de Wit, Elisa; Luck, Kai --- "The emerging threat of climate-related litigation for company directors" [2020] PrecedentAULA 11; (2020) 156 Precedent 45

---

THE USE OF ‘BEST AVAILABLE SCIENCE’ IN ENVIRONMENTAL AND NATURAL RESOURCES LAW

By Dr Bruce Lindsay

Environmental and natural resources management (NRM) depend on science to inform decision-making and action. This reliance operates in the context of complex and fragmented legal schemes, directed to specific domains of NRM and/or environmental protection, such as biodiversity conservation, water management, forestry, or climate change mitigation.

The focus of law, policy and science in these domains is management of public resources – such as air, water, land use, sub-surface minerals, and flora and fauna – under the supervisory authority of the state. These public resources are to be managed for the overall public benefit in an ‘ecologically sustainable’ manner, which is to say prudently, for the benefit of present and future generations.^[1] Ecologically sustainable development (ESD) itself relies on science to inform the operation of constituent principles, such as integrated decision-making, the precautionary principle, or the internalisation of environmental costs.

But even in the ordinary constellation of environmental and NRM decisions, scientific knowledge and methods are essential. These decisions include the granting of statutory permits, consents or approvals, as well as the preparation and implementation of strategic plans, policies and programs.

In this context, little regard is had typically, in legislative guidance or mandate, as to how science is to be integrated into the decision-making process or what norms or principles should underpin that science. Decision-making can be straightforward, such as in the grant of an uncontroversial licence or approval, or protracted and complex, where there are elaborate and contested resource planning or approvals. The Basin Plan 2012^[2] and strategic assessments under the Environment Protection and Biodiversity Conservation Act 1999 (Cth)^[3] are examples of the latter.

Legislators or regulators may regard the integration of science into ordinary administrative decision-making processes as axiomatic. Science provides key knowledge inputs, governed by objective, agreed, transparent and rational procedures. However, scientific knowledge does not provide infallible solutions. It is susceptible to limits of quality, rigour, uncertainty, error, and, on occasion, incursion of interests into the scientific model of disinterested objectivity.

In this light, express legislative mandate for the use of ‘best available science’ in environmental actions is remarkable both for its infrequency and for its potentially influential contribution to good governance and decision-making.

‘BEST AVAILABLE SCIENCE’ IN US AND AUSTRALIAN ENVIRONMENTAL STATUTES

The legislative obligation to use ‘best available science’ in decision-making under environmental statutes is most famously associated with the operation of the US Endangered Species Act (ESA).^[4] This ‘mandate’^[5] arises at various decision-making points under the ESA, such as endangered species determinations,^[6] critical habitat designation^[7] and control of federal agency actions that may jeopardise the survival of listed species.^[8] The precise formula of the requirement to use ‘best available science’ varies.^[9] For example, determination of endangered species status is ‘solely on the basis of the best scientific and commercial data available to [the Secretary]’. Progressively, some form of the ‘best available science’ mandate has rippled through US natural resources and environmental law, and through legal practice more widely.^[10]

In European Union law, there are similar requirements to incorporate scientific standards and knowledge into decision-making. Examples include the Water Framework Directive^[11] (‘the Community is to take account of available scientific and technical data’) and Habitat Directive (‘on the basis of relevant and reliable scientific information’).^[12] Again, the precise wording and formulae of provisions requiring incorporation of science into decision-making varies.

In Australia, it is rare for environmental and resources legislation to include express ‘best available science’ mandates. The most prominent example of a ‘mandate’ in Australian law arises under s21 of the Water Act 2007 (Cth) (Water Act), which obliges the Murray Darling Basin Authority and federal water resources minister to ‘act on the basis of the best available scientific knowledge and socio-economic analysis’^[13] in the preparation of the Basin Plan. Another example is Victorian planning law relating to bushfire protection.^[14] In other instances, the provisions do not amount to obligations to act on scientific knowledge or information. Variations of these provisions operate under marine and coastal legislation (‘best available and relevant environmental, social and economic understanding’)^[15] and Yarra River legislation (‘best practicably available information’).^[16] In administering the Environment Protection and Biodiversity Conservation Act 1999 (Cth) (EPBC Act), the federal Environment Minister is required to consider scientific advice in relation to certain actions, such as species listing, making plans and approving actions.

Two threads run through these examples. First, there is variation in the formulae used to express the scientific standard. Secondly, there are clear disparities around the use and legal effect of science as a basis for a standard of administrative or regulatory conduct.

‘BEST AVAILABLE SCIENCE’ MANDATE AND GOOD GOVERNANCE

A key benefit of using a ‘best available scientific knowledge’ standard combined with a legislative obligation rather than a discretion is its potential to contribute to good governance in resource management through transparent evidence-based decision-making.

While not an answer to poor, interest-driven or corrupt decision-making, elevating a scientific standard to a consideration upon which an administrative power must be exercised goes some way towards addressing issues of independence and rationality in decision-making. As Holly Doremus remarked, the mandate was ‘generally intended to ensure objective, value-neutral decision making by specially trained experts’.^[17] This does not give science an idealised status, but rather gives scope for environmental and NRM regulation to have to deal with a distinctive and separate source of authority, one that lies in scientific rationality.

‘Best’ science

Administrative action which is subject to a duty to act on the basis of science alone carries a risk of the science ‘providing a fig leaf’ to poor decision-making, whether as a consequence of interests, haste, lack of data (facts), lack of proficiency (competent opinion), or otherwise. Hence, the qualification of ‘best, available’ science.

The concept of ‘best’ science draws on attributes associated with established scientific protocols and codes, such as peer review, transparent methods of experimentation or observation, verifiable results, known analytical techniques, and reference to an accumulated body of knowledge.^[18] Wider knowledge may be relevant, such as ‘grey’ literature or other ‘suggestive’ or ‘supplementary’ sources of information.^[19] ‘Best’ may be context-specific, depending on the state of knowledge and enquiry in relation to the subject matter involved.^[20]

The implication of this is that ‘best’ scientific knowledge should demonstrate (as fully as can be achieved) rigour and credibility of process or technique, and depth or richness of its factual basis. It is interesting to observe that the US Supreme Court, interpreting the phrase in the ESA, found that it established only a minimal standard, akin to rationality, ‘to ensure that [that Act] is not implemented haphazardly, on the basis of speculation and surmise’.^[21]

‘Available’ science

Scientific processes are dynamic and iterative, with scope for new knowledge to emerge and existing knowledge to be falsified or disproved. Decision-making, however, is not open-ended but time-bound, often strictly so. There are pragmatic considerations to be made on the integration of knowledge into administrative action. The mandate therefore has been interpreted as ‘best available’ rather than ‘best possible’ scientific knowledge, not necessarily establishing a positive duty to collect further data or information, although superior evidence cannot be ignored.^[22]

‘Best available science’ includes consideration of the procedural context and end-users of the knowledge.^[23] Academic or research science works conscientiously to the scientific norms and it is ‘expected to be self-correcting over the long-term’.^[24] In the administrative context, ‘best available science’ informs specific decisions, policy-making or planning. Science is applied with potentially variable timeframes. Lastly, scientific knowledge can be used in the ‘short-term project’^[25] of the judicial process, by way of expert evidence.

In the administrative or judicial contexts, the scientific ‘inputs’ are typically distilled in reports of specialists and experts. For the regulatory process in particular, the focus is on the operationalising of scientific knowledge for the purposes of decision-making, policy or planning. This has been a key focus of the ‘best available science’ mandate, as the ESA and Water Act examples above indicate.

A potential tension that arises here is the confining or ‘packaging’ of scientific information for the sake of expediency or efficiency in decision-making. One increasingly common approach to managing and ‘operationalising’ scientific knowledge is to integrate knowledge through ‘decision support tools’.^[26] These tools include datasets and models synthesised into knowledge ‘platforms’, such as digital information systems, employed at the assessment stages of decision-making.^[27]

‘Best’ vs ‘available’ science?

This approach to decision-making informed by science is susceptible to a tension between ‘best’ science and ‘available’ science. ‘Information systems’ may lack transparency or be of variable quality. For example, data inputs and outputs may be clear, but information processing rules, models or assumptions may not be known or transparent.^[28] To qualify as scientific, in addition to ‘best’ scientific knowledge, such tensions must favour transparency and procedures for verification and testing.

Assessment methods based on ‘information tools’ must also expressly acknowledge uncertainties and the potential for adaptation of knowledge in the face of deficiencies and improvements in data, models, hypotheses and paradigms: ‘[m]anaging natural resources is ultimately about managing uncertainty’.^[29] A crucial legal response to science ‘operationalised’ in decision-making therefore is the precautionary principle.^[30]

This potential tension between ‘best’ and ‘available’ science suggests also that there may be only certain classes of decision or action to which a ‘best available science’ mandate should be applied. Are there relatively routine decisions to which, subject to a precautionary approach, the full rigour and efficacy of ‘best’ science is excessive? Are there more strategic decisions to which the legal mandate is better applied? The construction of laws such as the ESA and the Water Act would suggest so. Not all decisions are susceptible of the mandate but key ones are, such as threshold determinations (for example, listing of endangered species) and preparation of strategic plans governing management of water resources. Notwithstanding imperfections, a fuller, more robust application of science to these decisions seems an appropriate focus of resources and effort.

A failure to execute the science mandate

The water management arrangements for the Murray Darling Basin is a spectacular example of a failure to execute the law based on a scientific mandate. This failure was recognised in the 2019 Report of the South Australian Royal Commission into the Murray Darling Basin.^[31] The Commission undertook a wide-ranging review of the Water Act and its implementation in light of allegations of water mismanagement, including illegal taking of water. The Commissioner’s inquiry included detailed consideration of the meaning and function of ‘best available scientific knowledge’ in the Act.^[32]

The Water Act is primarily an environmental statute. Its key tasks include setting ‘sustainable diversion limits’ (SDLs) for the entire Basin, as well as for regional water systems within the Basin. These are to be given effect through the Basin Plan. SDLs are to reflect an ‘environmentally sustainable level of take’ (ESLT), requiring winding back of over-allocation of water and returning water to the environment for protection and restoration of water ecosystems. The setting of SDLs, the ESLT, and other aspects of water planning under the Act must occur on the basis of ‘best available scientific knowledge’. That knowledge base is, as the Commissioner expressed it:

‘neither secret nor classified. It is available to the scientific community, and the broader public. It involves processes and actions that represent science – that is, that are capable of being reviewed, checked and replicated.’^[33]

The Commissioner found that, in key respects, actions were not taken on that basis, including the setting of the Basin-wide SDL,^[34] planning on the basis of climate change projections,^[35] formulating SDL ‘adjustment mechanisms’,^[36] and the Northern Basin Review.^[37] Consequently, he found, those actions were not lawful.

Because the law relies on the science, decisions made and actions taken under the Water Act must be accountable, verifiable and disclosed. As the Commissioner put it, this is in the nature of science itself: ‘Science itself demands disclosure. Research, experimentation and decision-making are not science if they cannot be fully tested, and either validated or invalidated’.^[38] The Commissioner also drew parallels between this approach and the scrutiny inherent in representative democratic institutions.^[39]

Revision of ‘science’?

Finally, the meaning of ‘science’ in the ‘best available science’ mandate may seem self-evident, but some comment should be made on the limits of the conventional (Cartesian) scientific model used in decision-making. That model is based on a binary relationship between humans and nature, as subject and object, assuming human authority over nature.

One emerging source of revision of this epistemic model comes from the use and accommodation of Indigenous knowledge systems. Indigenous cultural knowledge systems^[40] contain features familiar to, and intersecting with, Cartesian science. Commonalities include ‘classifying, inferring, questioning, observing, interpreting, predicting, monitoring, problem-solving and adapting’.^[41] There are very significant epistemic differences as well, such as long-term cyclical knowledge, oral transmission, ‘high context’ perspective, relational rather than binary models, and close integration with spiritual and ethical-legal codes.^[42] Integrating conventional science and Indigenous knowledge systems occurs in certain environmental and NRM contexts, such as water resource assessments that inform planning.^[43] Statutory imperatives to integrate can be found, for instance, in obligations (albeit weak) on water planners to identify ‘Indigenous uses and values’.^[44]

The significance of ‘Indigenous science’ in knowledge systems informing ‘best available science’ is its capacity both to challenge and to extend the domains and the standards of science applied to statutory decision-making and planning. It is an opportunity to test scientific knowledge claimed as ‘best’ and ‘available’ by reference to additional sources of intellectual authority. In so doing, revision of ‘best available science’ is likely to include a broader challenge to sources of social power that are implicit in this concept. The outcome may well be, however, the strengthening of ‘science’ and the environmental and NRM governance based on it.

Dr Bruce Lindsay is a lawyer at Environmental Justice Australia. PHONE [phone] EMAIL [e-mail].

---

^[1] Telstra Corporation Ltd v Hornsby Shire Council (2006) NSWLEC 133, [107]–[124].

^[2] Basin Plan 2012, prepared under Water Act 2007 (Cth) (Water Act), pt2, div1.

^[3] Environment Protection and Biodiversity Conservation Act 1999 (Cth), pt10.

^[4] Endangered Species Act 16 USC 35 § 1531ff.

^[5] H Doremus, ‘The purposes, effects and future of the Endangered Species Act’s best available science mandate’, Environmental Law, Vol. 34, 2004, 397.

^[6] Endangered Species Act 16 USC 35 § 1533(b)(1)(a).

^[7] Endangered Species Act 16 USC 35 § 1533(b)(2).

^[8] Endangered Species Act 16 USC 35 § 1536(a)(2), 1536(c).

^[9] M Brennan, D Roth, M Feldman and A Greene, ‘Square pegs and round holes: Application of the “best scientific data available” standard in the Endangered Species Act’, Tulane Environmental Law Journal, Vol. 16, 2003, 387 at 400–2.

^[10] Ibid, 402–4; B Esch, A Waltz, T Wasserman and E Kalies, ‘Using best available science information: Determining best and available’, Journal of Forestry, Vol. 118, 2018, 473.

^[11] Directive 2000/60/EC, (12).

^[12] Directive 92/43/EEC, art 5(1).

^[13] Water Act, sub-s21(4)(b).

^[14] Victorian Planning Provisions, cl 13.02-1S.

^[15] Marine and Coastal Act 2018 (Vic), s11.

^[16] Yarra River Protection (wilip-gin Birrarung murron) Act 2017 (Vic), sub-s8(2).

^[17] Doremus, above note 5, 419.

^[18] Doremus, above note 5; Esch et al, above note 10; D Murphy and P Weiland, ‘Guidance on the use of best available science under the US Endangered Species Act’, Environmental Management, Vol. 58, 2016, 1; PJ Sullivan et al, ‘Defining and implementing best available science for fisheries and environmental science, policy and management’, Marine Sciences Faculty Scholarship, 2006, 30; D Ryder, M Tomlinson, B Gawne and G Likens, ‘Defining and using “best available science”: A policy conundrum for the management of aquatic ecosystems’, Marine and Freshwater Ecosystems, Vol. 61, 2010, 821; G Bisbal, ‘The best available science for the management of anadromous salmonids in the Columbia River Basin’, Canadian Journal of Fisheries and Aquatic Sciences, Vol. 59(12), 2002, 1952.

^[19] Bisbal, ibid.

^[20] Ryder et al, above note 18, 823.

^[21] Bennett v Spear [1997] USSC 24; 520 US 154, 176 (1997).

^[22] J Tavener Holland, ‘Regulatory Daubert: A panacea for the Endangered Species Act’s “best available science” mandate?’, McGeorge Law Review, Vol. 39, 2008, 299 at 303 and cases cited therein.

^[23] Doremus, above note 5, 409–17.

^[24] Ibid, 411.

^[25] Ibid, 412.

^[26] See, for example, K Bagstad et al, ‘A comparative assessment of decision support tools for ecosystem services quantification and valuation’, Ecosystem Services, Vol. 5, 2013, 27.

^[27] See, for example, Department of Environment, Land, Water and Planning (Vic), ‘Natureprint and Strategic Management Prospects’, <https://www.environment.vic.gov.au/biodiversity/natureprint>.

^[28] This obscurity of data processing was referred to as a ‘black box’ by the Victorian Civil and Administrative Tribunal in relation to the production of quantified values for biodiversity used in digital biodiversity assessment ‘tools’ under native vegetation clearing regulations: see Watermark Village Pty Ltd v Mornington Peninsula SC [2016] VCAT 1853, [74]–[75], [99].

^[29] Ryder et al, above note 18, 825.

^[30] SA, Royal Commission into the Murray Darling Basin (2019) <https://www.mdbrc.sa.gov.au/>, 151–2; Telstra Corporation Ltd v Hornsby Shire Council (2006) NSWLEC 133, [125]–[183].

^[31] SA, ibid.

^[32] Ibid, 150–2.

^[33] Ibid, 53, Finding 3.4.

^[34] Ibid, 54, Findings 5.2–5.7.

^[35] Ibid, 55, Findings 6.1–6.11.

^[36] Ibid, 57–8, 7.4–7.8.

^[37] Ibid, 62–3, Findings 10.2–10.6.

^[38] Ibid, 69, Finding 18.3.

^[39] Ibid, 69, Finding 18.4.

^[40] See, for example, E Ens et al, ‘Indigenous biocultural knowledge in ecosystem science and management: Review and insight from Australia’, Biological Conservation, Vol. 181, 2015, 133.

^[41] J Johnson, R Howitt, G Cajete, F Birkes, R Paulani Louis and A Kliskey, ‘Weaving indigenous and sustainability sciences to diversify our methods’, Sustainability Sciences, Vol. 11, 2016, 1 at 5.

^[42] Ibid, 5, 8.

^[43] MLDRIN, NBAN and NAILSMA, National Cultural Flows Research Project, Components 2–4, <http://culturalflows.com.au/index.php?option=com_content & view=article & id=38 & Itemid=131> G Tipa and L Teirney, A cultural health index for streams and waterways: A tool for nation-wide use (Report prepared for the Ministry for the Environment, 2006) <https://www.mfe.govt.nz/sites/default/files/cultural-health-index-for-streams-and-waterways-tech-report-apr06.pdf>.

^[44] Basin Plan 2012 (Cth), ch10, pt14.