By Alan Randall
The precautionary principle is fundamentally a claim that acting to avoid or mitigate threats of serious harm should be accorded high priority in public policy. Over the last three decades, governments and international bodies have endorsed it in principle, and some of them have incorporated it into some areas of policy practice. Yet, the precautionary principle is controversial in policy circles, public discussion, and scholarly discourse.
Here I review precautionary principle literature from the perspective of economics, where the tendency is to see theory and methods for rational decision-making under risk and uncertainty as well established and risk management tools as well developed, and to view the precautionary principle with some circumspection. The overarching objective of this review, parts of which are encapsulated in this essay, is to make some progress toward answering the following question: Can a precautionary principle be defined that is meaningful and coherent and offers a distinct contribution to the policy and management arsenal for dealing with the threat of harm?
Ordinary risk management
For the purposes at hand, ordinary risk management is defined to include all of the standard economic apparatus of decision theory and risk management. Relevant theoretical concepts for decision-making under risk and uncertainty include benefit-cost analysis and risk-benefit analysis, as well as other utilitarian approaches including the assignment of monetary values to environmental assets. At the heart of all these approaches, uncertain future prospects are reduced to expectations of value or utility, and expected costs are weighed against expected benefits. The toolkit for managing risks may still be organized into Ehrlich and Becker’s (1972) classic categories of market insurance, self-insurance, and self-protection. Self-insurance is mostly a matter of diversifying one’s risk exposure and setting aside resources sufficient to survive losses should bad outcomes eventuate – being rich offers serviceable protection from many threats. Self-protection involves actions and investments that help avoid or mitigate bad outcomes.
There are at least three problems with ordinary risk management. First, it is best adapted to everyday sorts of risks, as opposed to catastrophes. Second, it makes much more sense of insurable risks than those that are uninsurable; for example, it is much better adapted to risks that strike randomly than to systemic risks. Third, and quite a different sort of problem, it relies on risk assessments that assume the underlying systems are known and regular, as opposed to systems that can and do surprise us.
The role of principles in policy and management
What reasonably can, and cannot, be expected of a principle; and how do principles differ from laws and rules? In moral reasoning, principles are general statements of normative moral positions for a class of concerns. Examples familiar to economists include "change that benefits some while harming none is morally desirable" and "each person’s autonomy is morally valued." Some principles are so compelling that they are recognized by many competing moral systems. Nevertheless, important principles will come into conflict, and resolution requires a weighing of the principles involved and the facts of the case. Yet, a principle is much more than a preference – we feel a serious moral loss when we have to compromise a principle in a particular case.
Principle has a similar meaning in law. A legal principle is not a complete theory of normative law, but must be interpreted in ways that anticipate and account for competing and conflicting principles. Agreed principles cannot be implemented directly – a principle provides a powerful argument in a particular direction but must be operationalized in laws, rules, guidelines, etc.
The precautionary principle succeeds if it serves as a guide for law and policy, in the sense that it embodies a considerable moral intuition to be consulted along with other pertinent principles and values when the issues it addresses come into play.
Criticisms of the precautionary principle
What are the major objections to the precautionary principle?
The precautionary principle is meaningless and self-defeating (Sunstein 2005).
In Laws of Fear, Cass Sunstein defines his task as to critique an extreme version of the precautionary principle: effectively, "avoid all risk" because (he claims) more moderate and circumscribed versions are uninteresting. It is true that the injunction to avoid all risk is self-defeating, because there is risk in every conceivable course of action including the no-action alternative. Nevertheless, in this critique Sunstein is simply playing tennis with the net down. Serious precautionary principle proposals are focused on risks that pass some test of seriousness and are moderated by various provisions such as where the principle applies, protection of vulnerable groups from undue burdens, and in some cases, cost-effectiveness. Moderate versions are "uninteresting" only if real life, real problems, and real solutions are uninteresting.
This Sunstein criticism is best interpreted as an injunction to take risk-risk trade-offs seriously (a concern he shares with several other authors), and to specify the precautionary principle carefully so as to direct it toward risks that are in some sense unusually serious.
The precautionary principle is logically incoherent (M. Peterson 2006).
What Peterson actually showed is that a person applying rationality axioms and facing ideal conditions familiar to economists would choose ordinary risk management rather than a precautionary restraint. This is reasoning of impregnable circularity, since ordinary risk management is exactly what one gets when one derives optimal decisions from economic rationality assumptions and ideal conditions. The scope for a meaningful precautionary principle is concentrated on the class of cases where conditions ideal for ordinary risk management are absent.
The precautionary principle contains no normative content (M. Peterson 2007).
Specifically, it cannot guide action and decision because it is not structured and used as a stand-alone rule that tells us what to do and what not to do for each possible input of information. Here, Peterson is simply demanding too much. The precautionary principle is positioned as a principle, a general statement of a serious normative position for a class of concerns. A stand-alone rule that passes Peterson’s test of completeness would be something quite different.
The precautionary principle is just one piece of ordinary risk management done right (Farrow 2004) or, conversely, all that is good in the precautionary principle tool-kit is consistent with ordinary risk management (Gollier 2001).
This claim usually takes the form of an argument that the theory of decision making under risk and uncertainty, the hallmarks of which are rationality assumptions and utilitarian decision rules, allows for aversion to risk, desire to avoid catastrophe, and corner solutions. A corner solution is a case where it is optimal to impose a quantity restriction (perhaps, but by no means always, at the zero level) on risky undertakings.
If the precautionary principle were just one piece of ordinary risk management, it would be redundant in a world where ordinary risk management was well developed and implemented comprehensively. So, to avoid redundancy, it is important for a precautionary principle framework to be distinguishable from ordinary risk management both logically and practically.
The precautionary principle is foolish because it can be invoked by unfounded panic (Sunstein 2005).
It would be hard to defend a precautionary principle with inadequate defenses against drastic remedies provoked by unfounded panic. An essential piece of the task ahead is to address the issue of knowledge and the relationship between evidence and remedy – what can be said about the kinds of evidence that would justify invoking more or less drastic precautionary interventions?
The precautionary principle is best understood as a kind of soft norm (Mandel & Gathii 2006).
A soft norm is little more than a statement of a widely honored but nonbinding standard – in our case, a non-binding injunction to precaution in the face of possible adverse consequences. It has a status perhaps akin to the Hippocratic Oath. "First, do no harm" expresses an important caveat (be alert to the harm that treatment itself may cause), but cannot be taken literally in a world where well-intentioned and prudent intervention may be intrusive and risky, and the recovery slow and painful.
Soft norms should not be trivialized – there is social and political virtue in being clear about the things we value – but a serious precautionary principle must offer more.
The precautionary principle debate is just a quarrel about who owns the null hypothesis (Parson 2000).
The issue of burden of proof is raised directly in the Wingspread Statement (1998), which asserts that the proponent of a risky action bears the burden of proof. That is, the proponent would have to prove the action is safe. In hypothesis testing, two kinds of errors are possible: a true hypothesis might be rejected, and a false hypothesis might be accepted. To protect against incorporating false statements into the body of accepted knowledge, tests are usually constructed so that it takes overwhelming empirical evidence to reject the null hypothesis (usually that the treatment has no effect). The Wingspread Statement would define the null hypothesis as "the proposed action is harmful," which assigns the burden of proof to the proponent. In contrast, US environmental policy typically makes proof of harm a prerequisite for regulatory intervention – "the (thing or act) is safe" is treated as the null hypothesis. (The major exception is endangered species legislation.)
Who bears the greater burden of proof matters, but that question is intertwined with the question of evidence and its relationship to remedy. We might expect a coherent precautionary principle to set an evidentiary standard higher than merely "crying wolf" but lower than is appropriate for admitting empirical findings to the scientific knowledge base. Then the null hypothesis (and its implicit assignment of burden of proof) will be less important. There will be work to do on both sides of the argument – establishing that (the thing or act) is potentially harmful, and making the case for its safety.
There are about twenty different formulations of the precautionary principle in circulation, so how can it be taken seriously in real-world (law and) policy? (Sandin 1999).
This complaint ignores the distinction between a principle and a road-ready rule. The on-going discourse about how the precautionary principle might apply to real-world policy and management is an essential part of the process of developing rules, policies, and practices that are informed by principles.
Similarly, discussion of the precautionary principle in the context of international trade seems mostly to proceed in legalistic terms (D. Peterson 2006). Such discussion by its nature seeks enforceable interpretations and is uncomfortable with competing definitions and inconsistencies (either internal or with other widely honored values). However, the precautionary principle cannot be expected to provide alone the clear, coherent, and enforceable laws demanded by critics in the trade literature. At best, it is just one of the touchstones for the on-going process of developing trade laws and interpretations that respect precaution in the face of certain kinds of threats and a broad slate of legitimate trade concerns.
The precautionary principle conflicts with other important values that good law and policy would surely respect, and it fails to provide clear instructions for resolving these conflicts (Lofstedt et al 2002).
Again, what reasonably can be expected of a principle? A principle captures a serious moral intuition, but enunciating a principle neither claims nor establishes its priority over other principles. In complicated exercises in law and policy, important principles will come into conflict and resolution requires a weighing of the principles involved, the values at stake, and the facts of the case.
The Sandin and Lofstedt et al critiques and Peterson’s (2007) claim that the precautionary principle is devoid of normative content all share a common misconception. Critiques of the precautionary principle as incomplete, inconsistent with other considerable principles and values, and altogether not a good stand-alone decision rule simply misunderstand the nature and role of principles in moral and legal discourse.
An approach to defining and justifying a defensible precautionary principle
The task, then, is to outline and defend a meaningful precautionary principle that is neither foolish nor redundant. Such a precautionary principle would be:
- Clearly distinct from ordinary risk management (and therefore not redundant in that sense).
- Justified for circumstances that would be recognized as extraordinary – a plausible scenario that generates an extraordinary threat (and therefore avoids at least two kinds of foolishness – the imperative to avoid all risks, and susceptibility to unfounded panic).
- Clearly positioned as a principle (not a rule). A principle is capable of commanding broad agreement and identifies considerations to be taken seriously, but is neither a universal trump that defeats other valid considerations nor a road-ready rule.
A precautionary principle that passes these tests would be able, in carefully specified circumstances, to justify remedies beyond those justified by expected welfare considerations.
The framework offered here follows the broad direction suggested by Manson (2002) and Hughes (2006) – to define a precautionary principle that explicitly relates damage conditions, knowledge conditions, and remedies. Consider an ETR framework. A precautionary principle of the following form …
If there is evidence stronger than E that an activity raises a threat more serious than T, we should invoke a remedy more potent than R.
… would focus on the key issues: what sorts of threats might invoke precaution, what sorts of remedies might be appropriate, what sorts of evidence might justify a precautionary intervention, and how might these key elements, E, T, and R, interact? Conceding at the outset the virtues of ordinary risk management in a wide range of applications, and accepting the imperative to avoid redundancy, I start by asking what kinds of threats would justify remedies stronger than risk management.
It makes sense to focus the precautionary principle on threats of disproportionate and asymmetric harm. Disproportionate addresses the magnitude of possible harm, and describes the case where the outcome set includes possibilities of harm that are very much worse than the likely beneficial outcomes are good. Asymmetric describes likelihood – specifically the more-than-trivial likelihood that the disproportionately harmful outcome will occur. These are exactly the kinds of threats that pose difficulties for ordinary risk management, which is best suited to situations where outcome possibilities are normally distributed. A focus on disproportionate and asymmetric threats responds to the commonsense admonition, "Don’t risk great harm in pursuit of modest benefit."
Some things can be said about the sorts of threat that are likely to be disproportionate and asymmetric.
They are likely to be collective threats. The precautionary principle seems best adapted to collective threats, where individual exposure is in considerable degree involuntary. This may be because the threat affects a larger system to which we are inextricably bound, or because we are not sufficiently informed to exercise meaningful choice. Individuals may face genuine choices about voluntary exposure to disproportionate and asymmetric threats, and they may respond with precautionary rules for themselves and those for whom they are responsible: avoid activities and situations that are "too" hazardous. With collective threats, opportunities for effective self-protection at the individual level are limited, which strengthens the case for a public policy response.
Certain characteristics of threat situations seem more likely to pose the possibility of asymmetric and disproportionate harm.
Complexity. The resilience of complex systems – their tendency, following perturbations, to return toward their prior trajectory; or their capacity to tolerate disturbances while retaining their structure and function – may be undermined by sudden anthropogenic shocks that precipitate regime shifts, or by sustained anthropogenic pressure that reduces resilience and increases vulnerability to regime shifts (Holling 2001, Folke et al. 2004). Because regime shifts change the system in unpredictable ways (Folke et al.), the threats introduced may be disproportionate and asymmetric. Among complex systems, the least stable and resilient are more vulnerable, and it is harder to take effective avoidance/mitigation actions for those that signal early warnings only weakly.
Novelty. New and untested phenomena (things and events) may introduce disproportionate and asymmetric threats. We have better knowledge of outcomes and/or probabilities associated with familiar phenomena and it is likely that research and experience already have eliminated the extremely harmful possibilities, which implies that the remaining risks are more likely to be manageable with standard tools. Note that novelty may pertain to two quite different threat situations:
- Novel interventions – we introduce a new technology or release an alien organism that may substantially shock the outcome set and/or its probability distribution.
- Novel outcomes from systematic overstress under business-as-usual – a familiar and sustained intervention, perhaps exploitation or pollution, may increase vulnerability to regime-shifts with unpredictable consequences. Note that things that were novel interventions in the beginning (e.g. asbestos and PCBs) can become business-as-usual, with enormous harmful consequences, if the threat is recognized too late.
Large spatial and/or temporal scale. If the effects of an action cannot be confined readily in space or in time, then the chances are exacerbated that threats will be disproportionate and asymmetric. Uncontrolled dispersion of an innovation that is novel and incompletely understood may entail the possibility of large-scale damage, delayed detection and response, and, eventually, very expensive and disruptive remediation. Confinement in space permits quarantining the innovation until we can learn more about any harm it may cause, and confinement in time permits destroying it before general release should its adverse effects raise sufficient alarm. These safeguards allow us to approach a program of innovation more confidently, because each potential innovation can be tested extensively under low-risk conditions prior to decisions about general release.
The evidence that would justify a precautionary interruption of business as usual should be scientific and credible. First, the evidence underpinning decisions should reflect accurately what is broadly accepted in the scientific community, what controversies remain active and why, and what remains uncertain or unknown. Second, science and the institutions that support it should have adequate defenses against distortions (including inflated threat-claims that may induce unfounded panic, and unwarranted insistence that there is nothing to worry about) and the cynical exploitation of cognitive biases in service of various interests.
Finally, given scientific uncertainty, how much credible evidence of threat is enough to trigger a precautionary response?
In a 1999 court case, Cellular Telephone Co. v. Town of Oyster Bay, it was ruled that local governments could regulate cellular telephone towers despite scientific uncertainty about the extent of potential harm to human health from such installations. To justify regulation, the court wrote, scientific evidence of a threat does not have to be overwhelming: the requirement is "not a preponderance but more than a scintilla of evidence." This standard is imprecise. However, it is consistent with the concept of a principle (rather than a rule). It leaves room for case-by-case judgment, considering the principles that come into play and the facts of the case. It is possible, for example, that we may decide to relate the standard of evidence to the magnitude of threatened harm – perhaps the greater the threatened harm, the stronger the case for a precautionary response despite substantial scientific uncertainty. Nevertheless, the proposed evidentiary criterion would specify a clear lower bound on acceptable evidence: scientifically credible evidence of plausible threat.
Given the objective to explore the coherence of a precautionary principle that would make a distinct contribution beyond ordinary risk management, the set of appropriate remedies must include some that are stronger than those that would be prescribed by ordinary risk management.
On the surface, the set of appropriate remedies looks much like those offered by ordinary risk management. They include prohibiting the action or business-as-usual stress that gives rise to the threat; accommodations such as avoiding, mitigating, and/or adapting to the threat; and insurance, self-insurance, and self-protection. It is scarcely surprising that the standard instruments of ordinary risk management may well have roles in a precautionary policy package. The distinctions between ordinary risk management and precautionary principle remedies are likely to be matters of timing and degree. Consistent with precaution, remedies may be implemented in advance to head off uncertain threats; and they may be calibrated to provide safety margins, themselves a matter of degree (safety margins appear in ordinary risk management policy practice, too).
A comprehensive precautionary policy is sensitive to the need for:
- Early warning, to reduce the costs of remedies and increase their chances of success. Early warning is likely to be enhanced by research and investment (Bussiere and Fratcher 2008, Brock and Carpenter 2006).
- Scope for learning, to reduce ignorance, narrow the uncertainties, better specify the possible outcomes and probabilities, and develop and test strategies for avoidance, mitigation, and adaptation – in the best case, gross ignorance may be reduced to manageable risk (Sinha et al 2008).
- Affordable remedies, which are helpful in building and sustaining coalitions in support of solutions among contemporaries, and inter-generational commitments to solutions (UNESCO 2005).
- Specific policies to reduce or mitigate disproportionate negative impacts of precaution on particular locations and/or socio-economic groups (UNESCO 2005, Dickson 1999).
For novel interventions, a protocol of quarantine and stepwise release serves as a metaphor for a family of related remedies that incorporate many of these considerations. Quarantine and stepwise release is feasible if the effects of an innovation can be confined in space and time. Beginning with testing under laboratory conditions, the innovation goes through a stepwise release process with monitoring, study, reassessment of the threat and the remedy, and adjustments in remedy as warranted by emerging evidence, at every step – a process that may be iterated perhaps many times. This process cannot eliminate all risk – as it proceeds iteratively toward general release, capacity to confine the effects is diminished, so there remains some possibility that harmful effects emerge too late in the QSR process. But if all goes well, step by step, prohibitions may become pauses, precautionary accommodations may turn out to be temporary, and levels of insurance and self-protection may be adjusted as more is learned. Precaution is directed toward those innovations that present unusual threats, and is therefore a lesser impediment to innovation in general.
Novel outcomes from business-as-usual stresses may pose greater challenges. Various authors have linked the precautionary principle and sustainability (Deblonde and Du Jardin 2005). Economists often assume that there is a broad domain in which business-as-usual stresses can be managed by systematic implementation of policies to correct market failure. This involves getting the prices right throughout the economy – to internalize externalities, optimize provision of public goods, and the like – and through time to ensure that resource rents are re-invested. Such business-as-usual policies and institutions may be adequate for many natural resource and environmental sustainability problems, but I have argued (2008) that in particular cases that are especially difficult – for Australia, I identified climate change, endangered species and ecosystems, and dryland salinization – precautionary remedies may fail a conventional benefit-cost test but be justified nevertheless by the disproportionate potential harm from worst-case outcomes.
The hardest cases are risk dilemmas (Jablonowski 2007), where the threat has gotten so far out of hand that potential harm and costs of remedy are both enormous: we are doomed if we do and doomed if we don’t, the time for precaution is long past, and we fall back on benefit-cost thinking because the threat, while huge, cannot be considered disproportionate. There is on-going debate about the extent of the threat of global climate change and the costs of interventions. Weitzman (2009) offers a novel statistical "fat tails" argument to justify drastic measures in this case because, he argues, the level of threat is sufficiently great that the costs we are willing to bear for reducing greenhouse gas emissions should be very high. However, some observers argue that real reductions can be obtained rather cheaply and produce an array of environmental and health benefits in addition to greenhouse reduction that help offset the cost (Quiggin 2007). If Quiggin is right, climate change is not yet a risk dilemma, and precautionary thinking would recommend implementing the sorts of actions he discusses.
From Principle to Policy and Management Practice
I would argue that there is scope for a precautionary principle that addresses squarely the cases that expose the weaknesses of ordinary risk management: threats of relatively unlikely but high damage events. In some of these cases, the decision context is rendered more than ordinarily intractable by complexity, the constraints imposed by choosing one path rather than another, and extreme uncertainty or gross ignorance. This section suggests, but does not elaborate, a framework for implementing such a precautionary principle in a society that is governed with the consent, at least implicit, of the citizenry.
Moral foundations for the precautionary principle
First, the society would need to be persuaded to adopt the precautionary principle as a normative principle. The precautionary principle can be grounded morally in prior, widely accepted moral intuitions. The case for invoking the precautionary principle hinges on a scientifically credible threat of disproportionate and asymmetric harm. Moral principles that might support such a precautionary principle include the following variations on the general ideas that risk-taking should not be disproportionate to the expected gain, and that exposure to risks should be distributed roughly the same way as opportunity to gain:
One principle among many
- Do not sell out something unique and potentially very valuable for modest gain.
- Do not take inordinate risks for modest gain.
- Do not impose big risks on the public for modest private gain.
- Do not impose big risks on the future for modest immediate gain.
The precautionary principle is at best one among many principles, values, and commitments that society might take seriously. Calling it a principle does not grant it priority over other principles – in difficult policy cases, important and widely held principles, values, and commitments will come into conflict. Together, a set of agreed principles provides a framework for policy resolution that anticipates and accounts for competing and conflicting principles.
To pursue one example a little further, a framework for addressing issues of environmental sustainability should articulate clear principles and take them seriously in practice. A plausible set of commitments might include the following:
A precautionary principle that is both substantive and procedural
- Commitment to sustaining, at least, the well-being of present and future people, and economic and political arrangements that encourage it ("business-as-usual done right")
- Commitment to precautionary exceptions in cases that exceed a threshold of concern (e.g., a scientifically credible threat of disproportionate and asymmetric harm)
- Commitment to precautionary principle procedures that encourage learning, to reduce ignorance, narrow the uncertainties, better specify the threat and the possible outcomes and probabilities, and develop and test strategies for avoidance, mitigation, and adaptation
- Commitment to transparent and inclusive public decision processes
- Commitment to equitable sharing of burdens – especially burdens undertaken to benefit future generations
Attention should be given to integrating precautionary principle and ordinary risk management procedurally. The procedurally integrated process would incorporate all the tools of ordinary risk assessment and management. The precautionary principle perspective would be reflected in processes with these attributes:
- Informed by science and structured purposefully to encourage research, early warning, learning about the threat, and development and testing of strategies for avoidance, mitigation, and adaptation;
- Attentive to the virtues of affordable remedies and policies to reduce or mitigate disproportionate negative impacts;
- Dynamic and iterative, as in the case of quarantine and stepwise release;
- Open to public participation and scrutiny; and
- Incorporating complex systems thinking and alertness to the possibility of surprises.
The fundamental framework proposed here would make business-as-usual decisions according to familiar welfare criteria implemented in ordinary risk management, while providing precautionary exceptions for particular, scientifically credible threats of disproportionate and asymmetric harm. The framework would respect both the modern experience of technical progress and increasing welfare even as change in what we consume and how we produce it proceeds apace, and the reasonable instinct for caution as we continue to push at the frontiers of what can be known about our planet’s capacity to support future welfare.
Bussiere, M. and M. Fratzcher. 2008. Low Probability, High Impact: Policy Making and Extreme Events. J. Policy Modeling 30:111-121.
Deblonde, M. and P. Du Jardin. 2005. "Deepening a precautionary European policy." Journal of Agricultural & Environmental Ethics 18(4): 319-343.
Dickson, B. 1999. "The Precautionary Principle in CITES: A Critical Assessment." Natural Resources Journal 39(2): 211-28.
Ehrlich, I. and G. Becker, 1972. Market Insurance, Self-Insurance, and Self-Protection. Journal of Political Economy 80(4): 623-648.
Farrow, S. 2004. Using Risk Assessment, Benefit-Cost Analysis, and Real Options to Implement a Precautionary Principle. Risk Analysis 24(3):727-735.
Folke, C., S. Carpenter, B. Walker, M. Scheffer, T. Elmqvist, L. Gunderson, and C.S. Holling. 2004. Regime Shifts, Resilience, and Biodiversity in Ecosystem Management. Annual Review of Ecology, Evolution, and Systematics 35:557–81.
Gollier, C. 2001. Should we beware of the precautionary principle? Economic Policy 16:301-328.
Holling, C. S. 2001. Understanding the Complexity of Economic, Ecological, and
Social Systems. Ecosystems 4: 390–405.
Hughes, J. 2006. "How not to criticize the precautionary principle". Journal of Medicine and Philosophy 31:447-464.
Jablonowski, M. 2005. "High-Risk Decisions When Probabilities Are Unknown (or Irrelevant)" Risk Management 7: 57–61..
Mandel, G. and J. Gathii. 2006. "Cost-benefit analysis versus the precautionary principle: beyond Cass Sunstein’s Laws of Fear". University of Illinois Law Review 2006(5):1037-1079.
Manson, N. 2002. "Formulating the precautionary principle". Environmental Ethics 24: 263–274.
Peterson, D. C. 2006 "Precaution: principles and practice in Australian environmental and natural resource management" Australian Journal of Agricultural and Resource Economics 50:469-489.
Peterson, M. 2006. "The Precautionary Principle is Incoherent" Risk Analysis 26: 595–601.
Peterson, M. 2007. "Should the Precautionary Principle Guide our Actions or our Beliefs?" Journal of Medical Ethics 33: 5–10.
Quiggin, J. 2007. "Complexity, climate change and the precautionary principle". Environmental Health, 7 3: 15-21.
Randall, A. 2008. "Is Australia on a Sustainability Path? Interpreting the Clues," Australian Journal of Agricultural and Resource Economics 52:77-95.
Sandin, P. 1999. "Dimensions of the Precautionary Principle". Human and Ecology Risk Assessment 5:889-907.
Sinha, P., C. Kriegner, W. Schew, S. Kaczmar, M. Traister and D. Wilson. 2008. "Regulatory policy governing cadmium-telluride photovoltaics: A case study contrasting life cycle management with the precautionary principle." Energy Policy 36(1): 381-387.
Sunstein, C. 2005. Laws of Fear: Beyond the Precautionary Principle, Cambridge: Cambridge University Press.
UNESCO. 2005. The Precautionary Principle. World Commission on the Ethics of Scientific Knowledge and Technology. Paris.
Weitzman, M. L. 2009. "On modeling and interpreting the economics of catastrophic climate change". Review of Economics and Statistics 91(1): 1-19.
Wingspread Statement on the Precautionary Principle. 1998.