fact my perception was that the more information that was amassed, the more polarised viewpoints became. Increasingly detailed scientific analysis seemed to sustain (and indeed augment) viewpoints that were radically opposed to one another. In other words, the science became increasingly irrelevant as deeply held and different levels of risk aversion came to the fore.

The second feature revealed by this episode is that strenuous efforts to manage risk are always dependent on compliance with the law. Despite lengthy enquiries, no charges have ever been laid against those who illegally imported the virus, distributed it to desperate farmers and started the clandestine release. It has been assumed throughout that the deed was done by farmers driven to the point of exasperation by the slow (and from their point of view) ultimately fruitless process that had examined and then rejected first myxomatosis, then RCD. The immediate economic disaster they faced undoubtedly added to the high stakes' gamble they took which was, at the time, described as being evidence of their strength of feeling and their despair that the majority of New Zealanders had become indifferent or, worse, hostile to their concerns.

No one would, in any circumstances, condone a breach of the Biosecurity Laws. That it should, apparently, have been committed by members of a community who relied on the maintenance of strict biosecurity for their competitive advantage added to the enormity of the offence in the eyes of many people. But it does underline the fact that risk assessment is not a process that can be guaranteed to end in a consensus and that, at times, the passions that are unleashed will be so fierce that no amount of information will settle the matter. It also underlines the reality that the integrity of New Zealand's biosecurity laws depends on public acceptance of a particular understanding of their role in controlling risks.

This in turn draws attention to an important point about the limits of political action. The RCD decision was taken under laws that precluded political intervention. The subsequent illegal importation, however, forced the debate into the political arena. The HSNO Act, while giving the decision-making power to the ERMA, explicitly provides an avenue for the Minister for the Environment to take the final decision after the Authority has heard all the evidence.

The grounds on which the Minister may call-in a decision add little to those that already apply to decisions that remain in the hands of the Authority. But the Minister is given one potentially potent power. In deciding that he will take the final decision, the Minister is entitled to specify "in the circumstances of the particular case, what is or is not significant" for the purposes of applying section 36 of the Act. Section 36, you will recall, spelt out the risks which, if found to be significant, provide sufficient grounds alone for the Authority to turn down an application. In other words, a politician is given the power to provide a definitive and final definition of an operative provision of the law.

Jurisprudence scholars in this audience may well quail at the opportunity for executive excess this might imply. But from the point of view of any politician who might advocate its use, it raises as many problems as it might solve. The provision was framed while the RCD process was being considered and was frequently referred to by those who argued that a duly elected and publicly accountable office-holder should have to exercise the judgement required in hard cases such as those posed by the RCD application. It is an easy argument to advance for everyone other than the person - for the time being me - able to exercise the power. Imagine, in the RCD case I have just outlined to you, deciding that, for the purposes of section 36, adverse effects on human health or safety should not be regarded as significant. Quite apart from requiring the wisdom of Solomon, such a possibility assumes an ability to reflect through the person of a single politician, a level of risk aversion that somehow reflects community held values. The problem is that the circumstances in which such a consensus could be discerned would be the very circumstances in which the pressure to call-in a decision would be least pressing. Conversely, it is in precisely those cases where the community is irreconcilably divided that the pressure to call-in the decision (and thereby render it politically influenceable and accountable) will be most intense.

Given the low regard in which politicians are held in most contemporary democracies, it seems counter-intuitive that we should want politicians to involve themselves in weighing up the esoteric concoctions of science, ethics, values, costs, benefits and probabilities that are at stake. But this is what the New Zealand legislature has opted for. It is, at the very least, a frank admission that at their most profound, these issues of risk management enjoy no immunity from political engagement. Which, in a way, is exactly what the illegal importers of RCD proved.

There is scant chance that the sort of debate raised by the RCD affair will be an isolated occurrence before the ERMA. Controversy over the possible field release of genetically modified crops is already on the boil. In a country such as the United States where field release has been a fact for five years, this may seem strange but the acknowledged risks that some transgenic crops raise take on an added potency when injected into the debate over risk management in a country like New Zealand that has a long tradition of wrestling with biological risks.

The HSNO legislation should provide applications for the field release of GMOs with one of the most rigorous and transparent assessment procedures in the world today. Indeed, it is only the existence of the statute that stands between potential applications and widespread calls for an across the board moratorium on the release of GMOs. By and large, I am confident that the ERMA can handle the case-by-case risks more than adequately. But there are some special elements of the GMO debate that could benefit from separate and prior consideration. They have to do with irrevocability and the extent to which anyone has an 'ownership' stake in the status quo.

Some food producers - organic farmers and beekeepers, have raised the issue in particular - who argue that the introduction of genetically modified crops would irrevocably deny them the ability to market their products as coming from a GM-free environment. On the face of it, this is an argument that is more to do with marketplace perception than any real environmental risk - and then, only in markets that have themselves a high level of concern about genetically modified food.

It raises a claim to the preservation of the status quo ante which will, inevitably, be challenged by those who wish to benefit from the advantages that some genetically modified crops may offer (including environmental benefits in some cases). We have, as a society, generally accepted that the position of all of us is subject to change through technological innovation. Here, potentially, is a claim that an entire application of molecular biology should be forever contained to the laboratory to ensure not just a low level of risk but in this case the maintenance of a state of affairs that precludes the possibility of any risk materialising.

If this seems to be without precedent, I should remind you that New Zealand has once before taken a zero risk approach and that is with respect to nuclear energy. Starting with a Royal Commission in 1978 that recommended against nuclear power, New Zealand has set its face increasingly against any engagement with anything nuclear. The spillover into our foreign policy is well known in this country and has complicated an otherwise excellent relationship. But the political reality is that, whatever the assurances about risk, New Zealanders not only want no connection with things nuclear: they have turned the stance to one of positive advantage in promoting New Zealand as a nuclear free tourist destination and food producer. Whether that makes any rational sense, it is perceived by some businesses as a real advantage in the marketplace.

Given the biological nature of New Zealand's economy, it is to my mind inconceivable that New Zealand would adopt such an approach to the use of biotechnology. But there is no question that, given the wide provisions of the HSNO Act, an argument could be mounted against the field release of genetically modified crops on the grounds that any benefits alleged in support of their release were outweighed by the costs to those who sought to maintain a GM-free growing environment.

An application that did not pose any significant adverse consequences (in terms of section 36) and didn't raise the risk of undesirable self-sustaining populations (in terms of section 37) would, in my view, be likely to win approval from the ERMA. It was not, after all, established to eliminate risks - rather, to manage them. But it would be a pyrrhic victory if (as in the UK) the result were mindless destruction of the crops once planted. Indeed, in New Zealand, a Government owned research institute has already had contained trials destroyed by activists. Threats of vandalism are no basis for a defensible policy. But, equally, blunt assertions by experts that the risks are acceptable are not in themselves persuasive if people feel that the environment they live in is being exposed, irrevocably and without their consent, to a risk they do not understand.

There have been, in New Zealand, calls for a moratorium on the field release of genetically modified organisms pending an enquiry by a Royal Commission. Such a commission if established would, not withstanding the painstaking effort of enacting the HSNO legislation, trawl back through the debate about the way in which GMOs should be treated. The Government has rejected this approach, largely on the basis that it would not yield any new information and would be unlikely to narrow the gap between points of view that, beyond a certain point, are unlikely to be influenced by evermore fine-grained and voluminous analysis. In any case, there is something artificial about having two or three eminent persons spend months if not years hearing a vast amount of evidence if the outcome (a weighty volume of carefully balanced words) is immediately rendered obsolete by fast moving science. Such an approach also reinforces the view that this is a matter that great minds can pronounce on when, in fact, the issue is less about the quality of the science than the community's understanding of the nature of the risks and its ability to debate them.

On this basis we have opted, instead, for an independent advisory group whose brief is an interactive, community-based engagement in debate on an iterative basis rather than a single and purportedly final pronouncement. Its membership spans science, community and ethical expertise and it is tasked with staying ahead of the debate. As such it is in a good position to pick up the concerns of a group like the organic farmers and commence a dialogue that will, hopefully, place the contested issues in context before legal processes are activated. Between the two extremes of organic growers wanting the entirely understandable assurance that neighbours shouldn't be able to plant weedy, potentially hybridising close relatives of their own crops next door, and the removal of any risk by prohibiting any GM crops anywhere, there must be a middle ground. Whether there is a middle ground depends on just how risk averse in the face of uncertainty we are. Those in favour of a moratorium have, in a more or less explicit way, referred to the sorts of propositions that underlie the so-called precautionary principle. Characterised in extremes terms, the precautionary 'principle' demands that a moratorium or prohibition should issue when there is a threat of a reversible harm even if there are doubts about the causal link between the threatening activity and the possible harm, or the likelihood of that harm coming to pass.

A letter published in a very recent issue of Nature1 points out that this reading of the 'principle' leads to paralysis: that "in the case of genetically modified (GM) plants ... the greatest uncertainty about their possible harmfulness existed before anybody had yet produced one..." with the result that no experimental process that could show whether there are risks would ever be allowed to proceed. In putting this problem down to "a common problem in attempting to convert moral choices into legislation", Holm and Harris contend that "the precautionary principle will block the development of any technology if there is the slightest theoretical possibility of harm."

This, I believe, is an extreme interpretation of the Rio Declaration's version of the Principle that talks of taking "cost effective measures" to prevent degradation. This falls considerably short of an absolute, blocking approach that would, if Holm and Harris were correct, make intelligent risk management impossible. As I have noted (above) the Hazardous Substances and New Organisms Act in New Zealand makes reference to something called the 'precautionary approach'. I have deliberately avoided using the term 'precautionary principle' in any legislation that I have sponsored since it carries with it the promise of a formula - a talisman of certainty - that simply doesn't exist. A more pragmatic approach simply allows that there is no unique principle of rational choice that can assist us in these circumstances. I am clear, however, that there are real risks in a precautionary approach that is prepared to place such heavy weight on future risks that the status quo (with its attendant risks) is automatically favoured. This would represent the replacement of open scientific debate with a clumsy ideology of risk that would ultimately be inimical to the very scientific enquiry we need to manage pre-existing risks and those yet to be identified.

What, then, are the conclusions for the role of science that may be drawn from New Zealand's engagement with the new organisms debate - modified and unmodified? It goes without saying that high quality scientific commentary is central to any attempts to reduce uncertainty, provide a measure of the risks at stake and form the basis for weighing up costs and benefits. It also goes without saying that public consideration of complex issues would be greatly assisted by a more scientifically literate populace - something I have often enough lamented.

But the temptation to dismiss those who, in the face of low, well-described risks refuse to accept that they should be exposed to them, must be resisted. So must the temptation to provide ever more exhaustive evidence that, at the margin, adds little to the balance of the evidence. Beyond a certain point, the evidence ceases to be the limiting factor. It is, rather, a culturally derived aversion to risk that will reflect the particular experience of the community in question.

In New Zealand's case, the risks to a unique indigenous biodiversity and an economically crucial biological economy have generated a set of aversions that would have much less potency in a continental setting like this. Equally, New Zealand's reliance on a basically biological economy means that it will be under much greater pressure to use biologically engineered solutions to maintain its competitiveness and to combat environmental problems that have, until now, only invited crude solutions (such as the use of artificial pesticides and toxins) that carry their own environmental risks and attract mounting public concern. Indeed, because New Zealanders increasingly recognise the potential to market their food products on the basis that they are produced in one of the least polluted environments in the world, the need to develop more sophisticated and less intrusive biotechnologies is all the more pressing.

An interesting future awaits attempts to control (and hopefully eliminate) arguably New Zealand's worst introduced predator and vector for disease: the possum. Possums were introduced from Australia and, lacking any predators, have burgeoned in the primeval forests that still cover 25 percent of New Zealand and the adjacent farmland. They are responsible for threatening a number of forest plants (as well as radically altering the structure of the forest canopy) and contributing to the decimation of many of the remaining endemic bird species through preying on eggs and juveniles. Possums are also a vector for tuberculosis, which poses a risk to deer and cattle industries. And they wreak considerable havoc on ornamental and commercial forest plantings. New Zealanders swerve towards these animals when they freeze, blinded by headlights, on rural roads at night.

Control is largely effected through massive poisoning operations that have, as their mainstay, relied on sodium fluoroacetate, commonly known as 1080 - an extremely toxic substance that is, nevertheless, degraded by micro-organisms in soil and water to completely harmless and commonplace metabolites. It is a control agent that would be useless in most parts of the world since four-legged animals are normally among the species for which protection is being sought. Only in New Zealand's unique forest environment is the elimination of all mammals the conservation goal. New Zealand scientists have carried out a large amount of research on the environmental safety of the compound and have satisfied successive governments that its continued use is not only safe but the only feasible control mechanism presently available on the scale required (there may be as many as 70 million possums munching their way through our forest at the rate of 20,000 tonnes of dry matter per day).

Notwithstanding that, there has been a public uneasiness about the use of the toxin that has, from time to time, bubbled over into heated protests. And even among those who are comfortable with its use, concerns as varied as bait shyness and international tourist perceptions have led to ongoing pressure for a cost effective alternative. Biologically engineered solutions feature prominently on the list of alternative lines of attack and, if carried to fruition, will raise fascinating issues that will test the scientific and non-scientific parameters of risk assessment. The very fact that New Zealanders are prepared to tolerate the aerial bombardment of forests with poison when they will not sanction the widespread use of much less toxic compounds in other situations says something about the scale of the problem and the depth of feeling that supports the effort to control the devastation. It will be interesting to see what sort of public response is generated by attempts to interfere with possum fertility using genetically manipulated material. The challenge for researchers is made doubly complicated by the fact that in its native Australia, possums and other related marsupials are among the threatened species that that country is seeking to preserve. Swerving to kill a possum, which is almost a moral responsibility in New Zealand, is a criminal offence in Australia! New Zealand scientists have worked with their Australian counterparts closely from the outset on issues relating to possum fertility so it should be possible to engineer solutions that are acceptable from a scientific point of view. But public acceptability will be an altogether different challenge and the way scientists address this - and any other significant exercises in expanding risks - will be crucial to winning acceptance.

I would offer two observations on how scientists can contribute to public confidence about risk assessment on the basis of my experience in New Zealand. In the first place, it would be a mistake to rely purely on process (no matter how exhaustive) as a final guarantee of public acceptability. New Zealand has erected one of the most elaborate and publicly transparent processes for controlling the importation and release of new organisms. The community can be certain that before a decision is taken, all the information that is available in support and in opposition to release will see the light of day and be applied within a statutory framework that places a premium on caution in the face of uncertainty. For many, that will be adequate assurance that matters have been properly considered. But the fact that a public process has been carefully followed should not be used to support the proposition that any final determination amounts to a consensus that will be acceptable to all parties. As the RCD affair in New Zealand showed, strongly held opinions were not dislodged which led, in turn, to the flouting of the law by those who wanted to take the risks inherent in a release. Equally, telling people opposed to running the risk that they had had their day in court would have cut little ice with those who considered that they were being exposed, involuntarily, to risks that they believed to be unacceptable.

This leads me to my second observation: an application is likely to win a greater level of acceptance if the existence of public apprehension and specific risk aversions have been factored into the design of a solution from the outset. Public apprehensions of risk should be regarded as a prevailing background condition that moulds research every bit as much as the biophysical limits that constrain the potential solution set. Attempts to engineer brilliant solutions followed by the development of a risk assessment process that is designed to tackle pre-existing fears head on, seems to me a particularly risky strategy. I am not suggesting that research should be constrained by ill-conceived public fears or that scientists should regard the public as ineducable. On the contrary, scientists must be prepared to challenge conventional wisdom fearlessly. But at the same time, if novel techniques are going to be deployed in an increasingly risk averse world (which by my estimation is more superficially informed about more things than ever before in human history) then the community's particular concerns and aversions must be factored in from the outset if confidence in solutions with community-wide and irreversible consequences are to be sustained.

Finally, let me say a word about the uses to which science is put in the course of advocacy and some of the responsibilities that rest on scientists and politicians alike. I could easily enough illustrate this issue by referring again to the debate over releasing new organisms. But I should like, instead, to refer to the debate over climate change since it has proved to be so controversial and so much has been written about it. Besides, it has consumed more of my time, to less effect, than any other issue over the last nine years.

Again, I am not qualified to express a scientifically persuasive view on the climate change debate. I do, however, try to keep abreast of the burgeoning research in this area and, through the weekly pages on Nature, have a reasonable idea of the sorts of questions that are being tackled by modelers, atmospheric and oceanic scientists, palaeo-climate researchers and the huge range of inter-disciplinary enquiries that have been unleashed to gain a better understanding of the dynamics of the carbon cycle and its influence on climate.

Let me lay my cards on the table and say that my reading of the science leads me to the view that the unconstrained emission of anthropogenic greenhouse gases poses a real risk that should be not ignored by policy makers. That risk is not posed by historical emissions or confirmed by evidence of climate change to date. It is, rather, based on an assessment of what might happen in the second half of the 21st century if atmospheric concentrations of greenhouse gases approach one and a half to two times pre-industrial levels. The nature of that risk is still ill-defined. We don't know how big it is, how long the time frame is over which it may be unfolding or even what the `natural' background trends are against which that risk is being run. But we do know, from the Vostok ice cores for example, that we are already well outside the `operating band' for CO2 concentrations that have prevailed for the last 420,000 years. This, it seems to me, is a fair basis for concern given the high correlation between temperature and CO2 concentrations.

Nevertheless, it has been remarked on more than one occasion that if we had known in 1992 how little we would still know about the natural carbon cycle today, we might never have signed a Convention in the first place. I have been attacked for making this statement, but I see no point in being anything other than scrupulously honest about what we don't know. Indeed, ignorance is the very best platform to justify further intensive research. This is where the research community must think carefully about its relationship with the policy advocacy and political community.

We are faced with a classic case of decision-making under conditions of uncertainty - and on a truly global scale - which applies to very few other environmental concerns. Uncertainty normally occasions prudence based on a reasonable assessment of the risks. There is no unique calculus to be applied but we shouldn't be afraid to acknowledge the uncertainties that exist. To do so risks seriously harming the relationship between those whose task it is to judge scientific uncertainties and those responsible for managing the risks of those uncertainties.

Uncertainty has never been a very good basis for a rallying cry. Perhaps that is why activists on both sides of this debate seem determined to replace the cautious, sometimes equivocal and always provisional conclusions of research, with the sort of certainty that can only be summoned by those who, for reasons best known to themselves, have pre-determined the issue and don't want to be confused by any arguments to the contrary.

On the one hand, those who wish to see immediate and serious cuts in emissions (no doubt from some innate distaste for consumer society) point to 'changes' in natural patterns, such as the recent severe El Nino and La Nina cycles or the localised break-up of ice shelves, and claim them as 'evidence' for human-induced climate change and imminent catastrophe if action is delayed. On the other hand, we observe those with much to lose from any reduction in emissions, skilfully interpreting any uncertainties as evidence fatal to the case for early or indeed any action. This camp frequently seizes triumphantly on any skerrick of evidence that suggests 'natural', non-anthropogenic warming trends may be discernible.

The catastrophists fail to explain that our baselines are simply too short to make such definitive claims; the sceptics ignore the fact that if global temperatures are rising for reasons that have nothing to do with burning fossil fuel, our room for manoeuvre may be even more limited than we think.

In my view, the problem with this hectic search for 'evidence' is that it proceeds from a faulty premise: that this is a debate about proving or disproving the case for reduced emissions on the basis of current knowledge. The truth is, that if we were truly certain one way or the other, the world would either have taken fairly decisive steps to curb emissions already (as it did with ozone depleting substances) or it would have consigned the issue to the waste paper basket. The fact that we have done neither - for the best part of a decade now - is that we continue to wrestle with a risk whose dimensions we cannot be sure of.

About the only thing we can be sure of, is the likelihood that the 'science' will remain incomplete on this issue for some considerable time to come. And while by 2008 we should know a great deal more than we do today, we will still not know for sure the extent of the problem and its full implications.

With these observations in mind, I should like to make three points. The first is that there is a need for much more research. By its very nature, it will have to be funded by the world's taxpayers as a public good. Unlike the issues I described earlier, with respect to new organisms, I think we are a long way away from the point where additional research offers diminishing returns to our understanding of the risks. Unless the dimensions of the issue are made much clearer, I do not believe people will summon the determination to respond to a risk whose impact will only be felt on an inter-generational timescale.

We need a much more robust understanding of the global carbon cycle and our interference with it. We need to understand much more about the feedback loops that can suppress or amplify climate responses and try, through a mixture of palaeo-climate research and modelling, to gain a better understanding of the thresholds that separate incremental from non-linear responses. That won't be achieved quickly. It will require international co-operation.

This is where the bulk of New Zealand's response has focused to date. Our efforts to curb emissions have been modest. But for a small country we have made a significant contribution to climate change research by enabling our scientists to participate in global research into crustal, oceanic and atmospheric processes. I believe it is money well spent and the knowledge gained from this sort of research is (like so much basic science) likely to generate unforeseen and unforeseeable benefits in fields apparently unrelated to the original quest to understand the planetary climate system.

Secondly, policy makers must resist the temptation to dismiss research results not on the basis of their empirical validity or otherwise but their authorship. A sophisticated audience such as this might well be tempted to think itself immune from such tawdry behaviour. You would be unwise to assume that. The political world I inhabit is, to a large extent, powered by an examination of arguments on the basis of interests - real and alleged. It is unavoidable and can taint everything that comes in contact with it.

Climate change politics provides almost daily evidence of this. The testimony of research institutes that accept funding from the fossil fuel industry are routinely dismissed without serious analysis of their work by environmental groups on the basis that the results have been `bought'. A symmetrical claim is made by some private sector lobbyists, who charge that publicly funded climate scientists are simply fuelling the climate change debate to keep the research funds flowing in. In this vein, I have had respectable representatives of the business community tell me that the very fact that the Intergovernmental Panel on Climate Change has its roots in the UN system is a cause for serious scepticism.

There is a cynicism about these charges that is potentially pernicious. It is one thing to go along with the sensible observation that people in public and private employment are prey to the same self-interested calculations. It is another to convert this into a crude tool that automatically selects scientific opinion on the basis of an a priori judgement about the motivations of the researchers. Such an approach replaces empirical enquiry with ideology and represents an on-going risk to science in societies like ours that must be vigilantly watched for from whatever quarter it may emerge.

As far as the IPCC goes, I think it is absurd to characterise it as a self-justifying tool of the UN. It is a process that engages scientists around the world and attempts to provide policy makers with a scientific assessment on which they can base their assessment of the risks I have described. As a way of gathering together the vast amount of research being funded by governments around the world, it seems to me a very useful tool. But its credibility is crucially dependent on transparent processes and scrupulously careful attention to the full range of research results that are being generated.

This brings me to my third observation: that scientists must not feel pressed into framing their conclusion in a way that implies more certainty or unanimity than exists. You will recall the controversy that surrounded the IPCC's Second Assessment Report as a result of the statement in the Summary for Policymakers by Working Group I that "the balance of evidence suggests that there is a discernible human influence on global climate". It is, in fact, a much more equivocal statement than it appears at first sight. It is exquisitely poised language, which would allow any self-respecting scientist a large amount of latitude to distance herself on the basis of where she stood in the weighing of the evidence, how strongly she used the verb `suggest' and what she was prepared to interpret as `discernible'.

I would simply ask whether drawing a conclusion in these terms added any value. As I have already indicated, the policy issue is not whether we should seek to reduce the future trajectory of emissions on the basis of current evidence, but rather an assessment of the future risk of climate change. From my reading of the debate, the range of any likely warming has not changed markedly for some time (although the high end estimates have fallen somewhat).

Telling policymakers that there was a `discernible' anthropogenic signal in 1995 didn't add anything to our understanding of the risk. The case for taking out `insurance' did not depend on that. By publishing that sentence, Working Group I - perhaps unwittingly - pandered to the political appetite for `evidence' that I spoke of earlier. Certainly, it was seized on as if it had added something that I don't believe it did.

Scientists have to be prepared to tell politicians what they can reasonably ask for and what they must wait for (and to his credit Professor Bolin, the IPCC's first chairman, was prepared to do just that). Politicians in turn must accept that their job is to weigh up the costs and benefits of action in the absence of definitive information. That is what risk management is all about. They might also consider going into the matter a little more deeply than they sometimes do. I find myself, all too frequently, cringing in the presence of a colleague stating as facts matters for which there simply isn't unequivocal evidence. In 1995 they needed to look no further than the Technical Summary of Working Group I's report.

I am looking forward to the Third Assessment Report eagerly. The IPCC - as a process for drawing together widely dispersed knowledge - is breaking new ground and it is easy to be critical of it. I hope its critics in the scientific community will join it and see to it that we can be satisfied with the process. Providing it can express its findings as to scale in terms of ranges, and its forecasts in terms of probabilities, it should be able to improve the quality of the policy debate about the nature, the pace and the timing of our response to the risks we face. Even so, the climate change debate will probably remain the most difficult environmental risk management challenge of our time.

Conclusion

May I, in closing, thank the AAAS again for the privilege of allowing me to deliver this inaugural Robert C Barnard lecture. That a distinguished scientific body such as this should have reached into the political community - and one beyond these shores - to speak today demonstrates an openness on your part that is to be commended. I hope, in return, that the political community will be as prepared to listen to scientists on their terms. We have different roles and understanding those differences is the key to the advancement of science (your goal) in a socially, economically and environmentally responsible way (my goal).

1Soren Holm and John Harris at the Institute of Medicine, Law and Bioethics at the University of Manchester, Nature vol 400, 29th July 1999 29