Federal Court of Australia

Minister for the Environment v Sharma [2022] FCAFC 35

Appeal from:

Sharma by her litigation representative Sister Marie Brigid Arthur v Minister for the Environment [2021] FCA 560

Sharma by her litigation representative Sister Marie Brigid Arthur v Minister for the Environment (No 2) [2021] FCA 774

File number:

VID 389 of 2021

Judgment of:

ALLSOP CJ, BEACH AND WHEELAHAN JJ

Date of judgment:

15 March 2022

Catchwords:

NEGLIGENCE – representative proceeding on behalf of Australian children under 18 against the Minister for the Environment – threat of global warming and climate change to the world and mankind – novel duty of care – declaration that Minister owed Children duty of care when exercising power under ss 130 and 133 of the Environment Protection and Biodiversity Conservation Act 1999 (Cth) (EPBC Act) to approve or not approve extension of a coal mine to take reasonable care to avoid causing personal injury or death arising from emissions of carbon dioxide and subsequent global warming – declaration of duty disaggregated from breach and damage – nature of and correct approach to salient features analysis – whether sufficient relationship of neighbourhood between Minister and Children under the EPBC Act – duty of care analysis must begin with relevant statute – purpose and scope of EPBC Act and nature of statutory power in context of the inter-governmental arrangements for protection of environment in Australia – whether duty of care throws up for consideration at breach matters unsuitable for judicial determination – whether duty concerns matters of “core policy” – operational/policy distinction – nature of coherence within the law of negligence –whether duty is incoherent with the Minister’s statutory discretion under the EPBC Act – whether relevant harm reasonably foreseeable – nature of relationship between reasonable foreseeability and causation – consideration of law of causation in Australia – whether sufficient closeness and directness between Minister’s statutory power and the likely risk of harm to respondents and members of class – consideration of Minister’s control, responsibility and knowledge in relation to foreseeable harm – extent of children’s legal vulnerability to feared harm from exercise of Minister’s statutory power – parens patrie jurisdiction – whether potential liability indeterminate

NEGLIGENCE – challenge to five findings of fact by primary judge concerning catastrophic risks of global warming caused by greenhouse gas emissions – where findings based on unchallenged expert evidence led on a final basis – where Minister made forensic decision not to cross-examine expert or lead responsive evidence – factual findings open on unchallenged evidence

STATUTORY INTERPRETATION – whether human safety is an implied relevant mandatory consideration for exercise of power under ss 130 and 133 of the EPBC Act – mandatory consideration not supported by text, context or purpose of EPBC Act

Legislation:

Constitution

Australian Capital Territory (Self-Government) Act 1988 (Cth)

Australian National Registry of Emissions Units Act 2011 (Cth)

Australian Renewable Energy Agency Act 2011 (Cth)

Clean Energy Act 2011 (Cth) (repealed)

Clean Energy Finance Corporation Act 2012 (Cth)

Clean Energy Legislation (Carbon Tax Repeal) Act 2014 (Cth)

Environment Protection and Biodiversity Conservation Act 1999 (Cth)

Greenhouse and Energy Minimum Standards Act 2012 (Cth)

Judiciary Act 1903 (Cth)

National Greenhouse and Energy Reporting Act 2007 (Cth)

Product Emissions Standards Act 2017 (Cth)

Federal Court Rules 2011 (Cth)

Civil Law (Wrongs) Act 2002 (ACT)

Civil Liability Act 2002 (NSW)

Crown Proceedings Act 1988 (NSW)

Environmental Planning and Assessment Act 1979 (NSW)

Australian Capital Territory (Self-Government) Regulations 2021 (Cth)

State Environmental Planning Policy (Mining, Petroleum Production and Extractive Industries) 2007 (NSW)

State Environmental Planning Policy (State and Regional Development 2011 (NSW)

Explanatory Memorandum, Environment Protection and Biodiversity Conservation Bill 1998

Kyoto Protocol to the United Nations Framework Convention on Climate Change, opened for signature 16 March 1998, 2303 UNTS 162 (entered into force 16 February 2005)

United Nations Framework Convention on Climate Change (UNFCCC) Paris Agreement 2015

United Nations Framework Convention on Climate Change, opened for signature 4 June 1992, 1771 UNTS 107 (entered into force 21 March 1994)

Cases cited:

Adelaide Stevedoring Co Ltd v Forst [1940] HCA 45; 64 CLR 538

Agar v Hyde [2000] HCA 41; 201 CLR 552

Aiken v Municipality of Kingborough [1939] HCA 20; 62 CLR 179

Alcan Gove Pty Ltd v Zabic [2015] HCA 33; 257 CLR 1

Alchin v Commissioner for Railways (1935) 35 SR (NSW) 498

Alec Finlayson Pty Ltd v Armidale City Council (1994) 51 FCR 378

Amaca Pty Ltd v Booth [2011] HCA 53; 246 CLR 36

Amaca Pty Ltd v Ellis [2010] HCA 55; 240 CLR 111

Anns v Merton London Borough Council [1978] AC 728

Apotex Pty Ltd v Sanofi-Aventis Australia Pty Ltd [2013] HCA 50; 253 CLR 284

Australian Conservation Foundation Inc v Minister for the Environment and Energy [2017] FCAFC 134; 251 FCR 359

Baker v Carr (1962) 369 US 186

Barisic v Devenport [1978] 2 NSWLR 111

Barker v Corus UK Ltd [2006] 2 AC 572

Bass v Permanent Trustee Co Ltd [1999] HCA 9; 198 CLR 334

Bendigo and Country Districts Trustees and Executors Co Ltd v Sandhurst and Northern District Trustees and Executors and Agency Co Ltd [1909] HCA 63; 9 CLR 474

Bendix Mintex Pty Ltd v Barnes (1997) 42 NSWLR 307

Bennett v Minister for Community Welfare [1992] HCA 27; 176 CLR 408

Bennett v The Commonwealth [2007] HCA 18; 231 CLR 91

Betts v Whittingslowe [1945] HCA 31; 71 CLR 637

Bienke v Minister for Primary Industries and Energy (1996) 63 FCR 567

Boensch v Pascoe [2019] HCA 49; 268 CLR 593

Bonnington Castings Ltd v Wardlaw [1956] AC 613

Breen v Williams [1996] HCA 57; 186 CLR 71

British Celanese Ltd v A H Hunt (Capacitors) Ltd [1969] 1 WLR 959

Brodie v Singleton Shire Council [2001] HCA 29; 206 CLR 512

Brooks v Burns Philp Trustee Co Ltd [1969] HCA 4; 121 CLR 432

Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; 179 CLR 520

Bushell v Repatriation Commission [1992] HCA 47; 175 CLR 408

Caledonian Collieries Ltd v Speirs [1957] HCA 14; 97 CLR 202

Caltex Oil (Australia) Pty Ltd v The Dredge “Willemstad” [1976] HCA 65; 136 CLR 529

Caltex Refineries (Qld) Pty Ltd v Stavar [2009] NSWCA 258; 75 NSWLR 649

Cardile v LED Builders Pty Ltd [1999] HCA 18; 198 CLR 380

Carnie v Esanda Finance Corporation [1995] HCA 9; 182 CLR 398

Cattanach v Melchior [2003] HCA 38; 215 CLR 1

Chapman v Hearse [1961] HCA 46; 106 CLR 112

Chappel v Hart [1998] HCA 55; 195 CLR 232

Cole v South Tweed Heads Rugby League Football Club Ltd [2004] HCA 29; 217 CLR 469

Crane v City of New York 65 NY 2d 859 (1985)

Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; 200 CLR 1

Davis Contractors Ltd v Fareham Urban District Council [1956] UKHL 3; [1956] AC 696 

Donoghue v Stevenson [1932] AC 562

Dovuro Pty Ltd v Wilkins [2003] HCA 51; 215 CLR 317

Duyvelshaff v Cathcart & Ritchie Ltd (1973) 1 ALR 125

Espinal v Melville Snow Contractors Inc 98 NY 2d 136 (2002)

Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22; [2003] 1 AC 32

Financial Conduct Authority v Arch Insurance (UK) Ltd [2021] UKSC 1; [2021] AC 649

Flounders v Millar [2007] NSWCA 238

Fox v Percy [2003] HCA 22; 214 CLR 118

Gala v Preston [1991] HCA 19; 172 CLR 243

Gett v Tabet [2009] NSWCA 76; 254 ALR 504

Ghantous v Hawkesbury City Council [1999] NSWCA 51; 102 LGERA 399

Gifford v Strang Patrick Stevedoring Pty Ltd [2003] HCA 33; 214 CLR 269

Goldstein v Consolidated Edison Company of NY Inc 115 AD 2d 34 (1986)

Graham Barclay Oysters Pty Ltd v Ryan [2002] HCA 54; 211 CLR 540

H R Moch Co Inc v Rensselaer Water Co 247 NY 160 (1928)

Hall v Herbert [1993] 2 SCR 159

Hargrave v Goldman [1963] HCA 56; 110 CLR 40

Harriton v Stephens [2006] HCA 15; 226 CLR 52

Hawkins v Clayton (1988) 164 CLR 539

Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465

Henville v Walker [2001] HCA 52; 206 CLR 459

Hill v Van Erp [1997] HCA 9; 188 CLR 159

Hockey v Fairfax Media Publications Pty Ltd (No 2) [2015] FCA 750; 237 FCR 127

Hollis v Vabu Pty Ltd [2001] HCA 44; 207 CLR 21

Home Office v Dorset Yacht Co Ltd [1970] AC 1004

Hooper v Rogers [1975] Ch 43

Jackson v Goldsmith [1950] HCA 22; 81 CLR 446

Jackson v Harrison [1978] HCA 17; 138 CLR 438

Jaensch v Coffey [1984] HCA 52; 155 CLR 549

John Pfeiffer Pty Ltd v Canny [1981] HCA 52; 148 CLR 218

Johnson Tiles Pty Ltd v Esso Australia Pty Ltd (2003) Aust Torts Reports 81-692

King v Philcox [2015] HCA 19; 255 CLR 304

Kinsella v Gold Coast City Council [2014] QSC 65; 1 Qd R 274

Leadfree Enterprises Inc v United States Steel Corporation 711 F 2d 805 (1983)

Lipohar v The Queen [1999] HCA 65; 200 CLR 485

March v E & M H Stramare Pty Ltd [1991] HCA 12; 171 CLR 506

Mayfair Trading Co Pty Ltd v Dreyer [1958] HCA 55; 101 CLR 428

McGhee v National Coal Board [1973] 1 WLR 1

Melbourne Corporation v Commonwealth [1947] HCA 26; 74 CLR 31

Miller v Miller [2011] HCA 9; 242 CLR 446

Milliken & Co v Consolidated Edison Co 84 NY 2d 469 (1994)

Minister for Home Affairs v DUA16 [2020] HCA 46; 385 ALR 212

Minister for Immigration & Border Protection v SZVFW [2018] HCA 30; 264 CLR 541

MM Constructions (Aust) Pty Ltd v Port Stephens Council (No 7) [2012] NSWCA 417

Modbury Triangle Shopping Centre Pty Ltd v Anzil [2000] HCA 61; 205 CLR 254

Mount Isa Mines Ltd v Pusey [1970] HCA 60; 125 CLR 383

Mutual Life & Citizens’ Assurance Co Ltd v Evatt [1968] HCA 74; 122 CLR 556

National Australia Bank Ltd v Nautilus Insurance Pte Ltd (No 2) [2019] FCA 1543; 377 ALR 627

Naxakis v Western General Hospital [1999] HCA 22; 197 CLR 269

Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd (The Wagon Mound) [No 1] [1961] AC 388

Overseas Tankship (UK) Ltd v The Miller Steamship Co (The Wagon Mound (No 2)) [1967] AC 617

Patrick Stevedores Operations No 2 Pty Ltd v Maritime Union of Australia (No 3) [1998] HCA 30; 195 CLR 1

Perera v Genworth Financial Mortgage Insurance Pty Ltd (trading as Genworth) [2017] NSWCA 19; 94 NSWLR 83

Perre v Apand Pty Ltd [1999] HCA 36; 198 CLR 180

Plaintiff S156/2013 v Minister for Immigration [2014] HCA 22; 254 CLR 28

Progress and Properties Ltd v Craft [1976] HCA 59; 135 CLR 651

Pyrenees Shire Council v Day [1998] HCA 3; 192 CLR 330

Re Ditfort; Ex parte Deputy Commissioner of Taxation (1998) 19 FCR 347

Redland Bricks Ltd v Morris [1970] AC 652

Regent Holdings Pty Ltd v State of Victoria [2013] VSC 60

Roads and Traffic Authority (NSW) v Dederer [2007] HCA 42; 234 CLR 330

Roads and Traffic Authority of NSW v Refrigerated Roadways Pty Ltd [2009] NSWCA 263; 77 NSWLR 360

Roads and Traffic Authority v Royal [2008] HCA 19; 245 ALR 653

Roe v Minister of Health [1954] 2 QB 66

Romeo v Conservation Commission (NT) [1992] HCA 5; 192 CLR 431

Rosenberg v Percival [2001] HCA 18; 205 CLR 434

Rowling v Takaro Properties Ltd [1988] AC 473

San Sebastian Pty Ltd v Minister Administering the Environmental Planning and Assessment Act 1979 (1986) 162 CLR 340

SCM (United Kingdom) Ltd v WJ Whittall and Son Ltd [1971] 1 QB 337

Sean Investments Pty Ltd v MacKellar [1981] FCA 191; 38 ALR 363

Seltsam Pty Ltd v McGuiness [2000] NSWCA 29; 49 NSWLR 262

Sharma v Minister for the Environment [2021] FCA 560; 391 ALR 1

Sharma v Minister for the Environment (No 2) [2021] FCA 774

Shirt v Wyong Shire Council [1978] 1 NSWLR 631

Sienkiewicz v Greif (UK) Ltd (Willmoe v Knowsley Metropolitan Borough Council) [2011] UKSC 10; 2 AC 229

Smethurst v Commissioner of Police [2020] HCA 14; 376 ALR 575

Smith v Fonterra Co-operative Group Limited [2021] NZCA 552

Smith v Leurs [1945] HCA 27; 70 CLR 256

South Australia v The Commonwealth [1962] HCA 10; 108 CLR 130

Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973] 1 QB 27

St George Club Ltd v Hines (1961) 35 ALJR 106

State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (in liq) [1999] HCA 3; 73 ALJR 306

Stovin v Wise [1996] 2 AC 923

Strauss v Belle Realty Co 65 NY 2d 399 (1985)

Strong v Woolworths Ltd [2012] HCA 5; 246 CLR 182

Stuart v Kirkland-Veenstra [2009] HCA 15; 237 CLR 215

Sullivan v Moody [2001] HCA 59; 207 CLR 562

Sutherland Shire Council v Heyman [1985] HCA 41; 157 CLR 424

Sydney Water Corporation v Turano [2009] HCA 42; 239 CLR 51

Tabet v Gett [2010] HCA 12; 240 CLR 537

Tame v New South Wales [2002] HCA 35; 211 CLR 317

Tarkine National Coalition Inc v Minister for the Environment [2015] FCAFC 89; 233 FCR 254

Thomson Australian Holdings Pty Ltd v Trade Practices Commission [1981] HCA 48; 148 CLR 150

Thompson v Woolworths (Qld) Pty Ltd [2005] HCA 19; 221 CLR 234

Thorpe v Commonwealth (No 3) [1997] HCA 21; 144 ALR 677

Tubemakers of Australia Ltd v Fernandez (1976) 10 ALR 303

Ultramares Corp v Touche 255 NY 170 (1931)

Vairy v Wyong Shire Council [2005] HCA 62; 223 CLR 422

Victoria Park Racing and Recreation Grounds Company Limited v Taylor [1937] HCA 45; 58 CLR 479

Vozza v Tooth & Co Ltd [1964] HCA 29; 112 CLR 316

Vyner v Waldenberg Pty Ltd [1946] KB 50

Williams v Bermuda Hospitals Board [2016] UKPC 4; [2016] AC 888

Wilsher v Essex Area Health Authority [1988] AC 1074

Wintle v Conaust (Vic) Pty Ltd [1989] VR 951

Woolcock Street Investments Pty Ltd v CDG Pty Ltd [2004] HCA 16; 216 CLR 515

Wyong Shire Council v Shirt [1980] HCA 12; 146 CLR 40

X v State of South Australia (No 3) [2007] SASC 125; 97 SASR 180

X & Y (by her tutor X) v Pal (1991) 23 NSWLR 26

BM McLachlin, “Negligence Law – Proving the Connection” in MJ Mullany and AM Linden (eds), Torts Tomorrow: a Tribute to John Fleming (LBC Information Services, 1998)

Dixon J, Science and Judicial Proceedings (paper delivered on 30 September 1933)

Fleming, The Law of Torts (9th ed)

French CJ, Science and Judicial Proceedings – Seventy-Six Years On (paper delivered on 2 May 2009)

Hart and Honoré, Causation in the Law (2nd ed)

Heydon, Leeming & Turner, Meagher, Gummow and Lehanes Equity: Doctrine and Remedies (5th ed)

Spry, The Principles of Equitable Remedies (5th ed)

Stapleton, An ‘Extended But-For’ Test for the Causal Relation in the Law of Obligations (2015) 35 OJLS 697

Stapleton, Factual Causation [2010] Federal Law Review 467

Stapleton, Unnecessary Causes (2013) 129 LQR 39

Division:

General Division

Registry:

Victoria

National Practice Area:

Administrative and Constitutional Law and Human Rights

Number of paragraphs:

888

Date of last submissions:

29 October 2021

Date of hearing:

18 – 20 October 2021

Counsel for the Appellant:

Dr S Donaghue QC (Solicitor-General of the Commonwealth) with Mr S Free SC and Ms Z Maud

Solicitor for the Appellant:

Australian Government Solicitor

Counsel for the Respondents:

Mr N Hutley SC with Mr E Nekvapil, Ms K Brazenor, Ms S Brenker and Mr N Petrie

Solicitor for the Respondents:

Equity Generation Lawyers

ORDERS

VID 389 of 2021

BETWEEN:

MINISTER FOR THE ENVIRONMENT (COMMONWEALTH)

Appellant

AND:

ANJALI SHARMA AND OTHERS NAMED IN THE SCHEDULE (BY THEIR LITIGATION REPRESENTATIVE SISTER MARIE BRIGID ARTHUR)

Respondents

order made by:

ALLSOP CJ, Beach and wheelahan JJ

DATE OF ORDER:

15 March 2022

THE COURT ORDERS THAT:

1.    The appeal be allowed.

2.    Before the Court makes orders setting aside orders 1 and 3 made by the primary judge on 8 July 2021 and dismissing the application, the parties should seek to agree further orders necessary to give effect to these reasons for judgment, including any appropriate orders concerning represented parties and as to costs.

3.    Within 14 days, the parties file and serve brief written submissions annexing proposed short minutes of order dealing with costs and any further necessary or appropriate order.

4.    Final orders will then be determined on the papers and without a further oral hearing, unless a party objects to such a course in writing to the Chambers of the Chief Justice in which case the Court will consider the need for any further hearing.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

ALLSOP CJ:

Introduction

1    The threat of climate change and global warming was and is not in dispute between the parties in this litigation. The seriousness of the threat is demonstrated by the attention given to it by many countries around the world, and the attempts made by them to reach agreement and to co-operate to reduce the emission of carbon dioxide and other greenhouse gases in order to reduce the rate of increase of the Earth’s surface temperature. Those steps of international diplomacy and international co-operation in scientific matters, including research, have had the consequence that many countries and constituent political parts of countries have adjusted national and regional policy to meet the recognised threat. The debate over the appropriate steps to take at a national and international level has not been without its international and national political controversy.

2    At the outset it is important to appreciate the nature of the proceeding and the basis upon which the case was fought. The evidence led by the respondents (applicants before the primary judge) was not challenged by the appellant, whether by cross-examination or by the leading of contrary or supplementary evidence. This is a matter of some importance. It should not be seen merely as a strategic or tactical choice in a piece of inter partes litigation. This was not a demurrer procedure where the applicants’ case was to be taken at its highest for the purposes of striking out the claim. Evidence was led, on a final basis. Some objections were taken and ruled upon. There was no cross-examination. The appellant is the Minister for the Environment. The Minister is not any litigant. She is the Minister of the Commonwealth responsible, with her Department, for the very type of issue with which the Court was concerned and to which the evidence was directed. There are challenges to some of the primary judge’s findings (which should be rejected), but, by and large, the nature of the risks and the dangers from global warning, including the possible catastrophe that may engulf the world and humanity was not in dispute.

3    The Minister is responsible for decision-making under the Environment Protection and Biodiversity Conservation Act 1999 (Cth) (the EPBC Act or the Act). In the discharge of these responsibilities, the Minister is responsible to Parliament (and thereby the Australian people) for the decisions made and the policies implemented in the execution of the laws of the Parliament. The Minister’s decision-making is also subject to judicial review by the courts (the High Court of Australia under s 75(v) of the Constitution and the Federal Court of Australia under, at least, s 39B(1) and s 39B(1A)(c) of the Judiciary Act 1903 (Cth)) for the legality of the decision-making, and of the decisions made, by her. That simple, but basal, structure of responsible and representative parliamentary democracy in which the Executive is subject to the rule of law rooted in s 75(v) of the Constitution and is responsible to the Australian people through Parliament is an important part of the context for the claim. That claim is that apart from, indeed quite distinct from, the subjection to scrutiny for the lawfulness of the decision, the Minister has a personal duty, for breach of which she (and also the Commonwealth), may be found to be personally liable in damages, to take reasonable care, in the execution of her particular duties, powers and functions under ss 130 and 133 of the EPBC Act to avoid causing personal injury or death to all persons who were less than 18 years of age and ordinarily resident in Australia at the time of the commencement of the proceeding in this Court arising from emissions of carbon dioxide into the Earth’s atmosphere. Further, in so finding such a duty of care, the primary judge concluded that human safety was a distinct implied mandatory consideration in the decision about a controlled action that might endanger human safety, to be implied from the subject matter, scope and purpose of the EPBC Act.

4    In setting out at the outset the basic constitutional position of the Executive and the role of the Judiciary in pronouncing upon the legality of Executive decision-making, the potential liability of the Commonwealth and its officers in tort is to be recognised, as it is by s 75(iii) of the Constitution. Section 64 of the Judiciary Act recognises that in a suit to which the Commonwealth or a State is a party the rights of parties shall as nearly as possible be the same as in a suit between subject and subject. The law of torts applies to Ministers and the Commonwealth, as much as it applies to ordinary persons or companies. The “aspiration to equality” in s 64 (to use the words of Gleeson CJ in Graham Barclay Oysters Pty Ltd v Ryan [2002] HCA 54; 211 CLR 540 at 556 [12]) recognises within itself that “perfect equality is not attainable”: Gleeson CJ (in the same paragraph)). The nature and responsibilities of government are relevant to the operation of legal principle, here in determining whether a duty of care to avoid personal injury or death is to be recognised. The subjection of governments and public authorities, including Ministers of the Crown, to the rule of law encompasses not only the requirement of legality under s 75(v) of the Constitution, but also liability for tortious wrongs ascertained in accordance with the application of principle under the common law.

5    The EPBC Act takes its place within the federal structure under which the Commonwealth and the States and Territories have co-ordinate, and to a degree overlapping, responsibility and authority in relation to the environment. The EPBC Act, for its own part, is founded in significant part on the translation of international agreements into Commonwealth law. In this regard, it is important to keep in mind that there has been no attempt by the Commonwealth Parliament to translate those international agreements concerning climate change, in particular the Kyoto Protocol to the United Nations Framework Convention on Climate Change, opened for signature 16 March 1998, 2303 UNTS 162 (entered into force 16 February 2005) or the United Nations Framework Convention on Climate Change (UNFCCC) Paris Agreement 2015 (Paris Agreement) into Commonwealth law. In 2011, Parliament did legislate in relation to climate change issues, the central component being the Clean Energy Act 2011 (Cth). This legislation was repealed on 1 July 2014: Clean Energy Legislation (Carbon Tax Repeal) Act 2014 (Cth). See also the National Greenhouse and Energy Reporting Act 2007 (Cth); Australian National Registry of Emissions Units Act 2011 (Cth); Australian Renewable Energy Agency Act 2011 (Cth); Clean Energy Finance Corporation Act 2012 (Cth); Greenhouse and Energy Minimum Standards Act 2012 (Cth); and Product Emissions Standards Act 2017 (Cth).

6    Before the time of the hearing of the appeal, but after the primary judge’s decision and orders, the Minister made a decision and granted approval. There was discussion at the hearing as to whether the Court should receive the decision and any reasons. This course was opposed by the respondents. The Court did not receive this material.

The conclusion of the appeal in summary form

7    For the reasons that follow the Minister’s appeal against the imposition of a duty of care in the terms articulated and against the conclusion that human safety was an implied mandatory statutory consideration should be upheld. The latter implication cannot be derived from the EPBC Act. The primary judge’s conclusions in this respect were not sought to be supported by the respondents. The imposition of the duty should be rejected. First, the posited duty throws up for consideration at the point of breach matters that are core policy questions unsuitable in their nature and character for judicial determination. Secondly, the posited duty is inconsistent and incoherent with the EPBC Act. Thirdly, considerations of indeterminacy, lack of special vulnerability and of control, taken together in the context of the EPBC Act and the nature of the governmental policy considerations necessarily arising at the point of assessing breach make the relationship inappropriate for the imposition of the duty. These conclusions reflect differences of view that I have with the evaluative judgments of the primary judge in a field of contention, the imposition of a duty of care in novel circumstances, that is not without difficulty. The primary judge considered and dealt with the arguments of the respondents and the Minister in a careful, thorough and clear body of reasons.

The duty and its framing and its calling forth core policy-making and considerations unsuitable for resolution by the Judicial branch of government

8    In their application at [2] and their concise statement at [22] the respondents sought to express the posited duty at a high level of abstraction: Whether the Minister owed the respondents and the children whom they represented a duty to exercise the power under ss 130 and 133 of the EPBC Act with reasonable care not to cause the respondents harm. The duty was said to arise out of positive action, not omission. Expressed at that high level of abstraction and divorced from concrete facts and referable to any decision under ss 130 and 133 it can be seen to be of little assistance. A postulated duty of care must be stated by reference to the kind of damage that a plaintiff will suffer: John Pfeiffer Pty Ltd v Canny [1981] HCA 52; 148 CLR 218 at 241–242 (Brennan J); Sutherland Shire Council v Heyman [1985] HCA 41; 157 CLR 424 at 487 (Brennan J); Modbury Triangle Shopping Centre Pty Ltd v Anzil [2000] HCA 61; 205 CLR 254 at 262–263 [13]–[16] (Gleeson CJ); Cole v South Tweed Heads Rugby League Football Club Ltd [2004] HCA 29; 217 CLR 469 at 472–473 [1] (Gleeson CJ); see also Roads and Traffic Authority of NSW v Dederer [2007] HCA 42; 234 CLR 330 at 345 [43]–[44] (Gummow J) and Sydney Water Corporation v Turano [2009] HCA 42; 239 CLR 51 at 71 [47] (the Court).

9    The scope and content of the duty was illuminated by the expression of its anticipated breach in the concise statement at [23]: To exercise reasonable care not to cause the applicants harm by acting in a manner that materially contributes to increasing the minimum level at which carbon dioxide concentration can flatten. That expression of the matter, taken with the balance of the concise statement and the uncontested evidence presented, informs one that the duty concerns acting in connection with an approval of the extension of a coal mine in the light of the risks of global temperature warming and the consequent risks of harm to humans in the future caused by climate change, by not just the mining and transportation of the coal, but also by the emissions from the combustion of the coal mined from the extension of the mine.

10    The declaration made by the primary judge was in the following terms:

The first respondent has a duty to take reasonable care, in the exercise of her powers under s 130 and s 133 of the Environment Protection and Biodiversity Conservation Act 1999 (Cth) in respect of referral EPBC No. 2016/7649, to avoid causing personal injury or death to persons who were under 18 years of age and ordinarily resident in Australia at the time of the commencement of this proceeding arising from emissions of carbon dioxide into the Earth’s atmosphere.

11    That expression of the duty omits the express reference to “the material contribution to increasing the minimum level at which [carbon dioxide] concentration can flatten”. The primary judge’s reasons, however, demonstrate that such consideration is embedded within the duty declared and that such matter will be the very subject thrown up for consideration at the point of breach. Further, the duty as expressed in [23] of the concise statement, and the duty declared, is based on the evidence of the contribution of mining the coal to carbon dioxide emissions, not only by the activity of mining and transportation, but also by its combustion. This latter point becomes important as necessarily raising question of policy concerning so-called “Scope 3” emissions to which I will come.

12    The appropriate level of abstraction or focal length of perspective for the proper articulation of the scope and content of the duty is by reference to the whole of the asserted cause of action. As Brennan J said in John Pfeiffer v Canny 148 CLR at 241–242:

His duty of care is a thing written on the wind unless damage is caused by the breach of that duty; there is no actionable negligence unless duty, breach and consequential damage coincide … For the purposes of determining liability in a given case, each element can be defined only in terms of the others.

(emphasis added)

13    As Gleeson CJ said in Cole 217 CLR at 472–473 [1]:

The appellant, having suffered personal injuries, claims that the first respondent is liable to her in damages for negligence … In the circumstances of this case, it is of little assistance to consider issues of duty of care, breach, and damages, at a high level of abstraction, divorced from the concrete facts. In particular, to ask whether the respondent owed the appellant a duty of care does not advance the matter. Before she was injured, the appellant was for some hours on the respondent’s premises, and consumed food and drink supplied by the respondent. Of course the respondent owed her a duty of care. There is, however, an issue concerning the nature and extent of the duty. To address that issue, it is useful to begin by identifying the harm suffered by the appellant, for which the respondent is said to be liable, and the circumstances in which she came to suffer that harm ... As Brennan J said in Sutherland Shire Council v Heyman, ‘‘a postulated duty of care must be stated in reference to the kind of damage that a plaintiff has suffered’’. The kind of damage suffered is relevant to the existence and nature of the duty of care upon which reliance is placed. Furthermore, a description of the damage directs attention to the circumstances in which damage was suffered. ‘‘Physical injury’’, or ‘‘economic loss’’, may be an incomplete description of damage for the purpose of considering a duty of care, especially where, as in the present case, the connection between the acts or omissions of which a victim complains and the damage that she suffered is indirect.

(emphasis added, footnotes omitted)

14    The importance of the articulation of the nature and extent or scope and content of the posited duty (here, illuminated by [23] of the concise statement and by the declaration made) is that, if one posits the duty by reference to the asserted breach and the closely related evidence led to reveal the risk of harm, in order to decide the questions of the existence of the duty and of breach one is necessarily taken to consider all the evidence, material, policy and other considerations that attend the decision and that involve the question of the proper response to, including the adequacy of governmental policy in relation to, the risks of the emissions from the combustion of the coal mined as part of the worldwide risks of global warming and climate change. If one leaves the duty expressed at the high level of abstraction in [2] of the application or [22] of the concise statement one might well respond: Yes of course, but what are the circumstances? Possibly (though the Minister’s submissions contest the proposition) there could be a duty upon the Minister faced with the question of an approval of a mine of some description or of some other controlled action near a centre of urban population to exercise reasonable care for the health and safety of the nearby residents in exercising the power to approve or not approve the mine or controlled action, and if the former, on what conditions. Considerations and dangers in such a case might be so direct, so well understood, so immediately proximate, and attended by considerations in respect of which the court was entirely suited to adjudicate.

15    The question of duty is not to be placed at such a level of abstraction or generalisation as to elicit such an unhelpful response as yes, but depending on the facts, thereby leaving the real controversy and contest to breach. The duty here, however, is framed by reference to contributing to carbon dioxide emissions into the atmosphere by the combustion of the coal mined. That duty throws up for consideration at the point of assessing breach the question of the proper policy response to climate change and considerations unsuitable for resolution by the Judicial branch of government. In particular, the duty throws up at the point of assessing breach the question whether, and if so, how so-called Scope 3 emissions from the combustion of the coal that is to be exported should be or should have been taken into account in making a decision about whether to approve the extension of a coal mine, when the statutory focus and concern of the decision is the protection of identified species and communities of fauna and water resources. A duty that calls up such questions should not be imposed: It is one of core, indeed high, policy-making for the Executive and Parliament involving questions of policy (scientific, economic, social, industrial and political) which are unsuitable for the Judicial branch to resolve in private litigation by reference to the law of torts and potential personal responsibility for indeterminate damages, if harm eventuates in decades to come.

16    Later in these reasons, I will refer to other parts of the judgment of Gleeson CJ and of other members of the High Court in Graham Barclay Oysters 211 CLR 540. It is appropriate, however, at the outset to set out what the Chief Justice said at 553–554 [6], which is, in my view, the central framing consideration in this case that transcends any distinction between acts and omissions:

Citizens blame governments for many kinds of misfortune. When they do so, the kind of responsibility they attribute, expressly or by implication, may be different in quality from the kind of responsibility attributed to a citizen who is said to be under a legal liability to pay damages in compensation for injury. Subject to any insurance arrangements that may apply, people who sue governments are seeking compensation from public funds. They are claiming against a body politic or other entity whose primary responsibilities are to the public. And, in the case of an action in negligence against a government of the Commonwealth or a State or Territory, they are inviting the judicial arm of government to pass judgment upon the reasonableness of the conduct of the legislative or executive arms of government; conduct that may involve action or inaction on political grounds. Decisions as to raising revenue, and setting priorities in the allocation of public funds between competing claims on scarce resources, are essentially political. So are decisions about the extent of government regulation of private and commercial behaviour that is proper. At the centre of the law of negligence is the concept of reasonableness. When courts are invited to pass judgment on the reasonableness of governmental action or inaction, they may be confronted by issues that are inappropriate for judicial resolution, and that, in a representative democracy, are ordinarily decided through the political process. Especially is this so when criticism is addressed to legislative action or inaction. Many citizens may believe that, in various matters, there should be more extensive government regulation. Others may be of a different view, for any one of a number of reasons, perhaps including cost. Courts have long recognised the inappropriateness of judicial resolution of complaints about the reasonableness of governmental conduct where such complaints are political in nature.

17    The central error of the primary judge, in my respectful view, was to marginalise such questions and considerations as a miscellaneous control mechanism and as not relevant (see J[474]–[485]), and to construct the duty by individual analysis of salient features commencing with the risk of harm, assuming the matters thrown up by the duty were suitable for judicial determination as in any other tort case. That was the approach urged on his Honour and on this Court by the submissions of the respondents (applicants below). The submissions went so far (at least before us) that, it was said, to hold that the duty should not be imposed because it raised core policy-making and involved considerations unsuitable for judicial determination would be to abrogate judicial responsibility under Chapter III of the Constitution to quell controversies between subject and government and to introduce a “political question” doctrine foreign to Australian Constitutional government and the rule of law within it. That submission should be rejected by reference to cases of the highest binding authority. That which follows, especially [206][272] and [291][293], is an elaboration and explanation of the above. The immanent and central proposition for the existence of a duty of care, based on the central thesis of Professor Steffen, is that science dictates that for the Minister not to endanger the Children requires a re-evaluation and change to any government policy on climate change that remains fixed within the parameters of the Paris Agreement and the treatment of Scope 3 emissions.

The structure of these reasons

18    I approach the explanation of why the posited duty of care should not be imposed by dealing with the following:

The immediate factual context of the decision of the Minister: [19][42]

The statutory framework of the decision of the Minister: [43][92]

The legislative context of the EPBC Act: [93][98]

The reasons of the primary judge: [99][176]

The Minister’s grounds of appeal and submissions: [177][187]

The respondents’ submissions: [188][205]

Consideration and determination: [206][346]

Introduction: [206][213]

Grounds 2(a) and (b): human safety is not an implied mandatory consideration in the EPBC Act: [214][217]

The relationship between the Minister and the respondents and the class: [218][232]

The EPBC Act, core policy and incoherence: grounds 1 and 2(c): [233][272]

Challenge to the factual findings in ground 5: [273][290]

Coda to ground 5: [291][293]

The law of negligence: reasonable foreseeability and causation, control, vulnerability and reliance, and indeterminacy: grounds 3 and 4: [294][343]

Conclusion: [344][346]

Orders: [347]

The immediate factual context of the decision of the Minister

19    Vickery Coal Pty Ltd is a subsidiary of Whitehaven Coal Pty Ltd which holds development consent under the Environmental Planning and Assessment Act 1979 (NSW) (the EPA Act) for a coal mine in northern New South Wales near Gunnedah. No mining has commenced under the approval. In February 2016, Whitehaven applied under s 68 of the EPBC Act to the Minister to expand the existing approved project for coal mining. In July 2018, Vickery replaced Whitehaven as the proponent for the application. The extension of the mine would increase total coal extraction from the mine from 135 million to 168 million tonnes (Mt), that is by 33 Mt which, when combusted, would produce 100 Mt of carbon dioxide (CO2).

20    The application for an extension will not only increase the total coal extracted from the mine, but also increase the rate of extraction from 4.5 to 10 Mt per year, increase the area of land disturbed by mining, and there will be a new coal handling and preparation plant and rail facility at the site. The extension project itself will cause directly or indirectly emissions of greenhouse gases, especially CO2.

21    The Minister came to be concerned with the approval of the extension of the mine because on 14 April 2016 a delegate of the Minister determined that the extension of the mine was a “controlled action” under s 75(1) of the EPBC Act. The delegate decided that the relevant controlling provisions were ss 18 and 18A concerned with listed threatened species and communities and ss 24D and 24E concerned with a water resource, in relation to coal seam gas development and large coal mining development. The assessment approach was to be under the bilateral agreement with New South Wales, an approach provided for by the EPBC Act. The initial project had been approved as a “State Significant development” under the EPA Act and a delegate of the Minister had determined that it was not a “controlled action” under s 75. So the initial project being a little over four times larger than the extension did not involve the Minister under the EPBC Act.

22    As a consequence of the extension of the mine being a controlled action, the Minister was required to approve or not approve the application under ss 130(1) and 133 of the EPBC Act. The duty was binary: to approve or not, though, as will be seen, there was a power to attach some conditions. The EPBC Act requires that prior to that approval the matter be assessed. One method of assessment provided for by the EPBC Act (used in this case) was pursuant to bilateral agreement with the relevant State, here New South Wales. In May 2020, the New South Wales Department of Planning, Industry and Environment (the NSW Department) provided its assessment report to the Independent Planning Commission of New South Wales (the IPC), in the context of the Minister for Planning and Public Spaces (the State Minister) directing the IPC (under s 2.9(1)(d) of the EPA Act) to conduct a further public hearing concerning the extension of the mine with particular attention to the assessment report and any relevant public submissions. In August 2020, the IPC, after conducting a further public hearing as directed, granted development consent for the extension project. The report of the department and the reasons in the report of the IPC were provided to the Minister.

23    The IPC received a significant body of evidence during the public hearing, including a report of Professor Steffen that was in substantial accordance with the evidence led before the primary judge.

24    The two reports (of the NSW Department and the IPC) were prepared under the relevant State legislation, primarily the EPA Act. Under that statutory framework it was mandatory for the NSW Department and IPC to consider the question of greenhouse gases, including under the framework of the public interest: see s 4.15(1)(e) of the EPA Act. As discussed below, there is nothing in the EPBC Act which required the Minister to consider greenhouse gas emissions or global warming or climate change. The Minister’s necessary statutory focus was on the considerations concerned with listed threatened species and communities and the water resource in question. An approval of the extension would have greenhouse gas emissions consequences as discussed below, but they did not bear directly or indirectly upon the matters to which the Minister was directed by the EPBC Act. They did, however, bear directly upon the decision of the State Minister and upon the reports of the NSW Department and of the IPC.

25    The extension of the mine, being a development for the purpose of mining, was a “State significant development” under New South Wales legislation (s 4.36 of the EPA Act, as specified by cl 8 and cl 5 of Sch 1 of the State Environmental Planning Policy (State and Regional Development 2011 (NSW) (the SRD SEPP)). Pursuant to s 4.38 of the EPA Act, a consent authority is required either to grant (with or without modifications or conditions) or refuse consent to a development application concerning a State significant development. By operation of s 4.5(a) of the EPA Act and cl 8A(1)(b) of the SRD SEPP (due to over 50 public submissions objecting to the extension being provided to the NSW Department), the IPC became the designated consent authority. Under s 4.6(b) of the EPA Act, certain functions of the IPC are to be exercised by the Planning Secretary, including “undertaking assessments of the proposed development”, which provided the basis for the NSW Department preparing the assessment report for the IPC.

26    Unlike the requirements imposed on the Minister under the EPBC Act, the IPC as the consent authority (and as is reflected in the NSW Department’s report and the IPC’s reasons) was required specifically to consider the impact of the extension of the mine on greenhouse gas emissions and the risks to the environment and human consequences thereof. The evaluative task imposed by s 4.15 (applicable by operation of s 4.40) required the IPC to consider, among other things: the likely “environmental impacts on both the natural and built environment” (s 4.15(b)); the public interest (s 4.15(e)) (which requires consideration of the relevant objects of the EPA Act, including promoting the social and economic welfare of the community and a better environment (s 1.3(a)) and the principle of ecologically sustainable development (s 1.3(b))); and the provisions of any applicable environmental planning instrument (s 4.15(1)(a)(i)), which includes the State Environmental Planning Policy (Mining, Petroleum Production and Extractive Industries) 2007 (NSW) (the Mining SEPP). The Mining SEPP specifically required the IPC to “consider an assessment of the greenhouse gas emissions (including downstream emissions) of the development … having regard to any applicable State or national policies, programs or guidelines concerning greenhouse gas emissions” (cl 14(2)) and to consider whether or not consent should be granted with an aim to ensuring “that greenhouse gas emissions are minimised to the greatest extent practicable” (cl 14(1)(c)). The phrase “downstream emissions” encompassed the consequences of combusting or burning the coal mined.

27    The two reports considered the question of greenhouse gas emissions. In the Executive Summary of the report of the NSW Department the following appeared under the heading “Greenhouse Gas Emissions”:

The Greenhouse Gas (GHG) emissions of the Project have been assessed on a cumulative basis incorporating the Approved Project, but consideration has been given to the additional impacts over and above those associated with the Approved Project for comparative purposes.

The main sources of Scope 1, Scope 2 and Scope 3 Greenhouse Gas (GHG) emissions from the Project are from electricity consumption, fugitive emissions of carbon dioxide (CO2) and methane (CH4), diesel usage, and the transport and end use of product coal.

The Project would generate approximately 3.1 Mt carbon dioxide equivalent (CO2-e) of Scope 1 emissions, 0.8 Mt Scope 2 and 366 Mt CO2-e Scope 3 emissions.

In comparison to the Approved Project, there would be a reduction of about 1 Mt CO2-e of Scope 1 emissions, increase of about 0.15 Mt CO2-e Scope 2 emissions and an increase of about 100 Mt CO2-e of Scope 3 emissions over the life of the Project. The reduction in Scope 1 GHGE can be partially attributed to the inclusion of the Project CHPP, rail loop and rail spur, due to reduction in the consumption of diesel fuel associated with ROM coal haulage by truck to the Gunnedah CHPP.

The Project’s Scope 1 emissions would contribute to about 0.028% of Australia’s current annual GHG emissions and would remain a very small contribution when compared to Australia’s commitments under the Paris Agreement, as identified in the Commonwealth government’s nationally determined contribution (NDC).

The Department acknowledges that the Scope 3 emissions from the combustion of product coal is a significant contributor to anthropological climate change and the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

However, the Department notes that the Project’s Scope 3 emissions would not contribute to Australia’s NDC, as product coal would be exported for combustion overseas. These Scope 3 emissions become the consumer countries Scope 1 and 2 emissions and would be accounted for in their respective national inventories.

Importantly, the NSW or Commonwealth Government’s current policy frameworks do not promote restricting private development as a means for Australia to meet its commitments under the Paris Agreement or the long-term aspirational objective of the NSW Government’s Climate Change Policy Framework. Neither do they require any action to taken by the private sector in Australia to minimise or offset the GHG emissions of any parties outside of Australia, including the emissions that may be generated in transporting or using goods that are produced in Australia.

Overall, the Department considers that the GHG emissions for the Project have been adequately considered and that, with the Department’s recommended conditions, are acceptable when weighed against the relevant climate change policy framework, objects of the EP&A Act (including the principles of Ecologically Sustainable Development) and socio-economic benefits of the Project.

The Department has recommended conditions to manage the GHG emissions of the Project, including requiring Whitehaven to:

    take all reasonable steps to improve energy efficiency and reduce Scope 1 and Scope 2 GHG emissions for the Project; and

    prepare and implement an Air Quality and Greenhouse Gas Management Plan, including proposed measures to ensure best practice management is being employed to minimise the Scope 1 and 2 emissions of the Project.

(emphasis added)

28    In the detailed body of the report is the section dealing with the “public interest”. That public interest was described in [670] of the report as consideration of the objects of the State environmental legislation, including the principles of environmentally sustainable development, greenhouse gas emissions having regard to relevant climate change policy frameworks, and the (international) demand for coal and whether its sale would be to a country that is a signatory to the Paris Agreement.

29    Critical to the analysis and conclusions drawn in the report was the division of the emissions into Scope 1, 2 and 3 emissions. The protocols employed were of the World Business Council for Sustainable Development and the World Resources Institute. By that method of analysis Scope 3 emissions from the burning of the coal were part of the Scope 1 emissions of the country where the coal is combusted.

30    The coal to be mined in the extension of the mine was for export to Japan, South Korea and Taiwan.

31    At [683]–[715] the report considered “Climate Change Policy Consideration”. The report recognised that the requirements under cll 14(1) and 14(2) of the Mining SEPP to consider whether conditions should be attached to ensure the development is undertaken to minimise greenhouse gas emissions to the greatest extent possible (see [683]) and to consider an assessment of greenhouse gas emissions (including downstream emissions, which would include emissions from the combustion of the coal), such consideration to have regard to State or national policies, programs, or guidelines concerning greenhouse gas emissions (see [684]). Two key policy documents were identified (at [685]) for this assessment: the “Climate Change Policy Framework” of the New South Wales Government (CCPF) and the Commonwealth Government’s commitments to the Paris Agreement. In addition to these two policy frameworks, the report noted (at [686]) the State’s recently announced (March 2020) plan for net zero emissions by 2050 and a 35% cut in 2005 emissions by 2030.

32    At [688] the report described the State’s policy under the CCPF, as follows:

The CCPF does not set prescriptive emission reduction targets and sets policy directions for government action, for example, to improve opportunities for private sector investment in low emissions technology in the energy industry, which is needed for a transition to a net-zero emissions inventory.

33    The report stated (at [689]) that the project was not inconsistent with the CCPF, by reference to Scope 1 and Scope 2 emissions.

34    The report then considered the Paris Agreement and Australia’s obligations and commitments thereunder and thereto. The evident important consideration recognised by the report was Scope 3 emissions. The importance of Scope 3 emissions and their place in the State, Commonwealth and international policy frameworks (the last being the Paris Agreement) can be seen in [692]–[696]:

692.    The Department acknowledges that the Scope 3 emissions from the combustion of product coal is a significant contributor to anthropological [sic] climate change and the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

693.     Importantly, the Project’s Scope 3 emissions would not contribute to Australia’s NDC [nationally determined contributions], as product coal would be exported for combustion overseas. These Scope 3 emissions become the consumer countries Scope 1 and 2 emissions and would be accounted for in their respective national inventories.

694.    A regular 5-yearly review of NDCs is required under the Paris Agreement with the next review to be submitted by signatories in 2020. The Department acknowledges that ongoing review to meet emission targets by signatories may affect export markets for coal and that the UNFCCC global approach to nationally determined emission reduction targets is the appropriate mechanism for managing Australia’s Scope 3 emissions, rather than regulating Scope 3 emissions on a project by project basis in Australia.

695.    The Department also notes that the Department’s ‘Guidelines for Economic Assessment of Mining and Coal Seam Gas Proposals’ and the associated 2018 technical notes do not require the social cost of Scope 3 emissions to be incorporated into the economic evaluation when determining the net benefits to NSW or Australia of the development. This approach, where both the costs and benefits of consumption and use of the coal is considered by the country/ development where the coal is being used, is consistent with the global accounting framework for GHG emissions under the UNFCCC.

696.     Importantly, the NSW or Commonwealth Government’s current policy frameworks do not promote restricting private development as a means for Australia to meet its commitments under the Paris Agreement or the long-term aspirational objective of the CCPF guidelines. Neither do they require any action to [sic] taken by the private sector in Australia to minimise or offset the GHG emissions of any parties outside of Australia, including the emissions that may be generated in transporting or using goods that are produced in Australia.

35    The place of Scope 3 emissions in Commonwealth policy was made clear in the report at [697] in which a letter from the Commonwealth Minister to her State counterpart expressing Commonwealth Government policy was quoted as saying:

… “any requirement to consider scope three emissions within a sub-national or state jurisdiction is inconsistent with long accepted international carbon accounting principles and Australia’s international commitments”.

36    The report went on to explain (at [701]–[706]) why Scope 3 emissions should not be relevant:

701.     There is no NSW or Commonwealth policy that supports placing conditions on an applicant to minimise the Scope 3 emissions of its development. Any such policy is likely to result in significant implications for the NSW and Australian economy and it is not clear it would have any effect on reducing GHG emissions generated by parties in other jurisdictions outside Australia. Further, conditions must be for a proper planning purpose, must fairly and reasonably relate to the subject development, and must not be manifestly unreasonable.

702.    On this basis the Department has recommended conditions requiring Whitehaven to take all reasonable steps to improve energy efficiency and reduce Scope 1 and Scope 2 GHG emissions for the Project and to prepare and implement an Air Quality and Greenhouse Gas Management Plan, including a requirement to apply best practice to minimise the Scope 1 and 2 emissions of the Project.

703.    The Department also acknowledges that GHG emissions have attracted additional attention following a February 2019 judgement in the Land and Environment Court (Rocky Hill appeal – [Gloucester Resources Limited vs Minister for Planning] (2019 NSWLEC 7) and the Commission’s subsequent decisions relating to GHG emissions in coal mining projects, including the refusal of Bylong Coal Project and inclusion of conditions relating to Scope 3 GHG emissions for the United Wambo Open Cut Coal Mine.

704.    In its Statement of Reasons for Decision for the Rixs Creek Continuation of Mining Project (SSD 6300), the Commission noted that the Applicant for the Rixs Creek Project does not have direct control over Scope 3 emissions and accepted that Scope 3 emissions were the responsibility of the end customer for coal export. The Commission also noted that coal consumption in countries which are signatories to the Paris Agreement, or have enforced GHG reduction targets (such as Taiwan) would lead buyers to seek coal products which meet their product requirements and would also minimise GHG emissions to achieve domestic emission reduction targets.

705.    The NSW Government has since introduced a Bill into Parliament (Environmental Planning and Assessment Amendment (Territorial Limits) Bill 2019). The Bill was introduced in response to recent planning decisions and seeks to clarify that conditions under EP&A Act can only be imposed if they relate to impacts occurring within Australia or its external territories.

706.    This aligns with the intent that development consent conditions set and enforced in the NSW planning system are not an appropriate mechanism to control the impacts resulting from the activities of third parties in other countries.

37    The report then considered (at [707]–[710]) the international demand for coal, as follows:

707.    The Project would produce metallurgical coal (around 70% of the product coal) including semi-soft coking coal, pulverised coal injection (PCI) coal and thermal coal (around 30% of the product coal) to supply Whitehaven’s main export market customers in Japan, the Republic of Korea (South Korea) and the Republic of China (Taiwan).

708.    Japan and South Korea are signatories to the Paris Agreement and have developed GHG emission reduction targets, which would be managed under the NDCs of these countries. Taiwan is not a signatory to the Paris Agreement but has developed its own GHG emission reduction targets (enforced under its Greenhouse Gas Reduction and Management Act) that are comparable to those of countries who are signatories.

709.    Whitehaven recognises that global coal demands are shifting and has provided an economic sensitivity analysis (see Section 6.8 – Economic Evaluation) to account for changing trends in forecast coal pricing and demand. The sensitivity analysis shows that significant net benefits would accrue to NSW over a range of assumptions for coal prices, discount rates, exchange rates and employment related benefits.

710.    The Department notes that the majority of the coal is of metallurgical quality and that the thermal coal quality is a high calorific/ low ash/ low sulphur coal which is in stronger demand globally compared to lower quality (high ash/ high sulphur) coal. Whitehaven provided the Department with further information (see Appendix G6-10) on the Project’s coal quality relative to anticipated demand based on the three climate changes scenarios contemplated by the International Energy Agency (IEA) in its World Energy Outlook 2019. Under the Sustainable Development Scenario there would continue to be demand for high quality (low ash/ low sulphur/ high calorific energy) thermal and metallurgical coal, particularly in the Asia Pacific region, as provided by the Vickery coal resource.

38    Explicit in the above is an approach based on State and Commonwealth public policy stated to be conformable with international convention (the Paris Agreement) and in part reduced to State legislation that Scope 3 emissions are a matter for the countries buying and combusting the coal which have emission targets conformable with international treaty.

39    Appendix J to the report concerned matters relevant to the Minister’s decision under the EPBC Act. It dealt with impacts on listed species and communities in section J.1. This included the likely impact of certain native vegetation clearing on listed species and communities of fauna, including birds, koalas, bats and fish. Section J.2 dealt with offsetting impacts. Sections J.3 to J.6 dealt with other matters required to be considered under the EPBC Act. At s J.7 conclusions were set out on the controlling provisions as follows:

Threatened species and communities (Sections 18 and 18A of the Act)

For the reasons set out in Section 6.2, Appendix I and this Appendix, the Department recommends that the impacts of the action would be acceptable, subject to avoidance, mitigation measures described in Whitehaven’s EIS, Submissions Report and additional advice provided to the Department and the recommended conditions of consent in Appendix L.

A water resource, in relation to coal seam gas development and large coal mining development (Sections 24D and 24E of the Act)

For the reasons set out in Section 6.2 and this Appendix, the Department recommends that the impacts of the action on a water resource, in relation large coal mining development would be acceptable, subject to the avoidance, mitigation measures described in Whitehaven’s EIS, Submissions Report and the requirements of the recommended conditions of consent in Appendix L.

40    In appendix K, the NSW Department considered various aspects of the EPA Act requirements. Amongst these were “intergenerational equity”. The report stated in appendix K about this matter the following:

Intergenerational equity has been addressed through maximising efficiency and coal resource recovery and developing environmental management measures which are aimed at ensuring the health, diversity and productivity of the environment are maintained or enhanced for the benefit of future generations.

The Department acknowledges that coal and other fossil fuel combustion is a contributor to climate change, which has the potential to impact future generations. However, the Department also recognises that there remains a clear need to develop coal deposits to meet society’s basic energy requirements for the foreseeable future. The proposal includes measures to mitigate potential GHGE’s from the operation of the Project, which would be recommended as a requirement of the Project’s operating conditions and detailed in an Air Quality and Greenhouse Gas Management Plan.

The Department’s assessment of direct energy use and associated GHGE’s (ie Scope 1 and Scope 2 emissions) has found that these emissions would be low and comprise a very small contribution towards climate change at both the national and global scale (see Section 6.10).

The Department considers that the socio-economic benefits and downstream energy generated by the Project would benefit future generations, particularly through the provision of national and international energy needs in the short to medium term.

41    The report of the IPC reviewed the Department’s report. It similarly categorised the greenhouse gas emissions into Scope 1, 2 and 3 emissions. A public hearing was conducted. The submissions of members of the public expressed concern as to the total amount of greenhouse gas emissions from the extension and the project, including Scope 3 emissions. The IPC made its findings, consistent with Australia’s non-responsibility for Scope 3 emissions, saying at [215]–[216] and [220]–[223], as follows:

215.    The Commission acknowledges that the aim of the NSW Climate Change Policy Framework (CCPF) is to “maximise the economic, social and environmental wellbeing of NSW in the context of a changing climate and current and emerging international and national policy settings and actions to address climate change” with an aim to achieve net-zero emissions by 2050 and to ensure NSW is more resilient to a changing climate. The Commission notes that the CCPF does not set prescriptive emission reduction targets and sets policy directions for government action as stated by the Department in paragraph 207 above. The Commission also notes that the NSW Government released the Net Zero Plan Stage 1: 2020–2030 (Net Zero Plan) in March 2020 as referenced by the Department in paragraph 208 above. The Commission notes that the Net Zero Plan builds on the CCPF and sets out a number of initiatives to deliver a 35% cut in emissions by 2030, compared to 2005 levels. The Commission agrees with the Department’s assessment, in paragraph 207 above that the Project is not inconsistent with the CCPF and that the Applicant has committed to minimising its Scope 1 emissions over which it has direct control.

216.    The Commission notes that, under the Paris Agreement, the Australian Government committed to a nationally determined contribution (NDC) to reduce national GHG emissions by between 26 and 28 percent from 2005 levels by 2030. The Commission also notes that Australia does not require monitoring or reporting of Scope 3 emissions under the NGERS [National Greenhouse and Energy Reporting Scheme] and they are not counted in Australia’s national inventory of GHG emissions under the Paris Agreement. The Commission agrees with the Department’s statement in paragraph 209 above that the Project’s Scope 3 emissions would not contribute to Australia’s NDC, as product coal would be exported overseas. The Commission notes that these Scope 3 emissions become the consumer countries’ Scope 1 and 2 emissions and would be accounted for under the Paris Agreement in their respective national inventories.

220.     The Commission has acknowledged that the concerns raised by the public in paragraphs 188 to 191 above, including submissions made by the BFCG [Boggabri Farming and Community Group] in relation to the burning of fossil fuels as an energy resource and meeting the Paris Agreement climate targets. The Commission notes that the carbon budget approach suggested in some submissions is not endorsed by the Paris Agreement, the Australian Government or the NSW Government. Furthermore, neither the Australian nor NSW Government have indicated that the development of new coal mines or the expansion of existing mines be prohibited or restricted in any way for the purpose of achieving Australia’s NDC.

221.    The Commission agrees with the Department’s statement in paragraph 209 above and acknowledges that Scope 3 emissions from the combustion of product coal are a significant contributor to anthropological [sic] climate change and that the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

222.    The Commission notes that between 60-70% of the coal proposed to be extracted is likely to be metallurgical coal, with the remainder being thermal coal as stated above by the Applicant in paragraph 200 and by the Department in paragraph 205 of this report. The Commission notes that at this point in time, metallurgical coals are essential inputs for the current production of approximately 70% of all steel globally as stated by the Applicant in paragraph 200 above. The Commission is of the view that in the absence of a viable alternative to the use of metallurgical coal in steel making and on balance, the impacts associated with the emissions from the combustion of the project’s metallurgical coal are acceptable. The Commission also notes that the coal proposed for extraction is anticipated to be of a relatively high quality, as stated above by the Applicant in paragraph 194 and Department in paragraph 204. The Commission notes the Applicant’s statement in paragraph 194 above that the use of higher quality coal may result in lower pollutants.

223.    For the reasons set out above, the Commission is of the view that the GHG emissions for the Project have been adequately considered. The Commission finds that on balance, and when weighed against the relevant climate change policy framework, objects of the EP&A Act, ESD principles (section 4.10) and socio-economic benefits (section 4.9.6), the impacts associated with the GHG emissions of the Project are acceptable and consistent with the public interest. The Commission therefore imposes the Conditions B35, B36 and B37 as recommended by the Department.

(emphasis added)

42    The Minister thus came to her responsibility to decide whether to approve the application.

The statutory framework in the EPBC Act of the decision of the Minister

43    The statutory framework of the Minister’s decision needs to be explained. I will begin with the text of the statute. It is, however, necessary to say something of the context of the EPBC Act, in particular its passing as an expression of the agreement amongst governments of the federation to co-operate in their shared responsibilities in respect of the environment.

44    At the outset of this explanation it is appropriate to emphasise a point (correctly) submitted by the Minister to be central to the resolution of the controversy. The EPBC Act is not concerned generally with the protection of the environment. Nor is there any part of the EPBC Act that is expressly concerned with greenhouse gases, global warming or climate change. Using such Constitutional foundations as are available, most importantly the external affairs power in the implementation of conventions and international agreements (s 51(xxix) of the Constitution), the Parliament has directed the EPBC Act to nine “matters of national environmental significance”. In respect of some, but not all, of these matters the environment (as widely defined) is a relevant consideration. The two matters of national environmental significance which were the subject of the Minister’s concern and decision were listed endangered species and communities and the water resource. As will be seen, the environment generally (as widely defined) was not an express consideration in respect of those matters. Yet, it was only by the exercise of the power to approve the extension, derived from the duty to approve or refuse the application, by reference to the particular matters of concern (species, communities and water resource) that the question, called forth by the posited duty, of the affectation of the environment generally by emission of greenhouse gases arose. The decision for the Commonwealth Minister was not concerned with greenhouse gas emissions, although the State decision was so concerned. The duty of care was said to arise because the power to approve amounted to the effective grant of a licence to carry on activity which involved the foreseeable risk of personal injury or death to the respondents.

45    The objects of the EPBC Act are set out in s 3(1) as follows:

(1)    The objects of this Act are:

(a)    to provide for the protection of the environment, especially those aspects of the environment that are matters of national environmental significance; and

(b)    to promote ecologically sustainable development through the conservation and ecologically sustainable use of natural resources; and

(c)    to promote the conservation of biodiversity; and

(ca)    to provide for the protection and conservation of heritage; and

(d)    to promote a co‑operative approach to the protection and management of the environment involving governments, the community, land‑holders and indigenous peoples; and

(e)    to assist in the co‑operative implementation of Australia’s international environmental responsibilities; and

(f)    to recognise the role of indigenous people in the conservation and ecologically sustainable use of Australia’s biodiversity; and

(g)    to promote the use of indigenous peoples’ knowledge of biodiversity with the involvement of, and in co‑operation with, the owners of the knowledge.

46    The word “environment” (see ss 3(1)(a) and Ch 2) is defined in broad terms in s 528 of the EPBC Act as including:

(a)    ecosystems and their constituent parts, including people and communities; and

(b)    natural and physical resources; and

(c)    the qualities and characteristics of locations, places and areas; and

(d)    heritage values of places; and

(e)    the social, economic and cultural aspects of a thing mentioned in paragraph (a), (b), (c) or (d).

47    The phrase “ecologically sustainable use” (in para 3(1)(b)) is defined in s 528 as follows:

ecologically sustainable use of natural resources means use of the natural resources within their capacity to sustain natural processes while maintaining the life‑support systems of nature and ensuring that the benefit of the use to the present generation does not diminish the potential to meet the needs and aspirations of future generations.

48    Subsection 3(2) sets out how the EPBC Act seeks to achieve these objects:

(2)    In order to achieve its objects, the Act:

(a)    recognises an appropriate role for the Commonwealth in relation to the environment by focussing Commonwealth involvement on matters of national environmental significance and on Commonwealth actions and Commonwealth areas; and

(b)    strengthens intergovernmental co‑operation, and minimises duplication, through bilateral agreements; and

(c)    provides for the intergovernmental accreditation of environmental assessment and approval processes; and

(d)    adopts an efficient and timely Commonwealth environmental assessment and approval process that will ensure activities that are likely to have significant impacts on the environment are properly assessed; and

(e)    enhances Australia’s capacity to ensure the conservation of its biodiversity by including provisions to:

(i)    protect native species (and in particular prevent the extinction, and promote the recovery, of threatened species) and ensure the conservation of migratory species; and

(ii)    establish an Australian Whale Sanctuary to ensure the conservation of whales and other cetaceans; and

(iii)    protect ecosystems by means that include the establishment and management of reserves, the recognition and protection of ecological communities and the promotion of off‑reserve conservation measures; and

(iv)    identify processes that threaten all levels of biodiversity and implement plans to address these processes; and

(f)    includes provisions to enhance the protection, conservation and presentation of world heritage properties and the conservation and wise use of Ramsar wetlands of international importance; and

(fa)    includes provisions to identify places for inclusion in the National Heritage List and Commonwealth Heritage List and to enhance the protection, conservation and presentation of those places; and

(g)    promotes a partnership approach to environmental protection and biodiversity conservation through:

(i)    bilateral agreements with States and Territories; and

(ii)    conservation agreements with land‑holders; and

(iii)    recognising and promoting indigenous peoples’ role in, and knowledge of, the conservation and ecologically sustainable use of biodiversity; and

(iv)    the involvement of the community in management planning.

        (emphasis added)

49    Section 3A describes the principles of “ecologically sustainable development” as follows:

(a)    decision‑making processes should effectively integrate both long‑term and short‑term economic, environmental, social and equitable considerations;

(b)    if there are threats of serious or irreversible environmental damage, lack of full scientific certainty should not be used as a reason for postponing measures to prevent environmental degradation;

(c)    the principle of inter‑generational equity—that the present generation should ensure that the health, diversity and productivity of the environment is maintained or enhanced for the benefit of future generations;

(d)    the conservation of biological diversity and ecological integrity should be a fundamental consideration in decision‑making;

(e)    improved valuation, pricing and incentive mechanisms should be promoted.

50    Chapter 2 of the EPBC Act concerns “protecting the environment”. Section 11 contains a simplified outline of the chapter as follows:

This Chapter provides a basis for the Minister to decide whether an action that has, will have or is likely to have a significant impact on certain aspects of the environment should proceed.

It does so by prohibiting a person from taking an action without the Minister having given approval or decided that approval is not needed. (Part 9 deals with the giving of approval.)

Approval is not needed to take an action if any of the following declare that the action does not need approval:

(a)    a bilateral agreement between the Commonwealth and the State or Territory in which the action is taken;

(b)    a declaration by the Minister.

Also, an action does not need approval if it is taken in accordance with Regional Forest Agreements or it is for a purpose for which, under a zoning plan for a zone made under the Great Barrier Reef Marine Park Act 1975, the zone may be used or entered without permission.

51    Part 3 within Ch 2 concerns “requirements for environmental approvals”. Division 1 of Pt 3 concerns “requirements relating to matters of national and environmental significance”. Subdivisions A to G set out ten categories of such matters: “World Heritage” (Subdiv A), “National Heritage” (Subdiv AA), “wetlands of international importance” (Subdiv B), relevantly here “listed threatened species and communities” (Subdiv C), “listed migratory species” (Subdiv D), “protection of the environment from nuclear actions (Subdiv E), “marine environment” (Subdiv F), “Great Barrier Reef Marine Park” (Subdiv FA), relevantly here “protection of water resources from coal seam gas development and large coal mining development” (Subdiv FB), and “additional matters of national environmental significance” (Subdiv G).

52    The phrases “coal seam gas development” and “large coal mining development” are defined in s 528 as follows:

coal seam gas development means any activity involving coal seam gas extraction that has, or is likely to have, a significant impact on water resources (including any impacts of associated salt production and/or salinity):

(a)    in its own right; or

(b)    when considered with other developments, whether past, present or reasonably foreseeable developments.

large coal mining development means any coal mining activity that has, or is likely to have, a significant impact on water resources (including any impacts of associated salt production and/or salinity):

(a)    in its own right; or

(b)    when considered with other developments, whether past, present or reasonably foreseeable developments.

53    Sections 18 and 18A (in Subdiv C) and 24D and 24E (in Subdiv FB) were relevant to the delegate’s decision under s 75. Their structure mirrors the pattern of other sections in Pt 3: a prohibition on taking an action and an identification of the circumstances when that prohibition does not apply. Sections 18 and 18A set out the prohibitions and offences in relation to species and communities and are as follows:

18    Actions with significant impact on listed threatened species or endangered community prohibited without approval

Species that are extinct in the wild

(1)    A person must not take an action that:

(a)    has or will have a significant impact on a listed threatened species included in the extinct in the wild category; or

(b)    is likely to have a significant impact on a listed threatened species included in the extinct in the wild category.

Civil penalty:

  (a)    for an individual—5,000 penalty units;

  (b)    for a body corporate—50,000 penalty units.

Critically endangered species

(2)    A person must not take an action that:

    (a)    has or will have a significant impact on a listed threatened species included in the critically endangered category; or

    (b)    is likely to have a significant impact on a listed threatened species included in the critically endangered category.

        Civil penalty:

        (a)    for an individual—5,000 penalty units;

        (b)    for a body corporate—50,000 penalty units.

Endangered species

(3)    A person must not take an action that:

  (a)    has or will have a significant impact on a listed threatened species included in the endangered category; or

  (b)    is likely to have a significant impact on a listed threatened species included in the endangered category.

Civil penalty:

(a)    for an individual—5,000 penalty units;

(b)    for a body corporate—50,000 penalty units.

Vulnerable species

(4)    A person must not take an action that:

(a)    has or will have a significant impact on a listed threatened species included in the vulnerable category; or

(b)    is likely to have a significant impact on a listed threatened species included in the vulnerable category.

   Civil penalty:

   (a)    for an individual—5,000 penalty units;

   (b)    for a body corporate—50,000 penalty units.

Critically endangered communities

(5)    A person must not take an action that:

(a)     has or will have a significant impact on a listed threatened ecological community included in the critically endangered category; or

(b)    is likely to have a significant impact on a listed threatened ecological community included in the critically endangered category.

   Civil penalty:

   (a)    for an individual—5,000 penalty units;

   (b)    for a body corporate—50,000 penalty units.

Endangered communities

(6)    A person must not take an action that:

(a)    has or will have a significant impact on a listed threatened ecological community included in the endangered category; or

(b)    is likely to have a significant impact on a listed threatened ecological community included in the endangered category.

Civil penalty:

(a)    for an individual—5,000 penalty units;

(b)    for a body corporate—50,000 penalty units.

18A    Offences relating to threatened species etc.

(1)    A person commits an offence if:

(a)    the person takes an action; and

(b)    the action results or will result in a significant impact on:

(i)    a species; or

(ii)    an ecological community; and

(c)    the species is a listed threatened species, or the community is a listed threatened ecological community.

Note:    Chapter 2 of the Criminal Code sets out the general principles of criminal responsibility.

(1A)    Strict liability applies to paragraph (1)(c).

Note:    For strict liability, see section 6.1 of the Criminal Code.

(2)    A person commits an offence if:

(a)    the person takes an action; and

(b)    the action is likely to have a significant impact on:

(i)    a species; or

(ii)    an ecological community; and

(c)    the species is a listed threatened species, or the community is a listed threatened ecological community.

Note:    Chapter 2 of the Criminal Code sets out the general principles of criminal responsibility.

(2A)    Strict liability applies to paragraph (2)(c).

(3)    An offence against subsection (1) or (2) is punishable on conviction by imprisonment for a term not more than 7 years, a fine not more than 420 penalty units, or both.

(4)    Subsections (1) and (2) do not apply to an action if:

(a)    the listed threatened species subject to the significant impact (or likely to be subject to the significant impact) is:

(i)    a species included in the extinct category of the list under section 178; or

(ii)    a conservation dependent species; or

(b)    the listed threatened ecological community subject to the significant impact (or likely to be subject to the significant impact) is an ecological community included in the vulnerable category of the list under section 181.

54    Section 19 provides for certain actions not being prohibited. These include where an approval under Pt 9 for the taking of the action is in operation: subss 19(1) and (2), or for other reasons set out in subs 19(3) including if the Minister decides under Div 2 of Pt 7 that the provision is not a controlling provision: para 19(3)(b).

55    Sections 24D and 24E set out the prohibitions and offences in relation to water resources, and are as follows:

24D    Requirement for approval of developments with a significant impact on water resources

(1)    A constitutional corporation, the Commonwealth or a Commonwealth agency must not take an action if:

(a)    the action involves:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action:

(i)    has or will have a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Civil penalty:

(a)    for an individual—5,000 penalty units;

(b)    for a body corporate—50,000 penalty units.

(2)    A person must not take an action if:

(a)    the action involves:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action is taken for the purposes of trade or commerce:

(i)    between Australia and another country; or

(ii)    between 2 States; or

(iii)    between a State and Territory; or

(iv)    between 2 Territories; and

(c)    the action:

(i)    has or will have a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Civil penalty:

(a)    for an individual—5,000 penalty units;

(b)    for a body corporate—50,000 penalty units.

(3)    A person must not take an action if:

(a)    the action involves:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action is taken in:

(i)    a Commonwealth area; or

(ii)    a Territory; and

(c)    the action:

(i)    has or will have a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Civil penalty:

(a)    for an individual—5,000 penalty units;

(b)    for a body corporate—50,000 penalty units.

(4)    Subsections (1) to (3) do not apply to an action if:

(a)    an approval of the taking of the action by the constitutional corporation, Commonwealth, Commonwealth agency or person is in operation under Part 9 for the purposes of this section; or

(b)    Part 4 lets the constitutional corporation, Commonwealth, Commonwealth agency or person take the action without an approval under Part 9 for the purposes of this section; or

(c)    there is in force a decision of the Minister under Division 2 of Part 7 that this section is not a controlling provision for the action and, if the decision was made because the Minister believed the action would be taken in a manner specified in the notice of the decision under section 77, the action is taken in that manner; or

(d)    the action is an action described in subsection 160(2) (which describes actions whose authorisation is subject to a special environmental assessment process).

(5)    A person who wishes to rely on subsection (4) in proceedings for a contravention of a civil penalty provision bears an evidential burden in relation to the matters in that subsection.

24E    Offences relating to water resources

(1)    A constitutional corporation, or a Commonwealth agency that does not enjoy the immunities of the Commonwealth, commits an offence if:

(a)    the corporation or agency takes an action involving:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action:

(i)    results or will result in a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Penalty: Imprisonment for 7 years or 420 penalty units, or both.

(2)    A person commits an offence if:

(a)    the person takes an action involving:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action is taken for the purposes of trade or commerce:

(i)    between Australia and another country; or

(ii)    between 2 States; or

(iii)    between a State and Territory; or

(iv)    between 2 Territories; and

(c)    the action:

(i)    has or will have a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Penalty: Imprisonment for 7 years or 420 penalty units, or both.

(3)    A person commits an offence if:

(a)    the person takes an action involving:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the action is taken in:

(i)    a Commonwealth area; or

(ii)    a Territory; and

(c)    the action:

(i)    has or will have a significant impact on a water resource; or

(ii)    is likely to have a significant impact on a water resource.

Penalty: Imprisonment for 7 years or 420 penalty units, or both.

(4)    Subsections (1) to (3) do not apply to an action if:

(a)    an approval of the taking of the action by the constitutional corporation, Commonwealth agency or person is in operation under Part 9 for the purposes of this section; or

(b)    Part 4 lets the constitutional corporation, Commonwealth agency or person take the action without an approval under Part 9 for the purposes of this section; or

(c)    there is in force a decision of the Minister under Division 2 of Part 7 that this section is not a controlling provision for the action and, if the decision was made because the Minister believed the action would be taken in a manner specified in the notice of the decision under section 77, the action is taken in that manner; or

(d)    the action is an action described in subsection 160(2) (which describes actions whose authorisation is subject to a special environmental assessment process).

56    The matters of national environmental significance in Div 1 of Pt 3 are varied, see Subdivs A-G of Div 1 of Pt 3 referred to at [51] above. Division 2 of Pt 3 is also relevant: The protection of the environment from proposals involving the Commonwealth, including: “protection of the environment from actions involving Commonwealth land” (Subdiv A); concerning Commonwealth Heritage places outside Australia (Subdiv AA); and “protection of the environment from Commonwealth actions” (Subdiv B). The matters relevant to this decision are in Subdivs C and FB of Div 1 of Pt 3 (see above). But an appreciation of the variety of the others is relevant to the question of the imposition of the duty in decision making under the ss 130 and 133 because it assists in comprehending the structure and purpose of the EPBC Act and the particular provisions in respect of which decisions of the Minister may be required and the responsibility of the Minister to Parliament in respect thereof.

57    A feature of some importance in the understanding of the nature of the duties, powers and functions of the Minister in connection with the various features of the requirements relating to matters of national environmental significance is the place of the “environment” as defined in such considerations. The definition of the word is set out at [46] above. The word appears as part of the expression of the matters of national environmental significance in some, but not all, of the matters to which the EPBC Act makes reference: in Div 1 of Pt 3: in Subdiv E: the protection of the environment from nuclear actions where the nuclear action has, will have, or is likely to have a significant impact on the environment: (ss 21–22A); in Subdiv F: action which has, will have, or is likely to have a significant impact on the environment being a Commonwealth marine area: (ss 23–24A); Subdiv FA: action that has, will have, or is likely to have a significant impact on the environment in the Great Barrier Reef Marine Park: (ss 24B and 24C); and in Div 2 of Pt 3: in Subdiv A: concerning protection from action that has, will have, or is likely to have a significant impact on the environment on Commonwealth land (ss 26–27A); Subdiv AA: concerning protection from action that has, will have, or is likely to have a significant impact on the environment in a Commonwealth Heritage Place outside Australia: (ss 27B and 27C); and Subdiv B: concerning protection of the environment from action by the Commonwealth or a Commonwealth agency that has, will have, or is likely to have a significant impact on the environment: (s 28).

58    The other provisions of Pt 3 are not directed to protection of the “environment” as that word is defined, but rather the more specific features within the concept of the environment set out that are “matters of national environmental significance”. In the decision before the Minister under consideration here, those matters are contained in ss 18, 18A, 24D and 24E of the EPBC Act.

59    Part 4 within Ch 2 concerns “cases in which environmental approval are not needed”. In s 34, the matters protected by Pt 3 are identified. Items 3 to 8A concern ss 18 and 18A and items 13H and 13J concern ss 24D and 24E.

60    Chapter 3 concerns bilateral agreements between Commonwealth and States and Territories.

61    The object of bilateral agreements was set out in s 44 as follows:

The object of this Part is to provide for agreements between the Commonwealth and a State or self‑governing Territory that:

(a)    protect the environment; and

(b)    promote the conservation and ecologically sustainable use of natural resources; and

(c)    ensure an efficient, timely and effective process for environmental assessment and approval of actions; and

(d)    minimise duplication in the environmental assessment and approval process through Commonwealth accreditation of the processes of the State or Territory (and vice versa).

(emphasis added)

62    Section 46 provides, inter alia, for accreditation by the Minister of bilateral authorisation processes in accordance with criteria in the regulations.

63    Under subs 47(4):

If a bilateral agreement has (or could have) the effect that an action need not be assessed under Part 8 but the action must still be approved under Part 9, the agreement must provide for the Minister to receive a report including, or accompanied by, enough information about the relevant impacts of the action to let the Minister make an informed decision whether or not to approve under Part 9 (for the purposes of each controlling provision) the taking of the action.

The process in this case reflected the operation of subs 47(4).

64    Chapter 4 concerns “environmental assessments and approvals”. Section 66 sets out a simplified outline of Ch 4 as follows:

This Chapter deals with assessment and approval of actions that Part 3 prohibits without approval (controlled actions). (It does not deal with actions that a bilateral agreement declares not to need approval.)

A person proposing to take an action, or a government body aware of the proposal, may refer the proposal to the Minister so he or she can decide:

(a)    whether his or her approval is needed to take the action; and

(b)    how to assess the impacts of the action to be able to make an informed decision whether or not to approve the action.

An assessment may be done using:

(a)    a process laid down under a bilateral agreement; or

(b)    a process specified in a declaration by the Minister; or

(c)    a process accredited by the Minister; or

(ca)     information included in the referral; or

(d)    preliminary documentation provided by the proponent; or

(e)    a public environment report; or

(f)    an environmental impact statement; or

(g)    a public inquiry.

Once the report of the assessment is given to the Minister, he or she must decide whether or not to approve the action, and what conditions to attach to any approval.

(emphasis added)

65    In Pt 7 within Ch 4, s 67 defines a “controlled action” as follows:

An action that a person proposes to take is a controlled action if the taking of the action by the person without approval under Part 9 for the purposes of a provision of Part 3 would be (or would, but for section 25AA or 28AB, be) prohibited by the provision. The provision is a controlling provision for the action.

Sections 25AA and 28AB referred to in s 67 concern limitations of liability for actions of third parties.

66    Section 67A prohibits taking a controlled action without approval, as follows:

A person must not take a controlled action unless an approval of the taking of the action by the person is in operation under Part 9 for the purposes of the relevant provision of Part 3.

67    Section 68 allows persons who propose to take action who think it may be a controlled action to refer the matter to the Minister. By s 69, a State or Territory may refer a proposal to the Minister. By s 70 the Minister may request a person or a State or Territory to refer a proposal to him or her. By s 71 a Commonwealth agency may refer a proposal to the Minister.

68    Section 74 provides for the invitation of information on a referred proposal from other Commonwealth Ministers, from a State or Territory or the invitation of public comment.

69    Division 2 of Pt 7 deals with the Ministerial decision whether action needs approval and thus whether action becomes a controlled action. Section 75 deals with the question whether the proposed action needs approval. The Minister must decide whether the subject of the proposal is a controlled action. The delegate of the Minister here decided that approval was necessary. In making the decision under s 75 the Minister must consider all adverse impacts (if any) the action has or will have or is likely to have on the matter protected by each provision of Pt 3 and must not consider any beneficial impacts the action has or will have or is likely to have on the matter protected.

70    Part 8 within Ch 4 concerns “assessing impacts of controlled actions”. The Part does not apply to actions that a bilateral agreement says are to be assessed another way: subs 83(1).

71    Under subs 82(1) of the EPBC Act, if the Minister has decided under Div 2 of Pt 7 (which includes s 75) that an action is a controlled action, the relevant impacts of the action are the impacts that the action has, or will have, or is likely to have on the matter protected by the provisions of Pt 3 that the Minister has decided (here by her delegate under s 75) is a controlling provision for the action: Here, such controlling provisions are ss 18, 18A, 24D and 24E.

72    The word “impact” is defined in s 527E as follows:

(1)    For the purposes of this Act, an event or circumstance is an impact of an action taken by a person if:

(a)    the event or circumstance is a direct consequence of the action; or

(b)    for an event or circumstance that is an indirect consequence of the action—subject to subsection (2), the action is a substantial cause of that event or circumstance.

(2)    For the purposes of paragraph (1)(b), if:

(a)    a person (the primary person) takes an action (the primary action); and

(b)    as a consequence of the primary action, another person (the secondary person) takes another action (the secondary action); and

(c)    the secondary action is not taken at the direction or request of the primary person; and

(d)    an event or circumstance is a consequence of the secondary action;

then that event or circumstance is an impact of the primary action only if:

(e)    the primary action facilitates, to a major extent, the secondary action; and

(f)    the secondary action is:

(i)    within the contemplation of the primary person; or

(ii)    a reasonably foreseeable consequence of the primary action; and

(g)    the event or circumstance is:

(i)    within the contemplation of the primary person; or

(ii)    a reasonably foreseeable consequence of the secondary action.

73    A consequence of the action (thus the effect on future global warming of the mining of the coal produced by the extension of the mine) must be either a direct consequence of the action, or if an indirect consequence, the mining of the coal must be the substantial cause of future global warming.

74    Part 9 within Ch 4 concerns “approval of actions”. Subsection 130(1) provides for the Minister’s approval for a controlled action, as follows:

(1)    The Minister must decide whether or not to approve, for the purposes of each controlling provision for a controlled action, the taking of the action.

75    The balance of s 130 delimits the timeframes for decisions to be made.

76    Sections 131, 131AA and 131A provide for inviting comments from other Ministers, from the person proposing the action and the “designated proponent”, and from the public. Section 131AB requires the Minister to obtain advice from the Independent Expert Scientific Committee on Coal Seam Gas and Large Coal Mining Development. Subsection 131AB(1) sets out when this is necessary:

(1)    This section applies if:

(a)    the taking of an action, for the purposes of a controlling provision, involves:

(i)    coal seam gas development; or

(ii)    large coal mining development; and

(b)    the Minister believes that the taking of the action:

(i)    is likely to have a significant impact on water resources, including any impacts of associated salt production and/or salinity; and

(ii)    may have an adverse impact on a matter protected by a provision of Part 3.

77    Section 133 deals with “grant of approval” and provides as follows:

Approval

(1)    After receiving the assessment documentation relating to a controlled action, or the report of a commission that has conducted an inquiry relating to a controlled action, the Minister may approve for the purposes of a controlling provision the taking of the action by a person.

(1A)    If the referral of the proposal to take the action included alternative proposals relating to any of the matters referred to in subsection 72(3), the Minister may approve, for the purposes of subsection (1), one or more of the alternative proposals in relation to the taking of the action.

Content of approval

(2)    An approval must:

(a)    be in writing; and

(b)    specify the action (including any alternative proposals approved under subsection (1A)) that may be taken; and

(c)    name the person to whom the approval is granted; and

(d)    specify each provision of Part 3 for which the approval has effect; and

(e)    specify the period for which the approval has effect; and

(f)    set out the conditions attached to the approval.

Note:    The period for which the approval has effect may be extended. See Division 5.

Persons who may take action covered by approval

(2A)    An approval granted under this section is an approval of the taking of the action specified in the approval by any of the following persons:

(a)    the holder of the approval;

(b)    a person who is authorised, permitted or requested by the holder of the approval, or by another person with the consent or agreement of the holder of the approval, to take the action.

Notice of approval

(3)    The Minister must:

(a)    give a copy of the approval to the person named in the approval under paragraph 133(2)(c); and

(b)    provide a copy of the approval to a person who asks for it (either free or for a reasonable charge determined by the Minister).

Limit on publication of approval

(4)    However, the Minister must not provide under subsection (3) a copy of so much of the approval as:

(a)    is:

(i)    an exempt document under section 47 of the Freedom of Information Act 1982 (trade secrets etc.); or

(ii)    a conditionally exempt document under section 47G of that Act (business documents) to which access would, on balance, be contrary to the public interest for the purposes of subsection 11A(5) of that Act; or

(b)    the Minister believes it is in the national interest not to provide.

The Minister may consider the defence or security of the Commonwealth when determining what is in the national interest. This does not limit the matters the Minister may consider.

Notice of refusal of approval

(7)    If the Minister refuses to approve for the purposes of a controlling provision the taking of an action by the person who proposed to take the action, the Minister must give the person notice of the refusal.

Note:    Under section 13 of the Administrative Decisions (Judicial Review) Act 1977, the person may request reasons for the refusal, and the Minister must give them.

Definition

(8)    In this section:

assessment documentation, in relation to a controlled action, means:

(a)    if the action is the subject of an assessment report—that report; or

(b)    if Division 3A of Part 8 (assessment on referral information) applies to the action:

(i)    the referral of the proposal to take the action; and

(ii)    the finalised recommendation report relating to the action given to the Minister under subsection 93(5); or

(c)    if Division 4 of Part 8 (assessment on preliminary documentation) applies to the action:

(i)    the documents given to the Minister under subsection 95B(1), or the statement given to the Minister under subsection 95B(3), as the case requires, relating to the action; and

(ii)    the recommendation report relating to the action given to the Minister under section 95C; or

(d)    if Division 5 of Part 8 (public environment reports) applies to the action:

(i)    the finalised public environment report relating to the action given to the Minister under section 99; and

(ii)    the recommendation report relating to the action given to the Minister under section 100; or

(e)    if Division 6 of Part 8 (environmental impact statements) applies to the action:

(i)    the finalised environmental impact statement relating to the action given to the Minister under section 104; and

(ii)    the recommendation report relating to the action given to the Minister under section 105.

78    Section 134 deals with conditions of approvals. The conditions which may be attached are those which are necessary or convenient for protecting a matter protected by a relevant provision of Pt 3 or repairing or mitigating damage to a matter protected by Pt 3.

79    Subdivision B of Div 1 of Pt 9 (ss 136–140A) concerns “consideration for approvals and conditions”. Section 136 deals with general considerations and is as follows:

Mandatory considerations

(1)    In deciding whether or not to approve the taking of an action, and what conditions to attach to an approval, the Minister must consider the following, so far as they are not inconsistent with any other requirement of this Subdivision:

(a)    matters relevant to any matter protected by a provision of Part 3 that the Minister has decided is a controlling provision for the action;

(b)    economic and social matters.

Factors to be taken into account

(2)    In considering those matters, the Minister must take into account:

(a)    the principles of ecologically sustainable development; and

(b)    the assessment report (if any) relating to the action; and

(ba)    if Division 3A of Part 8 (assessment on referral information) applies to the action—the finalised recommendation report relating to the action given to the Minister under subsection 93(5); and

(bc)    if Division 4 of Part 8 (assessment on preliminary documentation) applies to the action:

(i)    the documents given to the Minister under subsection 95B(1), or the statement given to the Minister under subsection 95B(3), as the case requires, relating to the action; and

(ii)    the recommendation report relating to the action given to the Minister under section 95C; and

(c)    if Division 5 (public environment reports) of Part 8 applies to the action:

(i)    the finalised public environment report relating to the action given to the Minister under section 99; and

(ii)    the recommendation report relating to the action given to the Minister under section 100; and

(ca)    if Division 6 (environmental impact statements) of Part 8 applies to the action:

(i)    the finalised environmental impact statement relating to the action given to the Minister under section 104; and

(ii)    the recommendation report relating to the action given to the Minister under section 105; and

(d)    if an inquiry was conducted under Division 7 of Part 8 in relation to the action—the report of the commissioners; and

(e)    any other information the Minister has on the relevant impacts of the action (including information in a report on the impacts of actions taken under a policy, plan or program under which the action is to be taken that was given to the Minister under an agreement under Part 10 (about strategic assessments)); and

(f)    any relevant comments given to the Minister in accordance with an invitation under section 131 or 131A; and

(fa)    any relevant advice obtained by the Minister from the Independent Expert Scientific Committee on Coal Seam Gas and Large Coal Mining Development in accordance with section 131AB; and

(g)    if a notice relating to the action was given to the Minister under subsection 132A(3)—the information in the notice.

Note:    The Minister must also take into account any relevant comments given to the Minister in response to an invitation under paragraph 131AA(1)(b). See subsection 131AA(6).

Person’s environmental history

(4)    In deciding whether or not to approve the taking of an action by a person, and what conditions to attach to an approval, the Minister may consider whether the person is a suitable person to be granted an approval, having regard to:

(a)    the person’s history in relation to environmental matters; and

(b)    if the person is a body corporate—the history of its executive officers in relation to environmental matters; and

(c)    if the person is a body corporate that is a subsidiary of another body or company (the parent body)—the history in relation to environmental matters of the parent body and its executive officers.

Minister not to consider other matters

(5)    In deciding whether or not to approve the taking of an action, and what conditions to attach to an approval, the Minister must not consider any matters that the Minister is not required or permitted by this Division to consider.

80    Section 139 concerns “requirements for decisions about threatened species and endangered communities” and provides as follows:

(1)    In deciding whether or not to approve for the purposes of a subsection of section 18 or section 18A the taking of an action, and what conditions to attach to such an approval, the Minister must not act inconsistently with:

(a)    Australia’s obligations under:

(i)    the Biodiversity Convention; or

(ii)    the Apia Convention; or

(iii)    CITES; or

(b)    a recovery plan or threat abatement plan.

(2)    If:

(a)    the Minister is considering whether to approve, for the purposes of a subsection of section 18 or section 18A, the taking of an action; and

(b)    the action has or will have, or is likely to have, a significant impact on a particular listed threatened species or a particular listed threatened ecological community;

the Minister must, in deciding whether to so approve the taking of the action, have regard to any approved conservation advice for the species or community.

(emphasis added)

81    The posited duty of care was one concerned with the execution of the statutory duties, powers and functions by the Minister conferred by the statute. It is thus necessary to identify the duties, powers and functions of the Minister under the statute in order to assess whether the common law duty of care can be seen as harmonious, compatible and coherent with them.

82    The decision under ss 130 and 133 is required because, without approval, the taking of action by the extension project would be unlawful because the action has been determined by a delegate of the Minister to have, or that it will have, or that it is likely to have a “significant impact” on a listed threatened species in a number of categories or a listed threatened ecological community in a number of categories: see ss 18 and 18A; and to have, or that it will have, or that it is likely to have a “significant impact” on a water resource: see ss 24D and 24E.

83    The decision of the Minister is circumscribed by s 136. Subsection (1) provides for mandatory considerations. These are: (a) matters relevant to any matters protected by provision of Pt 3 that the Minister has decided (here, by her delegate under s 75) is a controlling provision (here, Subdivs C and FB of Div 1 of Pt 3 within Ch 2: ss 18, 18A, 24D and 24E, being matters relevant to the threat to species and communities and the impact on water resources); and (b) “economic and social matters”.

84    The Full Court (Kenny, Jessup and Middleton JJ) in Tarkine National Coalition Inc v Minister for the Environment [2015] FCAFC 89; 233 FCR 254 dealt with the scheme of the EPBC Act, and relevantly here, s 136. At 233 FCR 265–266 [25]–[28] and 269 [44]–[45], Jessup J said (with the agreement of Kenny J (at 256 [1]) and Middleton J (at 278 [70])) the following:

[25] Returning to s 136, I would make four observations about the structure and content of this section. First, subss (1) and (2) made a distinction between the matters that the Minister “must consider” (subs (1)) and the things that the Minister “must take into account” in considering those matters (subs (2)). The purpose of subs (1), as it seems to me, was to mark out the broad categories of consideration to which the Minister was required to turn his mind, and specifically to require consideration not only of the matters protected by Pt 3 of the EPBC Act but also of matters that, otherwise, appear to be of no concern under that Act, namely, “economic and social matters”. Neither para (a) nor para (b) of s 136(1) dealt, at the level of detail, with particular matters that required consideration. For example, what, if any, particular “social matter” might have required consideration in a proposal that came before the Minister was, it seems, a matter for the Minister.

[26] Secondly, the expression “matters relevant” in s 136(1) was not defined in the EPBC Act. By contrast, the expression “relevant impacts”, used in s 136(2)(e), was defined and gave content, at the level of detail, to the Minister’s obligation to take things into account. I shall return to this definition below.

[27] Thirdly, while the range of things that the Minister was to take into account under subs (2) was extensive, with the exception of those referred to in paras (a) and (e), each was a concrete document or some similar existing artefact. In effect, what the Minister had to take into account were the contents of those documents or artefacts. This approach to regulation is to be contrasted with a situation in which the things to be taken into account were identified by description, or generically, such as, for example, where a decision-maker was required to take account of the condition of the habitat of a particular species. Subject to the exceptions mentioned, the scheme of s 136 was one in which it was assumed that specific subjects of this and similar kinds were already dealt with in the documents or artefacts referred to. The role of the Minister was to take into account the things that were before him in this way, rather than being either obliged or entitled to undertake additional research or investigations.

[28] Fourthly, the terms of s 136(5) should be noted. While they require no further explanation, they confirm the impression that Subdiv B established a closed system of the matters that the Minister was to consider in making his decision, and the things that should be taken into account.

[44] In what I have written above, I have taken a course through the terms of the legislation that appears, to me at least, to have the greatest potential to yield a positive outcome for the appellant. In that regard, I have focused on s 136(2)(e) of the EPBC Act. In their submissions, counsel for the appellant placed some emphasis on para (a) of subs (1) of this section. As already suggested, I do not regard this as a provision from which the Minister’s obligation to take particular matters into account, at the level of detail, may be discerned. Rather, its purpose was, in a sense, categorical. The very same statutory formula was to be found in other, analogous, provisions of the EPBC Act: ss 37B(1), 145D(3) and 146F(1).

[45] With respect to s 136(1)(a), the primary judge referred to, and adopted, what had been said by North J in Blue Wedges Inc v Minister for Environment, Heritage and the Arts (2008) 167 FCR 463 at [115]:

Section 136(1)(a) left it to the Minister to decide what were the matters relevant to the protected matters which he should take into account. The section does not suggest that there was a defined set of specific matters to be taken into account such as might be intended if the section had referred to “all matters relevant” or “the matters relevant”.

Neither side on the appeal submitted that these paragraphs contained any error.

85    Thus, specific types of environmental impact (ss 18, 18A, 24D and 24E) must be considered in a broader context of economic and social matters. Inhering in these required considerations is the balance of specific environmental considerations and broader economic and social considerations. A statutory balancing of such considerations is mandated, directed to the subjects enlivening the duty to make the decision: the matters in ss 18, 18A, 24D and 24E of the EPBC Act.

86    Subsection 136(2) deals with factors to be taken into account. In considering the matters in subs (1), that is in the balance of the specific environmental matters relevant to ss 18, 18A, 24D and 24E with the economic and social matters, the Minister must take into account the matters in paras (2)(a)–(g), to the extent relevant.

87    The “economic and social matters” are, in their placement in para 136(1)(b), distinct from matters relevant to the affectation of the matters protected by Pt 3. They are apt to be the wider considerations of economic and social benefit that may weigh in the balance to approve the action notwithstanding adverse impacts on the matters protected: here, the species, communities and water resource. Nevertheless, it was not put by the Minister that consideration of the environmental effect of greenhouse gas emissions and the danger to humans therefrom could not be taken into account as a “social matter” for the purposes of para 136(1)(b). Thus, it was accepted by both parties that the Minister could take such into account.

88    The Minister’s responsibility at the point in time of the making the approval decision under the EPBC Act was confined in scope. The Minister only assumed responsibility for approving the extension of the mine because of its likely significant impact, not on climate change, the environment more generally, or the foreseeability of risk to health and safety of people in the future, but rather specified matters of national environmental significance, being in this instance listed threatened species and communities (ss 18 and 18A) and water resources (ss 24D and 24E) respectively. It is those matters to which the Minister’s binary statutory task under ss 130 and 133 of the EPBC Act was directed by the Act. Whilst the Minister must take into account broader considerations, including “economic and social matters” and the principle of “ecologically sustainable development”, the Minister’s fundamental responsibility, and therefore concern under the Act, was whether the extension of the mine should have been allowed to proceed notwithstanding its likely significant impact upon a matter of national environmental significance protected by Pt 3, in this case listed threatened species and communities and water resources. To state the Minister’s responsibility any higher than that is to misunderstand the position of the EPBC Act in its application and operation here, within the intergovernmental arrangements for the management of the environment between Commonwealth and State governments. It is to expand the Minister’s responsibility to address issues dealt with by New South Wales. The Minister’s position and responsibility is to be contrasted with the position of the relevant consent authority in New South Wales under the EPA Act (namely the IPC), which, in deciding whether or not consent should be granted, was given responsibility to consider whether the extension of the mine would be undertaken in an environmentally responsible manner, including that greenhouse gas emissions are minimised to the greatest extent practicable. In doing so, the IPC was required to have regard to both State and national “policies, programs or guidelines concerning greenhouse gas emissions”: see s 4.15(a)(i) of the EPA Act and cl 14(2) of the Mining SEPP. Unlike the Minister under the EPBC Act, greenhouse gas emissions was a fundamental concern for the IPC in exercising its statutory function.

89    Pursuant to a bilateral agreement between the Commonwealth and New South Wales, the environmental assessment process conducted by New South Wales under the EPA Act is accredited by the Commonwealth, and therefore a separate assessment of the impact of controlled actions is not required under Pt 8 of the EPBC Act: see s 83(1). Rather, the Minister in this case was provided with the assessment report of the NSW Department and the IPC’s reasons before making her decision, the former of which was required to contain sufficient information about the relevant impacts of the Extension Project for the purposes of each controlling provision: s 47(4). Whilst those reports before the Minister also contained detailed assessment of the likely impact on greenhouse gas emissions of the extension of the mine, this was due to the New South Wales legislative scheme for environmental approvals (as accredited by the Commonwealth) and thus its responsibilities for the environment, not by any requirement of the EPBC Act or bilateral agreement between the Commonwealth and New South Wales. The Minister’s statutory function directed her primarily in those reports to the State’s assessment of the impact of the extension of the mine to listed threatened species and communities and water resources. Nevertheless the wider considerations of the IPC and the NSW Department were to be before the Minister.

90    As will be developed below, recourse to the terms of the EPA Act and the function of the relevant consent authority in New South Wales is not to suggest that the posited duty of care cannot be sustained because of the New South Wales legislative regime for environmental approvals, by reason of incoherency or otherwise. Rather, that scheme and its operation here serves to highlight the incoherency of the posited duty with the EPBC Act, with proper appreciation of the Minister’s statutory task under ss 130 and 133 within the inter-governmental arrangements for the management of the environment in Australia.

91    Section 520 of the EPBC Act provides for regulations to be made under the Act. Section 520(3) provides for regulations for and in relation to various international agreements. The only agreement referable to climate change in subs (3) is the United Nations Framework Convention on Climate Change, opened for signature 4 June 1992, 1771 UNTS 107 (entered into force 21 March 1994) (UNFCCC). No other agreements referable to climate change are mentioned. No regulations under subs (3) have been promulgated.

92    The EPBC Act did not require the Minister in this case, for the reasons below, to take into account the risks of personal injury or death resulting from an increase in greenhouse gas emissions consequent upon the Minister’s decision and its resultant impact upon climate change. Though the parties approached the appeal on the basis that such could be taken into account. As described above, the climate change and global warming consequences of combusting the coal mined were considered in detail by the NSW Department and IPC under the relevant State legislation, which views and analysis were before the Minister in making her decision under s 133 of the EPBC Act, in a process that reflected the division of co-ordinate responsibility for the protection of the environment between the Commonwealth and the State and Territories, and the EPBC Act’s particular role and function within the Commonwealth’s sphere of responsibility.

The legislative context of the EPBC Act

93    The co-ordinated division of responsibility between potentially overlapping responsibilities for the protection of the environment is to be seen not only in the text of the EPBC Act, but also in the context to it, being the inter-governmental review of the subject and agreement that led to the passing of the EPBC Act.

94    The EPBC Act flowed from agreement reached at the Council of Australia Governments (COAG) in 1997. COAG had undertaken a review of Commonwealth and State roles and responsibilities for the environment. This review was described in the Explanatory Memorandum to the Environment Protection and Biodiversity Conservation Bill 1999 (Cth) at p 8, as follows:

The Review of Commonwealth/State Roles and Responsibilities for the Environment was conducted by a senior level Working Group of the Intergovernmental Committee for Ecologically Sustainable Development.

In November 1996 the Government endorsed the objectives and approaches pursued by the Commonwealth in the Review. In September 19097 the Government agreed its position for both the final negotiations and the COAG meeting which considered the reforms resulting from the Review. The Government also noted that amendments to Commonwealth environment legislation will be required to implement the outcomes of the COAG Review, to proceed immediately after the Review had been concluded.

In November 1987 COAG gave in-principle endorsement to a Heads of Agreement on Commonwealth/State Roles and Responsibilities for the Environment. Fundamental changes to Commonwealth Environmental Legislation are required to give effect to the Agreement. A majority of States and Territories have now signed the agreement, and it is a Government priority to introduce legislation into Parliament to implement to agreement. There is an expectation, particularly on the part of business and industry, that the Government will introduce legislation quickly to provide certainty of outcome for the review process and deliver its benefits to the community.

95    The Explanatory Memorandum at pp 8-9 then explained the objectives of the EPBC Bill flowing from the COAG review:

In summary, the major outcomes of the Review process to be reflected in the Environment Protection and Biodiversity Conservation Bill 1998 are:

-    The Commonwealth focussing on matters of national environmental significance. This will result in the Commonwealth not being involved in matters of only State or local significance.

-    That for activities or proposals involving both the Commonwealth and a State, the Commonwealth environmental assessment and approval process will be triggered only by those actions which may have a significant impact on matters of national environmental significance. This will overcome the problem of Commonwealth legislation being triggered in an indirect manner by Commonwealth decisions that are not directly related to the environment, such as export approval and foreign investment, and funding decisions.

-    Improving the efficiency and timeliness of environmental and development approvals processes;

-    Greater transparency and certainty in decision making in relation to development proposals;

-    A reliance on State processes and management approaches which will, as appropriate accommodate Commonwealth interests;

-    Recognition of the Commonwealth’s role in international and national environmental matters with strengthened Commonwealth/State partnership arrangements for dealing with these matters.

96    In the second reading speech (29 June 1999, House of Representatives, Hansard at p 7770) the following was stated:

Over the last three years, the federal coalition government has worked cooperatively with the state governments to identify the reforms needed to produce a more effective and efficient national approach to environmental management. The result was an agreement, given in-principle endorsement by the Council of the Australian Governments in 1997, which defines the Commonwealth’s role by reference to certain matters of national environmental significance. The COAG agreement also seeks to ensure the seamless integration of Commonwealth and state laws through a transparent mechanism for Commonwealth accreditation of state processes.

The Environment Protection and Biodiversity Conservation Bill 1999 implements the COAG agreement. In doing so, it provides the framework for a more effective national approach to environmental management, ensuring resources are focused on delivering better environmental outcomes at all levels of government. The Commonwealth’s role in this national approach will, for the first time, be clearly and logically defined.

97    The COAG agreement, the EPBC Bill, and the EPBC Act reflected COAG’s and Parliament’s desire to develop a more effective framework for intergovernmental collaboration and a division of responsibilities with respect to matters concerning the environment within the broader context of the Australian federal system of government, as reflected in the methods of achieving the objects of the EPBC Act, especially in paras 3(2)(b), (c) and (g) of the EPBC Act set out at [48] above.

98    Some particular aspects of the COAG Agreement should be noted. Clause 4 recognised that “the Commonwealth’s environmental assessment and approval processes should only be triggered by proposals which may have a significant impact on matters of national environmental significance as described in Pt I of Attachment 1, which notably did not contain reference to matters concerning greenhouse gas emissions. Clause 5 agreed that in matters of national environmental significance, State processes would be relied upon as the “preferred means of assessing proposals”, with Commonwealth decisions limited “to only those aspects of proposals [which have a significant impact upon: cl 4] matters of national environmental significance”. While noting the Commonwealth had interests and obligations relating to reducing emissions of greenhouse gases under the UNFCCC, the agreement specifically stated that such a matter should not be a trigger, of its own, for the Commonwealth’s environmental assessment and approval processes (cl 4).

The reasons of the primary judge

99    The primary judge expressed his reasons under nine headings, considering, respectively: the parties and their claims; the application for approval to extend the coal mine; the risk of harm; whether a duty of care was owed by the Minister to the Children; the affirmative salient features; the negative salient features; conclusions on the duty of care; whether an injunction should have been issued; and conclusions and further steps. Those reasons comprise a close engagement, under the heading “the risk of harm”, with the expert evidence as to the consequences of climate change, and, under the headings pertaining to the existence of the putative duty and the relevant salient features, with the authorities relevant to the imposition of a novel duty of care. The reasons of the primary judge are careful and comprehensive, and the recitation of those reasons that follows must necessarily be (to an extent) cursory, although it is embarked upon with the intention that those reasons will be engaged with more fully when setting out the submissions of the parties, and the consideration thereof. The structure of the judgment below will be largely adhered to, for ease of reference.

100    By way of introduction to the structure of the judgment and the approach of the primary judge, three matters are of particular relevance to an understanding of the conclusion as to duty and, with great respect to his Honour, to an appreciation of the errors involved in that conclusion.

101    The first matter is the misconstruction of the EPBC Act. As earlier discussed, the Act is not concerned generally with the protection of the environment nor with any response to global warming and climate change. The Commonwealth has its particular concerns and focus, and the decision in question also has its particular concerns and focus. Nor, for the reasons later set out is the protection of the interests and safety of human beings in the environment a primary object of the Act, nor is human safety an implied mandatory consideration in the exercise of the Minister’s statutory function under ss 130 and 133.

102    Secondly, against the background of these misconstructions of the EPBC Act, the primary judge focused first and overwhelmingly in his reasons upon risk of harm and the reasonable foreseeability of harm, at the cost, with respect, of appreciating the nature of the relationship between the Minister and the Children.

103    Thirdly, and related to both the first and second matters, the primary judge gave inadequate consideration to the true relationship between the Minister and the Children, and in that respect gave inadequate, indeed with respect, cursory consideration to the questions of policy and the nature of the enquiry that would be thrown up at the point of breach by the imposition of the duty, including the inappropriateness of the judicial branch of government dealing with matters raised by the duty. This was despite his Honour’s clear finding at J[479] (as cannot be doubted) that the Executive branch (and one might add, Parliament) “is better placed to deal with the complex task of addressing climate change than the common law”.

The parties and their claims: J[4]–[17]

104    The applicants were eight (now six) Australian children who brought the proceeding on their own behalf and on behalf of the children they represent, seeking a declaration that a duty of care was owed by the Minister to exercise her powers under ss 130 and 133 of the EPBC Act with reasonable care so as not to cause them harm, and an injunction restraining an apprehended breach of that duty by the Minister in exercising her discretion under the EPBC Act in favour of the approval of a substantial extension of the Vickery Coal Project, being a coal mine in Northern New South Wales: J[4], [9]–[10]. The particular harm that was apprehended was originally framed as personal injury (being mental or physical harm), as well as economic and property loss and damage: J[11]. The primary judge concluded, however, that the duty of care could only pertain to personal injury and death: J[148]. The duty was framed such that it was only owed to the Children on the basis that the harm will occur towards the end of this century, and that only today’s children will live to experience that harm: J[12].

The risk of harm: J[29]–[90]

105    In the third section of his Honour’s reasons, the primary judge set out the unchallenged expert evidence as to the different possible “future worlds” that the world and all humanity (not just the Children) collectively face at this critical juncture: J[33]. The primary judge set out in broad strokes the nature of the greenhouse effect at J[37]–[40], reciting the conclusion of Professor Steffen at J[42] that the concentration of atmospheric CO2 is currently rising at a rate of 2.5 parts per million per year, which is driving increasing temperatures at a rate of 0.24℃ per five-year period, or nearly 0.5℃ per decade.

106    At J[30]–[32], the primary judge described the applicants’ case as follows:

[30] In a nutshell, the applicants’ case is that the scientific evidence demonstrates the plausible possibility that the effects of climate change will bring about a future world in which the Earth’s average surface temperature (currently at about 1.1°C above pre-industrial temperature levels) will reach about 4°C above pre-industrial temperature levels by about 2100. Supported by unchallenged expert evidence, the applicants contended that a 4°C future world may come about in one of two ways: first, where the greenhouse effect upon the Earth’s increasing temperature is driven by an approximately linear relationship between increased human emissions of CO2 and increased temperatures, and second, in circumstances where continuing human emissions of CO2 will result in ‘Earth System changes, which diminish the Earth’s current ability to reflect heat, absorb CO2, and retain CO2 currently held in carbon sinks, triggering ‘tipping cascades’ which propel the Earth into a 4°C trajectory. That scenario was referred to in the evidence as “Hothouse Earth”. Under this scenario, humans will lose the capacity to control climate change and global surface temperatures will continue warming even if human emissions of CO2 are curtailed.

[31] Further, the unchallenged evidence of the applicants is that the best available outcome that climate change mitigation measures can now achieve is a stabilised global average surface temperature of 2°C above pre-industrial levels. However, at that temperature and beyond, there is an exponentially increasing risk of the Earth being propelled into an irreversible 4°C trajectory because of ‘Earth System’ changes.

[32] Given the plausible prospect of Earth’s temperature stabilising at 4°C or greater if stabilisation at 2°C is not achieved, the applicants contended that 100 Mt of CO2 emissions, attributable to the Extension Project, will be significant and material to future increased global average surface temperatures. This, in turn, will expose the Children to a greater risk of injury.

107    At J[44]–[53], the primary judge considered the “Earth System” and the possibility of “feedback processes” such as melting ice or forest dieback which accelerate the warming of the Earth System and thus compound climate change: J[50]. The closer the average global temperature becomes to a 2℃ increase from pre-industrial levels, the more likely the activation of those feedback processes which may have a ripple effect on each other: J[51]. At J[54], the primary judge set out Professor Steffen’s evidence as to the sobering effects of human CO2 emissions experienced to date.

108    The primary judge relied principally upon the report of Professor Steffen and reports prepared by the Commonwealth Scientific and Industrial Research Organisation and the Bureau of Meteorology. It is unnecessary at this point to set out the detailed reasoning of the primary judge and the detailed descriptions of the role played by greenhouse gases in warming of the earth’s surface temperature. Professor Steffen posited various possible “future worlds”. The most benign would result in global average surface temperature rise of below 2ºC above pre-industrial level by 2100. The least benign would lead to a temperature rise of 4ºC above pre-industrial level by 2100. The first and most benign scenario accorded with the Paris Agreement that had a target of limiting temperature rise to “well below 2ºC” with an ambition to limit temperature to 1.5ºC above pre-industrial level. Professor Steffen saw this scenario now as unlikely. He considered that the lowest realistic increase would be 2ºC. At J[66], the primary judge set out a fundamental point made by Professor Steffen:

A fundamental point made by Professor Steffen’s analysis is that if sufficient measures are not taken to reduce human emissions of CO2 so as to stabilise surface temperature at 2°C, global average surface temperatures will then enter an irreversible 4°C Future World or Hothouse Earth trajectory. Professor Steffen opined that ‘feedback processes’ will be activated by a 3°C (or even lower) temperature rise with a consequent “significant risk” that a tipping cascade will be triggered taking the global average surface temperature beyond 3°C and onto the 4°C Future World trajectory. That is depicted in Figure 4 of Professor Steffen’s report.

109    At J[67]–[69], the primary judge set out the global impacts and impacts in Australia of the 2ºC, 3ºC and 4ºC “future worlds”. The 3ºC and 4ºC future worlds would be catastrophic. The evidence was that for Australia the following can be expected for the 2º C, 3º C and 4º C “future world”:

[67] In relation to each of the three scenarios postulated by Professor Steffen, he described the projected global impacts followed by a description of the impacts in Australia. He noted that the risks and impacts described were linked to the stabilisation of the global average surface temperature for each of the three scenarios. He emphasised that stabilisation will take multiple decades at a minimum and stated that, therefore, the risks and impacts described were relevant to the current generation of children and the following generation or two.

Scenario 1: Stabilisation at a rise in global average surface temperature of about 2℃ above the pre-industrial level (IPCC 2018).

    37% of the global population will be exposed to extreme heat at least once every five years. This will have severe impacts on human health and wellbeing, as well as on worker productivity.

    Sea-level will rise by 0.46 m by 2100, leading to large increases in coastal flooding, saltwater intrusion in low-lying areas, and more damaging storm surges. The most vulnerable countries include small island states, Bangladesh, low-lying Southeast Asian cities and settlements, and many regions along the African coast.

    99% of coral reefs will be dead from severe bleaching; this means that the Great Barrier Reef will cease to exist as we know it today, as well as other coral reefs around the world.

    A decline of 3 million tonnes in marine fisheries, with the most severe impacts on developing countries that rely on marine fish for a large fraction of protein in their diets.

    Ecosystems will shift to a new biome on 13% of Earths land, leading to large rates of extinctions as well as a surge in invasive species as individual organisms migrate in response to a changing climate.

    6.6 million square kilometres of Arctic permafrost will thaw, releasing large amounts of CO2 and methane to the atmosphere, accelerating the warming trend.

    7% reduction in maize harvests in the tropics, with the poorest countries suffering the most damaging impacts.

    16% of plant species will lose at least half of their current range, leading to significant within-ecosystem changes as well as an increase in extinction rates.

For Australia, Scenario 1 would significantly increase the likelihood in any given year of extreme weather events (King et al. 2017): (i) 77% likelihood of severe heatwaves, power blackouts and bushfires; and 74% likelihood of severe droughts, water restrictions and reduced crop yields. More generally, CSIRO and BoM 2020, have used simulations from the latest generation of climate models to project changes to Australias climate over the next few decades. These projections would thus be relevant for a 1.5-2℃ world, and thus provide useful insights for Scenario 1:

    Continued warming, with more extremely hot days and fewer extremely cool days.

    A decrease in cool season rainfall across many regions of the south and east, likely leading to more time spent in drought.

    A longer fire season for the south and east and an increase in the number of dangerous fire weather days.

    More intense short-duration heavy rainfall events throughout the country.

    Fewer tropical cyclones, but a greater proportion projected to be of high intensity, with ongoing large variations from year to year.

    Fewer east coast lows particularly during the cooler months of the year. For events that do occur, sea level rise will increase the severity of some coastal impacts.

    More frequent, extensive, intense and longer-lasting marine heatwaves leading to increased risk of more frequent and severe bleaching events for coral reefs, including the Great Barrier and Ningaloo reefs.

    Continued warming and acidification of its surrounding oceans.

    Ongoing sea level rise. Recent research on potential ice loss from the Antarctic ice sheet suggests that the upper end of projected global mean sea level rise could be higher than previously assessed (as high as 0.61 to 1.10 m global average by the end of the century for a high emissions pathway, although these changes vary by location).

    More frequent extreme sea levels. For most of the Australian coast, extreme sea levels that had a probability of occurring once in a hundred years are projected to become an annual event by the end of this century with lower emissions, and by mid-century for higher emissions.

[68] The effects forecast by Professor Steffen for a 3℃ Future World were as follows:

Scenario 2: Stabilisation at a rise in global average surface temperature of about 3℃ above the pre-industrial level. Here I focus on projected impacts on Australia of this scenario, based on a recent assessment by the Australian Academy of Sciences (Hoegh-Guldberg et al. 2020, and references therein):

    Many of Australias ecological systems, such as coral reefs and forests, would be unrecognisable, accelerating the decline or Australias natural resources through the loss or change in the distribution of thousands of species and ecological processes. (As noted for scenario 1, the Great Barrier Reef will no longer exist at temperature rises of 2℃ or more).

    Much larger climate change-driven changes to water resources are likely, leading to increasingly contested supplies for natural flows, irrigated agriculture and other uses.

    At 3℃, living in many Australian cities and towns would be extremely challenging due to more frequent and severe extreme weather events, including much higher temperatures and more severe water shortages.

    Sea levels will rise by 0.4 to 0.8 metres by 2100 and by many metres over subsequent centuries. These changes will cost hundreds of billions of dollars over coming decades as coastal inundation and storm surge increasingly impact Australias coastal communities, infrastructure and businesses. Between 160,000 and 250,000 properties are at risk of flooding when sea levels rise to 1 metre above pre-industrial.

    The probability of large-scale extreme events, such as large storms, floods, droughts, hail storms, tropical cyclones, heatwaves and other climate-related phenomena will increase rapidly.

    High fire danger weather will increase significantly, leading to more catastrophic fire seasons such as the 2019/2020 Black Summer fires.

    Grain, fruit and vegetable crops will suffer more severe reductions in yields in a 3℃ world, and rising heat stress will negatively affect extensive and intensive livestock systems.

    Rural communities will face increasingly harsh living conditions due to increasing debt from diminishing crop yields, insurance losses from worsening extreme weather events, and more challenging working conditions due to increasing extreme heat.

    Australia at 3℃ will be hotter, drier and more water stressed with impacts on water security, availability, quality, economies, human health and ecosystems. Many locations in Australia in a 3℃ world would be very difficult to inhabit due to projected water shortages.

    Multiple impacts of a 3℃ world would damage the health and wellbeing of Australians. These include escalating heat stress, more frequent and intense bushfires, reduced access to food and water, increasing risk of infectious disease, and deteriorating mental health and general wellbeing.

[69] The projected effects of a 4 Future World were described by Professor Steffen as follows:

Scenario 3: The Hothouse Earth scenario, with stabilisation in the 22nd century at a global average surface temperature level at least 4℃, and probably higher, above the pre-industrial level. There has been much less research on the impacts of a 4-5℃ temperature rise in global average surface temperature. However, a few of the potential impacts that could arise from such a high level of warming were summarised in Steffen et al. (2018: Supplementary Information). These include:

    Multiple impacts on agricultural regions, including depletion of soil fertility, changes in water availability and loss of coastal agricultural lands, with the risk of widespread starvation in the most vulnerable regions and/or large migrations out of those regions, increasing the risk of conflict elsewhere.

    Destruction of coral reefs from ocean warming and acidification, and consequent loss of livelihoods for those communities and societies dependent on reefs.

    Amazon rainforest at risk of conversion to savanna from both climate and land-use change. This would lead to large releases of CO2 to the atmosphere as well as large increases in extinction rates of species that depend on the rainforest.

    Tropical drylands at risk of becoming too hot and dry for agriculture, and too hot for human habitation. This has very large implications for many regions in Africa in particular, but also parts of Asia and much of Australia (see below).

    Very large risks from coastal flooding to transport, infrastructure and coastal ecosystems. Economic damages could trigger regional or global economic collapse as major coastal cities on all continents become uninhabitable.

    Reliability of South Asian (Indian) Monsoon vulnerable to high aerosol loading and to the warming of the Indian Ocean and adjacent land. Well over 1 billion people in south Asia depend on a reliable monsoon system. Failure of the monsoon would very likely lead to large-scale starvation, migration and conflict.

    Mountain glaciers melting at rapid rates, changing amount and timing of run-off. Freshwater resources of over 1 billion people at risk.

    Large changes to riparian and wetlands, with loss of water of some places and increased flooding in others.

For Australia, the corresponding impacts (harms) of Scenario 3 are:

    Much of Australias inland areas (savanna and semi-arid zones) will become uninhabitable for humans, except for artificial enclosed environments.

    The southeast and southwest agricultural zones will become largely unviable, due to extreme heat and a reduction in cool season rainfall. This would lead to a large depopulation of regional Australia.

    Australias large coastal cities (Brisbane, Sydney, Melbourne, Adelaide, Perth) will suffer increasing inundation and flooding from storm surges as sea level rises to metres above its pre-industrial level over the coming centuries. This will drive severe economic challenges, both because of direct damage from flooding and the large costs of adaptation.

    The Great Barrier Reef will no longer exist.

    Most of the eastern broadleafed (eucalypt forests) will no longer exist due to repeated, severe bushfires.

110    At J[70]–[73], the primary judge set out the steps that must be taken in order to achieve a 2℃ Future World. At J[71] his Honour summarised the conclusions of Professor Steffen that there is a 67% probability of achieving a 2℃ Future World if cumulative CO2 emissions from 2021 onwards are restricted to approximately 855 Gt of CO2, which would require all major emitting countries to reach net-zero emissions by 2050. At J[72], his Honour recited the expert evidence that 90% of Australia’s existing coal reserves cannot be burnt to be consistent with a 2℃ temperature target. At J[73], his Honour set out Prof Steffen’s conclusions from various research using a “carbon budget framework” as follows:

The definition of “reserves” used by McGlade and Ekins would appear to include the 100 Mt of coal from the Extension Project, it being “economically and technologically viable to exploit now”. On the basis of the carbon budget analysis used by McGlade and Ekins to predict a 50% probability of meeting a 2 Future World, Professor Steffen offered this conclusion:

The obvious conclusion from the carbon budget analysis above is that currently operating coal mines must be phased out as soon as possible (preferably no later than 2030), and that no new coal mines, or extensions to existing coal mines, can be allowed.

111    The “carbon budget” approach was referable to research of McGlade and Ekins in 2015 and Professor Steffen’s own work. The IPC appeared to be aware of this approach and dealt with it in [220] of its report (see [41] above), largely by reference to prevailing Commonwealth and New South Wales government policy on Scope 3 emissions.

112    The primary judge then turned to his conclusions on the question of harm. At J[74] the primary judge set out three “plausible scenarios” demonstrated by the evidence:

(i)    the Paris Agreement target of limiting global average surface temperature to well below 2°C, with the ambition to limit temperature to 1.5°C above the pre-industrial level, is now unlikely to be achieved without significant overshoot;

(ii)    the best future stabilised global average surface temperature which can be realistically contemplated today, is 2°C above the pre-industrial level; and

(iii)    if the global average surface temperature increases beyond 2°C, there is a risk, moving from very small (at about 2°C) to very substantial (at about 3°C), that Earth’s natural systems will propel global surface temperatures into an irreversible 4°C trajectory, resulting in global average surface temperature reaching about 4°C above the pre-industrial level by about 2100.

113    The primary judge said the following at J[75]:

Furthermore, the evidence demonstrates that the risk of harm to the Children from climatic hazards brought about by increased global average surface temperatures, is on a continuum in which both the degree of risk and the magnitude of the potential harm will increase exponentially if the Earth moves beyond a global average surface temperature of 2°C, towards 3°C and then to 4°C above the pre-industrial level.

114    The primary judge summarised the parties’ positions as to the prospective causal relationship between the emission of 100 Mt of CO2 which would result from the extension of the mine, and any future personal injury or death occasioned to the Children as a result of anthropogenic climate change. The relevant connection between the controlled action and the increased risk of harm is as to the contribution of 100 Mt of CO2. (As shall be seen in ground 5(d) of the appeal, the Minister contested before us the legitimacy of his Honour’s assumption or conclusion that the extension of the mine would produce an additional 100 Mt of CO2 emissions.) At J[80] and J[84] the primary judge set out the applicants’ case:

[80] The applicants contended that the 100 Mt of CO2 from the Extension Project would make a material contribution to future increases in the global surface temperature and thus the degree and magnitude of the risk of harm faced by the Children. That was put in two ways although primary reliance was placed on the second. First, the applicants contended that the approval, extraction, export and combustion of carbon from the Extension Project will emit a material quantity of CO2 into the atmosphere. They contended that the more CO2 that is emitted, the higher the level of CO2 concentration will be before it reaches its zenith. The higher the level of CO2 concentration when it reaches its zenith, the worse the harm to today’s children will be.

[84] The second way the case was put by the applicants was to adopt what an economist might call a marginal analysis. This contention was made by reference to the contribution that 100 Mt of CO2 may have on the level at which the global average surface temperature will stabilise. In that respect, the applicants first relied on Professor Steffen’s evidence that CO2 emissions from the Extension Project “would increase the level at which atmospheric CO2 concentration is eventually stabilised, and thus would increase the level at which the global average surface temperature is eventually stabilised”. The applicants then relied on the Future World scenarios identified already and the propositions set out at [74] above including that there is a risk, moving in degree from very small to very substantial as the global average surface temperature increases from 2 to 3 above the pre-industrial level, that a ‘tipping cascade’ will trigger a 4 Future World trajectory. The applicants contended that once global average surface temperatures reach or exceed 2 above the pre-industrial level, the risk of a 4 Future World increases exponentially and that with that heightened realm of risk in prospect, the emission of an additional 100 Mt of CO2 is material. On that basis and given that the evidence demonstrates an increase in both the degree and magnitude of risk of harm to the Children as between a 2 Future World and a 4 Future World, the applicants contended that the emission of 100 Mt of CO2 in the context of the risk profile just described, is a material contribution to the risk of exposure to harm.

115    At J[83] in dealing with the Minister’s submission that the 100 Mt of carbon dioxide would only contribute a tiny fraction (one eighteen-thousandth of a degree celsius) to temperature rise, an increase in magnitude said by the Minister be de minimis, the primary judge said: “I am unable to say that the evidence itself demonstrates the extent, if any, that a fractional increase in global temperature of the kind in question poses an additional risk of harm to the Children.” But his Honour noted (also at J[84]) that this did not answer the applicants’ case (by which phrase his Honour was referring to the first case):

… they argue that it is the accumulation of CO2 which causes exposure to the risk of harm and accumulated CO2, including the contribution to that accumulation which the 100 Mt of CO2 will make, that will bring about increased temperatures and the harm that the evidence demonstrates will follow. In that way, the applicants say there will be a material contribution to injury.

116    Later in his reasons at J[253] in drawing a conclusion about the foreseeability of harm, the primary judge said (to some degree inconsistently with what he had said at J[83] above):

I accept that, even on the marginal risk assessment referred to at [84], the prospective contribution to the risk of exposure to harm made by the approval of the extraction of coal from the Extension Project may be characterised as small. It may fairly be described as tiny. …

117    The primary judge stated this in the context of dealing with the second way the respondents (applicants below) put their case, set out at J[84] (see [114] above). If, as that case put it, and as Professor Steffens put it, only a small increase could tip the world into a cascade once the surface temperature reaches above 2, even a tiny contribution bringing the world’s temperature to or above 2 can be characterised as a material contribution.

118    This is not really a different “way” of putting the case. The cases at J[80] and [84] are really bound together.

119    Further at J[86]–[87] the primary judge said:

[86] Putting aside for the moment what I think is a mischaracterisation of the applicants’ case, there is not sufficient evidence before me on which I could conclude that there is no real prospect of the 100 Mt of CO2 being burnt outside the available fossil fuel budget necessary to meet a 2℃ target. The Minister called no evidence. The Minister essentially contended that the Court should infer that the 100 Mt of CO2 would likely be emitted in accordance with the Paris Agreement. There is no sufficient basis for that inference. The Minister relied upon little else than speculation, in circumstances where the evidence showed that at least one of the potential consumers of the coal is not a signatory to the Paris Agreement.

[87] Further and in any event, there is evidence before me which tends to support the proposition that the 100 Mt of CO2 will not be emitted as part of the available carbon budget necessary to achieve a 2℃ target. Professor Steffen’s opinion was that it was “obvious” from the carbon budget analysis, that “no new coal mines, or extensions to existing coal mines, can be allowed”. There can be no doubt that in making that statement Professor Steffen had the Extension Project in mind. True it is that he did not go on to explain why, but to say it is “obvious” by reference to the carbon budget analysis he relied on implies that the reason is to be found in his prior reliance on the study made by McGlade and Ekins, who had analysed the position for Australia and had calculated that over 90% of Australia’s existing coal reserves cannot be burnt to meet a 2℃ target. That observation reveals the logic behind Professor Steffen’s conclusion and it is logic which may be relied upon irrespective of whether the conclusion he proffered was based upon his specialist expertise. If there is no capacity to include 90% of existing Australian reserves of coal in the carbon budget, it seems unlikely that a capacity for new reserves to be included exists. Even “existing” reserves, by which Professor Steffen must have meant those already being exploited, logically have only a 1 in 10 chance of being included in the budget. There is no evidence sufficient to support a contention that the 100 Mt of CO2 from the Extension Project is earmarked for some priority treatment relative to other coal sufficient to put it in the top 10% of candidates for inclusion in the budget.

120    At J[90], the primary judge set out what he accepted as to the applicants’ case as to the 2 Future World:

In any event, the applicants did not say that a 2 Future World is the most likely scenario. Their contention was that a 2 Future World is a plausible possibility in circumstances where at temperatures at or slightly lower than 2℃, there is a small (but non-zero) probability that a tipping cascade will trigger a 4℃ Future World trajectory. Professor Steffen’s unchallenged evidence establishes that trajectory as a plausible scenario, should the global average surface temperature exceed 2 or slightly lower. That was a necessary element of the applicants’ contention and it was established.

The existence of a duty of care: J[91]–[183]

121    The primary judge then turned to the principal issue in the proceedings: the existence of the posited (novel) duty of care. At J[96]–[115] his Honour summarised the legal principles applicable to the ascertainment of a novel duty of care. At J[116]–[137] the primary judge traced the emergence of the law of negligence from its roots in the assize of nuisance. At this point, it is unnecessary to engage with his Honour’s expression of principle other than to say that the proper approach (broadly recognised by his Honour) is to undertake a close analysis and evaluation of the facts bearing on the relationship between the applicant and putative tortfeasor by reference to salient or relevant features or factors affecting the appropriateness of imputing a legal duty to avoid harm and where the duty posited concerns the exercise of statutory powers or functions the necessary starting point for any analysis is the legislation in question. Any posited duty must be coherent, compatible and harmonious with the duties, powers and functions of the repository and with the structure and purpose of the statute: Caltex Oil (Australia) Pty Ltd v The Dredge “Willemstad” [1976] HCA 65; 136 CLR 529 at 576–577 (Stephen J); Perre v Apand Pty Ltd [1999] HCA 36; 198 CLR 180 at 192 [5] and 194–195 [11]–[15] (Gleeson CJ), 218–231 [100]–[133] (McHugh J), 252–261 [196]–[221] (Gummow J), 300–307 [330]–[348] (Hayne J) and 326–327 [406]–[413] (Callinan J); Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; 200 CLR 1 at 13 [3] (Gleeson CJ, agreeing with McHugh J), 23–24 [42]–[43] (Gaudron J), 39–51 [93]–[133] (McHugh J), 96–97 [270]–[272] (Hayne J) and 113–117 [343]–[360] (Callinan J); Modbury Triangle (2000) 205 CLR 254 at 262–269 [13]–[36] (Gleeson CJ) and 288 [98] and 291 [108]–[118] (Hayne J, agreeing with Gleeson CJ); Sullivan v Moody [2001] HCA 59; 207 CLR 562 at 577–583 [42]–[64] (the Court); Tame v New South Wales [2002] HCA 35; 211 CLR 317 at 329–335 [6]–[28] (Gleeson CJ), 341 [53]–[54] (Gaudron J), 361 [123]–[125] (McHugh J), 397–399 [237]–[241] (Gummow and Kirby JJ) and 425–431 [323]–[336] (Callinan J); Graham Barclay Oysters 211 CLR at 555–564 [9]–[40] (Gleeson CJ), 570 [58] (Gaudron J, agreeing with Gummow and Hayne JJ), 577–583 [84]–[99] (McHugh J), 596–610 [145]–[186] (Gummow and Hayne JJ), 617 [213] and 629–631 [245]–[251] (Kirby J) and 663–664 [320]–[321] (Callinan J); Woolcock Street Investments Pty Ltd v CDG Pty Ltd [2004] HCA 16; 216 CLR 515 at 529–533 [19]–[33] (Gleeson CJ, Gummow, Hayne and Heydon JJ) and 547–560 [74]–[116] (McHugh J); Thompson v Woolworths (Qld) Pty Ltd [2005] HCA 19; 221 CLR 234 at 243 [24] (the Court); Vairy v Wyong Shire Council [2005] HCA 62; 223 CLR 422 at 442–448 [58]–[78] (Gummow J); Roads and Traffic Authority (NSW) v Dederer [2007] HCA 42; 234 CLR 330 at 345 [44] (Gummow J); Stuart v Kirkland-Veenstra [2009] HCA 15; 237 CLR 215 at 248–254 [87]–[114] (Gummow, Hayne and Heydon JJ) and 259–265 [130]–[146] (Crennan and Kiefel JJ).

122    The primary judge discussed (at J[116]–[137]) some aspects of the law’s adaption to altering social conditions and at J[138]–[142] the methodology of the common law. At J[140], the primary judge described the applicants’ approach here, as follows:

The applicants submitted that their case proceeds by analogical reference to two categories of negligence. First, the applicants referred to the Speirs line of authority (Caledonian Collieries Ltd v Speirs (1957) 97 CLR 202), which holds that statutory powers must be exercised with reasonable care and that the common law may impose liability for harm caused by their negligent exercise. Second, the applicants referred to what they called the Rylands v Fletcher line of authority, which holds a person liable for harm caused by dangerous things which escape from that person’s land.

The statutory scheme

123    After a brief introduction at J[143]–[148] of the salient features relied upon by the parties, the primary judge commenced (at J[149]–[183]) with the statutory scheme of the EPBC Act. With respect, that was the correct place to commence to understand the relationship of the Minister and the Children.

124    The primary judge referred (at J[149]–[159]) to the nature of the power in ss 130 and 133 and the objects of the EPBC Act, and in particular paras 3(1)(a) and (b), s 3A (especially para (c)), and the scheme of protection in Ch 2 Pt 3 and Ch 4 Pts 7 and 9.

125    The primary judge recognised that not all aspects of the environment were protected under the EPBC Act, constitutional authority for the Commonwealth largely shaping its form and reach. At J[157]–[159] the primary judge drew an important conclusion about the object and purpose of the EPBC Act as a whole, as follows:

[157] It is pertinent to note that whilst protection is afforded to various listed species and their habitats, neither the health, wellbeing nor survival of human beings, nor their habitats (by which I mean homes or private real property) are protected aspects of the environment directly specified by Pt 3. Actions with significant impacts on those subject matters are not prohibited subject to either the Minister’s approval or the Minister’s decision that approval is not required.

[158] That people and in particular future generations of people, should be able to enjoy the “health, diversity and productivity” of the environment is, however, a matter emphasised by the “principle of inter-generational equity” expressed in s 3A(c). Further, “protection of the environment” is an object of the EPBC Act (s 3(1)(a)) and the definition of the term “environment” given by s 528 refers specifically to “people and communities”. The term is there defined as follows:

environment includes:

(a)    ecosystems and their constituent parts, including people and communities; and

(b)     natural and physical resources; and

(c)     the qualities and characteristics of locations, places and areas; and

(d)     heritage values of places; and

(e)     the social, economic and cultural aspects of a thing mentioned in paragraph (a), (b), (c) or (d).

[159] Paragraph (a) addresses “people” directly and, by reason of para (e), the social, economic and cultural aspects of the subject matters otherwise dealt with by the definition must also be understood as dealing with the interests of people. From that definition and the statute read as a whole, the conclusion may be drawn that the object of the EPBC Act is not the protection of environment per se but the protection of the interests of human beings in the environment including, in particular, those aspects of the environment which are specified in Pt 3.

(emphasis added)

126    This conclusion as to statutory purpose of the EPBC Act as a whole was at the centre of the complaints on appeal by the Minister, and, with respect, at the centre of the erroneous approach of his Honour.

127    The primary judge then directed himself (at J[160]–[162]) to ss 18, 18A, 24D and 24E as the “controlled actions” of relevance for the approval of the extension of the mine.

128    The primary judge then referred to the assessment and approval regimes and Pts 8 and 9 in Ch 4. At J[164] and [165] the primary judge described the assessment documentation here as follows:

[164] For the application to approve the Extension Project, the “assessment documentation” referred to by s 133(1) included the NSW Department Report referred to above at [26]. That was submitted to the IPC and given to the Minister pursuant to cl 6.2 of the Bilateral Agreement, which was made between the Commonwealth and New South Wales pursuant to s 47(1) of the EPBC Act: s 133(8) and s 130(2).

[165] As required by s 47(4) of the EPBC Act, the bilateral agreement (cl 6.2(a)) provides that:

NSW will ensure there is sufficient Information in the Assessment Report on the impacts of a controlled action covered by this Agreement on each relevant Matter of [national environmental significance] so that the Commonwealth decision-maker may consider those impacts when determining whether to approve the action and, if so, on what conditions. The extent of the assessment will be proportionate to the level of likely environmental risk.

129    As to the proper construction of s 136 the primary judge referred to and relied upon what the Full Court said in Tarkine 233 FCR 254 especially at 265–266 [25]–[28] and 269 [44] and [45], as to which see [84] above.

130    Within that framework the primary judge said at J[180]:

… However, no party contended that the potential for harm to Australia’s children was not a matter that the Minister may permissibly take into account in deciding whether or not to approve an action. The applicants contended that such a matter fell within the expression “economic and social matters” in s 136(1)(b). The Minister did not contend to the contrary. That expression is unqualified and there is no basis for thinking that it was intended to be confined to those economic and social matters which are a beneficial rather than an adverse consequence of the “controlled action” subject to the Minister’s approval. The Act’s concern with adverse economic and social matters can be seen from ss 270(3)(c), 287(3)(c) and 464(3) as well as the objects in s 3 and, in particular, the object in s 3(1)(b) in light of the elaboration provided by s 3A(a).

The affirmative salient features: J[184]–[315]

Reasonable foreseeability: J[184]–[257]

131    The primary judge commenced with reasonable foreseeability of harm. At J[186], [187], [188] and [194] the primary judge recognised (correctly, with respect) that in considering the question of duty the task is a “generalised enquiry” (Wyong Shire Council v Shirt [1980] HCA 12; 146 CLR 40 at 47 (Mason J)), at a “higher level of abstraction” than when considering breach: Vairy v Wyong Shire Council 223 CLR at 446–447 [72] (Gummow J). The foreseeable real (that is not far-fetched or fanciful) risk of harm: Shirt at 48 does not need to be to the plaintiff or some particular person or persons, but “it is sufficient if the injury is to a class of persons of which the plaintiff was one might reasonably have been foreseen”: Chapman v Hearse [1961] HCA 46; 106 CLR 112 at 121, and that foreseeability is not of the particular harm in character or sequence of events but of like kind: Mount Isa Mines Ltd v Pusey [1970] HCA 60; 125 CLR 383 at 402 and Rosenberg v Percival [2001] HCA 18; 205 CLR 434 at 455 [64]. At J[191], his Honour considered that the common shared circumstances of the Children, being their age and thus temporal connection to particular harms, their location and thus geographical connection to particular harms enabled the finding that events induced by climate change expose each of the children to a real risk of harm.

132    At J[195]–[200], the primary judge examined whether the posited steps in the chain of events asserted by the respondents (applicants below) were, because of their complexity or otherwise, incapable of being seen as a real risk.

133    The chain of events relevant to the foreseeability inquiry was said by the primary judge to begin with the Minister’s conduct in granting the approval: J[195]. None of the contingencies of extraction, sale, combustion, and the occasioning of emissions were said to deny the real risk of harm. The coal will be sold and combusted. The primary judge then considered the next steps in the chain of events: emission of CO2 from the combustion of coal leading to an increase in global average surface temperature: J[196]; this increase in temperature leading to increased frequency or gravity of extreme climatic events such as heatwaves, bushfires, and so on: J[197]; and, after setting out what the Minister accepted as to the effects of global warming in Australia at J[198]–[199], the primary judge considered the exposure of the Children to the risk of personal injury as a result of climate change induced extreme climatic events: J[200].

134    At J[201]–[246], the primary judge considered the risks of harm identified and relied upon by the respondents (applicants below). Leaving to one side economic harm which his Honour elsewhere rejected, the primary judge examined personal harm being physical and mental harm by “direct impacts”, “indirect impacts” and “flow-on impacts”. The primary judge considered that the evidence did not support the reasonable foreseeability of harm in the latter two categories: J[204] and J[237]–[246].

135    At J[205]–[225], the primary judge examined the evidence on heatwaves, at J[226]–[235] on bushfires and at J[236] on other “direct impacts”.

136    At J[211], the primary judge set out the findings of Dr Mallon (whose expertise was in physics, engineering, energy and emissions modelling and climate change analysis) in terms of the Children’s likely future experience of heatwaves, and the consequences arising therefrom. Throughout Australia, Dr Mallon’s evidence was that heat stress related presentations to doctors, hospitals and paramedics will increase by 850%: J[211]. The burden of such will largely fall on the Children, who will be vulnerable to heat stress related presentations by that time, as they pass 75 years of age: J[213]. At J[218] the primary judge considered the evidence of Dr Meyricke (an actuary specialising in climate change mitigation and mortality risk) that, by between 2060 and 2080, excess mortality from heat is projected to be 12% in persons aged over 65 years. At J[222] his Honour referred to evidence projecting days of 55℃ in summer and 33℃ in winter in Victoria, and the conclusions of Professor Capon that heatwaves are the most deadly hazard in Australia: J[223]. The primary judge accepted this evidence as reliable, and concluded that it established a sufficient link between the age characteristic of the Children and the exposure (of each of the Children) to the harm in question: J[225]. No challenge to these findings was made on appeal.

137    At J[226]–[235], the primary judge set out the risk of harm posed to the Children by bushfires. The evidence surveyed by his Honour disclosed a 60% increase in the projected number of high fire danger days in the 2050s as compared to 1986–2005 (J[226]), a 77% likelihood of severe bushfires in any given year (J[230]), and extrapolating from the sequelae of the 2019–2020 Australian bushfire season (known as the Black Summer bushfires), predictions with respect to significant smoke-inhalation related mortality, hospitalisations for cardiovascular and respiratory problems, and emergency department presentations for asthma (J[231]–[232]). These sequelae are not confined to Australia, with the spread of smoke from a catastrophic bushfire likely to extend throughout the world: J[233]. The primary judge concluded, on the basis of this evidence, that some of the Children, perhaps many hundreds or thousands of them, will be killed or injured by future climate change induced bushfires: J[234]. His Honour linked the exposure to this harm to both the age characteristic and geographical characteristic of the children: J[235]. No challenge to these findings was made on appeal.

138    At J[236], his Honour dealt with other direct impacts, being inland and coastal flooding and cyclones, as follows:

Other climatic events relied upon by the applicants were inland and coastal flooding and cyclones. The evidence about those events was scant and, insofar as it existed, was pitched at a high level of generality which, in terms of identifying risks to health, was not sufficiently directed to the Children to enable a conclusion that each child is exposed to a real risk of injury or that some approximate number of them are at risk. The evidence mainly relied upon by the applicants was that given by Professor Capon. But that evidence barely touched upon the field in question. Insofar as it was given, it was given in the context of a global report of the IPCC which specifically acknowledged local geographical variations and in any event the only concrete example provided was sea level rises which (without specification) were said to “[threaten] population health”. Professor Steffen briefly mentioned cyclones suggesting the likelihood of fewer but more extreme tropical cyclones for Australia. There is some evidence in the report of Dr Mallon about inland and coastal flooding and cyclones but only in the context of his consideration of future damage to property. However, in the “undemanding” assessment here being undertaken it may be assumed that a reasonable person would have a general appreciation that, because of their hazardous nature, severe flooding and extreme cyclones pose a risk of harm to humans and, if those events are more frequent or more severe, that risk is increased. On that basis it may be said that some harm to some of the Children is reasonably foreseeable.

139    At J[247]–[257], the primary judge set out his conclusions on reasonable foreseeability. Having concluded that a reasonable person in the Minister’s position would foresee that, by reason of the effect of increased CO2 in the Earth’s atmosphere and the consequential increase in global average temperature, each of the Children is exposed through the occurrence of heatwaves or bushfires, to the risk of death or personal injury, his Honour turned to the question of whether that personal injury or death was a foreseeable consequence of the approval of the extension of the mine: J[247]. The primary judge expressed the question as follows at J[247]:

… Accordingly, I need to be satisfied that a reasonable person in the Minister’s position would foresee that a risk of injury to the Children would flow from the contribution to increased atmospheric CO2 and consequent increased global average surface temperature brought about by the combustion of the coal which the Minister’s approval would facilitate.

140    The primary judge concluded at J[248] that the proposition that combustion of 33 Mt of coal would contribute to an increase of atmospheric CO2 was both obvious and foreseeable. That that increase in CO2 from the combustion of coal from the extension of the mine would increase global average surface temperatures may not be, his Honour concluded, immediately obvious to the reasonable lay person, but the Minister should be taken to be aware of the unchallenged evidence of Professor Steffen that CO2 emissions caused by the extension of the mine would increase global average surface temperature and the level at which that temperature would eventually stabilise: J[248].

141    At J[249], the primary judge referred to the discussion at J[74]–[90] (see [112]–[120] above) as to risk of harm and stated that that demonstrated why he considered:

that the emission of 100 Mt of CO2 from the Extension Project increases the risk of the Children being exposed to harm and particularly so in the realm of the risk profile which plausibly arises should the ‘tipping cascade’ be triggered and engage a 4 Future World trajectory.

142    At J[250], his Honour made reference to the view of the IPC and the NSW Department that Scope 3 emissions from the extension of the mine would be a significant contributor to anthropological (sic: anthropogenic) climate change.

143    At J[251] and [252], the primary judge repeated aspects of the underlying risks necessary for the imposition of a duty of care: that the foreseeable risk need only be real, it may be remote or unlikely, as long as it is not far-fetched or fanciful: Shirt 146 CLR at 46 and 47; Overseas Tankship (UK) Ltd v The Miller Steamship Co (The Wagon Mound (No 2)) [1967] AC 617 at 642–643; Rosenberg v Percival 205 CLR at 455 [64], and that foreseeability is not, in itself, a test of causation: Chapman v Hearse.

144    In light of this, his Honour concluded at J[253]:

I accept that, even on the marginal risk assessment referred to at [84], the prospective contribution to the risk of exposure to harm made by the approval of the extraction of coal from the Extension Project may be characterised as small. It may fairly be described as tiny. However, in the context of there being a real risk that even an infinitesimal increase in global average surface temperature may trigger a 4 C Future World, the Minister’s prospective contribution is not so insignificant as to deny a real risk of harm to the Children. The risk of harm in question is reasonably foreseeable even without regard to the unparalleled severity of the consequences of that risk crystallising. But the magnitude of the danger to which the Minister’s conduct is likely to contribute must also be taken into account. When that is done, the conclusion that, by reference to “contemporary social conditions and community standards” a reasonable person in the Minister’s position would foresee the risk and take reasonable and available steps to eliminate it, is established.

This conclusion was, for the primary judge, only bolstered by the inclusion of the precautionary principle in s 3A of the EPBC Act (J[254]) which his Honour found “attunes both the foresight and response required of a reasonable person in the Minister’s position to the risks that the plausible scientific evidence confirms will be faced by the Children.”: J[256].

145    At J[257], the primary judge summed up foreseeability and duty as follows:

In sum, this is a case where the foreseeability of the probability of harm from the defendant’s conduct may be small, but where the foreseeable harm, should the risk of harm crystallise, is catastrophic. The consequent harm is so immense that it powerfully supports the conclusion that the Children should be regarded as persons who are “so closely and directly affected” that the Minister “ought reasonably to have them in contemplation as being so affected when…directing [her] mind to the acts…which are called in question”: Donoghue v Stevenson at 580 (Lord Atkin). Accordingly, ‘reasonable foreseeability’ is a strong salient feature in support of the posited duty of care being recognised by the law of negligence.

146    As will be discussed in due course, influential in the primary judge’s conclusions as to the imposition of a duty of care was this view of the first salient feature of risk of harm and reasonable foreseeability of harm. It dominated his Honour’s approach.

Control, responsibility and knowledge: J[258]–[288]

147    After surveying the relevant authorities at J[259]–[270], the primary judge concluded at J[271]:

The risk of harm to the Children is not remote, it is reasonably foreseeable and it is therefore a real risk for reasons already explained. The Minister has direct control over the foreseeable risk because it is her exercise of power upon which the creation of that risk depends. To my mind, there is therefore a direct relation between the exercise of the Minister’s power and the risk of harm to the Children resulting from the exercise of that power. The entirety of the risk of harm flowing from that exercise of power is therefore in the Minister’s control.

148    Furthermore, the primary judge considered the functions conferred by the EPBC Act to be of relevance, insofar as the Minister occupies a position of responsibility over the environment, and the interests of current and future generations of Australians therein: J[273]–[274].

149    The primary judge rejected the submission that, in order for ‘control’ in the relevant sense to be established, it must be exercised at each point in the causal chain (J[278]–[280]), concluding that the Minister had “very substantial, if not exclusive, control” over the risk that would flow from the approval of the extension of the mine: J[284]. His Honour further found that the Minister had knowledge of the risk of harm to the Children (J[286]), which served to support the proposition that the Children were so closely and directly affected by the approval that the Minister ought to have had them in her contemplation: J[287].

Vulnerability, reliance and recognised relationships: J[289]–[315]

150    The primary judge held the evidence to demonstrate that the Children were extremely vulnerable to a real risk of harm wrought by climate change: J[289]. His Honour’s conclusions on this point were set out as follows at J[293]–[294]:

[293] It is difficult to characterise in a single phrase the devastation that the plausible evidence presented in this proceeding forecasts for the Children. As Australian adults know their country, Australia will be lost and the World as we know it gone as well. The physical environment will be harsher, far more extreme and devastatingly brutal when angry. As for the human experience – quality of life, opportunities to partake in nature’s treasures, the capacity to grow and prosper – all will be greatly diminished. Lives will be cut short. Trauma will be far more common and good health harder to hold and maintain. None of this will be the fault of nature itself. It will largely be inflicted by the inaction of this generation of adults, in what might fairly be described as the greatest inter-generational injustice ever inflicted by one generation of humans upon the next.

[294] To say that the Children are vulnerable is to understate their predicament. However, it is not vulnerability in the abstract which is relevant for determining whether a duty of care is owed to them by the Minister. Their vulnerability must be connected to their relation with the Minister or their reliance upon the Minister: Stuart at [134] (Crennan and Kiefel JJ). And it is.

151    The primary judge said at J[295] that the source of the Children’s exposure to risk includes the impugned conduct of the Minister (J[295]), and the relevant vulnerability was said to be partly a function of the magnitude of the potential risk, and partly a function of the Children’s powerlessness to avoid that harm: J[296]. Vulnerability, the primary judge held, is not diminished by virtue of the fact that others are equally, or perhaps more, vulnerable: J[297]. The general nature of the Minister’s responsibility for Australians, derived from the EPBC Act was relevant to reliance: J[298].

152    The primary judge concluded that the Children were reliant upon the Minister by virtue of the control she exercised over the potential harm and the responsibility she had under the EPBC Act for the health of the environment and for those persons whose safety may be compromised by conduct which endangers it: J[299]. While the Minister’s contribution to the risk of harm was minute, the primary judge held that the magnitude of harm was sufficient to render the risks flowing from the conduct to be significant, and to therefore establish reliance in the relevant sense: J[300].

153    Finally, at J[301]–[311], the primary judge invoked the parens patriae jurisdiction, and the fact that the Children were all minors, to demonstrate the inherent vulnerability of the Children, and the protective aspect inherent in the relationship between the Minister in her capacity as a member of the Executive, and the Children: J[301]–[302], [311]. The primary judge also placed emphasis on the ‘innocence’ of the children, in the sense that they bear no responsibility for climate change, but will bear the burden of its consequences: J[312].

The negative salient features: J[316]–[489]

Coherence of the duty with the statutory scheme and administrative law: J[316]–[427]

154    It was at this point that the primary judge turned to a feature of fundamental importance: that of (in)coherence. His Honour’s explanation of the elusive concept of coherence was that of “a policy consideration … deployed by the common law to assist in the development and application of the common law, primarily in respect of its interaction with statute law, but also internally as between different principles of the common law” (J[321]) which requires that (in this case) statute and law cohere, with “one sitting compatibly alongside the other without ‘incongruity’ or ‘contrariety’” (J[322]).

155    The Minister’s argument at first instance with respect to incoherence was that the posited duty of care would foreclose the statutory discretion vested in the Minister by ss 130 and 133 of the EPBC Act, insofar as the duty would dictate a particular outcome and the elevation of the avoidance of harm to a mandatory and paramount consideration: J[329] and [340]. His Honour considered, following a survey of the relevant authorities, that the search for incoherence and inconsistency (J[356]) ought not be overly compartmentalised, that it may be revealed by the nature of the statutory power, or the process required by the performance of the statutory task, but that in each case, statutory purpose will be relevant: J[356]. The primary judge noted that constraints or impairments imposed by a duty upon the exercise of a discretionary power are not sufficient to preclude the finding of the existence of that duty: J[357].

156    Engaging closely with Crimmins 200 CLR 1 the primary judge concluded that the capacity to exercise validly a statutory power is not necessarily in tension with a co-extensive duty of care which may cut across the exercise of that power: J[369]–[380], especially [373]. That is, his Honour found that even where the valid exercise of a statutory discretion is impaired by the imposition of liability in negligence, so long as that duty is consonant with the statutory purposes, there will be no incoherence: J[376], [394]. Surveying such authorities as Crimmins, Brodie v Singleton Shire Council [2001] HCA 29; 206 CLR 512, and Alec Finlayson Pty Ltd v Armidale City Council (1994) 51 FCR 378, his Honour concluded at J[395]:

In reconciling the authorities, what must be recognised is that coherence-based reasoning places great importance on statutory purpose cohering with the imposition of liability in negligence. Consistency between statutory purpose and the duty of care imposed by the law of negligence is apt to be regarded as a potent consideration favouring a conclusion of coherence. An interference or impairment of a statutory discretion conferred by the statute has negative implications for coherence. However, both considerations must be weighed. As statutory discretion is subordinate to statutory purpose because a discretion is to be exercised “only in accordance with the objects and policy of the Act” (Walton v Gardiner (1993) 177 CLR 378 at 409 (Brennan J)), consistency with purpose will be the paramount consideration.

His Honour went on to remark that in MM Constructions (Aust) Pty Ltd v Port Stephens Council (No 7) [2012] NSWCA 417, X v State of South Australia (No 3) [2007] SASC 125; 97 SASR 180 and Sullivan v Moody 207 CLR 562 there was “no consistency or coherence with statutory purpose capable of negating the inconsistency with the discretionary function”: J[396].

157    The primary judge concluded that the posited duty of care, which is concerned with the avoidance of personal injury to, and death of, the Children, is consonant with the purpose of the EPBC Act, and is a relevant consideration which must be taken into account by the Minister when exercising her power under ss 130 and 133 of that Act: J[397]. This, it was said, was because an expectation that a “statutory power will not be used without care being taken to avoid killing or injuring persons will almost always cut across the exercise … [of] a broad discretionary power”: J[398]. This, the primary judge said, was likely sufficient to negate concerns of incoherence. His Honour found (without any reference to authority) that unless the legislation clearly identifies considerations which take priority over human safety, Parliament may be taken to have intended that priority be given to such safety (J[399]). Sullivan v Moody and X v South Australia are examples of an exception to this: where various fundamental liberties and rights collide: J[400]–[401].

158    Extrapolating out from ss 193(1), 212, 236 and 255 of the EPBC Act (which each make reference to the protection of human health and human life), the primary judge concluded that human health and safety is a primary concern of the legislation: J[403]. His Honour went on to find, considering the subject matter, purpose and scope of the EPBC Act, that human safety is a relevant mandatory consideration in relation to a controlled action which may endanger that safety: J[404]. On this basis, the primary judge held the posited duty to be in harmony with the statutory scheme and its overarching objectives: J[408]. His Honour also referred to McHugh J’s observations in Crimmins 200 CLR 1 at 50–51 [132] that the imposition of liability in negligence, where statutory purpose and the objective of a duty of care are aligned, would likely result in increased vigilance as opposed to defensiveness. At J[416], the primary judge declined to extend this conclusion to that aspect of the posited duty of care pertaining to property damage or economic loss.

159    As to whether the imposition of the duty of care would have the consequence of dictating the decision to be made by the Minister, his Honour considered at J[411] that it would not, and that ultimately, liability in negligence, assessed at the level of breach, includes considerations as to the reasonableness of a defendant’s response.

160    As to inconsistency with administrative law, and the susceptibility of the duty of care analysis to seep into the realm of merits review, the primary judge said the following at J[426]:

… The subject of the posited duty is not the validity of any decision made or to be made by the Minister under the EPBC Act. The posited duty, whether assessed at the level of duty or at the level of breach, is not that the Minister must exercise reasonable care not to make a flawed decision either generally or by reference to any particular instance of flawed decision-making. The subject of the posited duty is not, in either form or substance, legally invalid decision-making. No part of the applicants’ case in negligence, neither in their assertion of a duty nor in their assertion of a prospective breach, relies upon a contention that any decision taken or to be taken, or any step taken or to be taken, in the process of decision-making, is or will be legally invalid. Their action in negligence is “not brought in addition to or in substitution for any public law remedy”: Pyrenees Shire Council at [172] (Gummow J).

Indeterminacy: J[428]–[473]

161    With respect to the Minister’s charge that the posited duty was plagued by indeterminacy, the primary judge adverted (at J[430]) to the nature of the harm (being physical or psychiatric injury or death, as opposed to economic harm) as itself providing a limiting factor, and control mechanisms (such as causation) operating to mitigate indeterminacy: J[433]–[434]. At J[436] the primary judge reiterated the observations made in Perre 198 CLR at 233 [139], and Cattanach v Melchior [2003] HCA 38; 215 CLR 1 at 20 [32], that the size of the potential class of claimants is irrelevant to the question of indeterminate liability. At J[443] his Honour concluded that any indeterminacy was not sufficient to deny the existence of the posited duty. In response to the Minister’s contention that the duty would not be confined to Australians, the primary judge held that the affirmative salient features rely, at least in part, upon the Minister’s particular responsibility for Australians under the EPBC Act: J[455]. In response to the contention that the claimant class would extend to adults, the primary judge accepted that confining the duty to the Children was arbitrary (J[459]) but that, by virtue of these proceedings, the Minister had the benefit of identifying the potential claimants and the nature of their likely claims (limited to personal injury and death): J[465], [469]. The primary judge concluded that, together with the work done already by the controlling mechanisms of foreseeability and coherence, and the work to be done by others (such as causation), a reasonable person in the Minister’s position would be sufficiently informed with respect to her potential liability, both in terms of the likely number of potential claimants and the likely nature of their claims: J[470]. His Honour further held that the fact that others around the world share responsibility for the projected harm to the Children greatly diminishes the charge of indeterminacy: that is, fractional responsibility for the harm may result in fractional liability therefor J[471].

Other control mechanisms: J[474]–[489]

162    As the last considerations before expressing his conclusions as to duty of care at J[490]–[491], and under the (almost miscellaneous) heading of “other control mechanisms”, the primary judge turned to questions of policy and the question whether the matters raised by the posited duty were suitable for the judicial function. With respect, this somewhat cursory and after-the-fact consideration illuminates the fundamental problem with his Honour’s approach to which I will come in due course: the use of conventional salient features as if this was an ordinary tort case, and the failure to address at the outset the question of the relationship between the Minister and the Children and the scope and content of the duty to be imposed and the type of considerations necessarily thrown up by it at the point of breach, especially the complex social, economic and political considerations and choices as well as scientific considerations reflected in the uncontested evidence of Professor Steffen that face all governments in countries around the world as to the proper or appropriate response to the risks of global warming and potential climate catastrophe, for the world and mankind.

163    The primary judge framed the question at J[474], as follows:

It is sometimes said that a duty of care cannot be imposed where it would cut across a “policy decision” or, in other words, that no duty of care should be owed in respect of the exercise of a power by a statutory authority involving public policy: Dansar at [68]-[69] (Macfarlan JA). It is to that question that I now turn.

164    To express the matter thus was, with respect, hardly to do justice to the fundamental centrality of the scope and content of the duty and what it threw up for consideration at the point of breach.

165    At J[475], the primary judge referred to the value-laden exercise in the Minister’s broad evaluative judgment given to her. This expression of the matter focused on the decision under ss 130 and 133 and what might be called the usual considerations in such a decision. That of course was relevant. In relation to this, the primary judge commented that the policy/operational dichotomy was of dubious validity, citing Gummow J in Pyrenees Shire Council v Day [1998] HCA 3; 192 CLR 330 at 393–394 [182] and Vairy 233 CLR at [86]. Earlier at J[387] the primary judge had said that the “policy/operational dichotomy” was “largely … discredited”. At J[387] the primary judge aligned the approval of a development application in Alec Finlayson 51 FCR 378 with the relevant statutory decision under ss 130 and 133 of the EPBC Act.

166    At J[476], the primary judge referred to considerations of legislative or quasi-legislative domain and “core policy”, saying:

In a representative democracy some decisional fields are necessarily the exclusive domain of the legislature. Legislative and quasi-legislative decisions fall into that category. As to quasi-legislative decisions, the abundant authorities are clear that those decisions do not attract a duty of care: Heyman at 469 (Mason J); Pyrenees Shire Council at [182] (Gummow J); Crimmins at [32] (Gaudron J), at [87] and [93] (McHugh J), at [170] (Gummow J), and at [292] (Hayne J); Graham Barclay Oysters at [14] (Gleeson CJ); Vairy at [81] and [85]-[86] (Gummow J). “Core policy-making functions” also find support as a further exception: Crimmins at [87] and [93] (McHugh J). There are, however, many examples of a duty of care being recognised in relation to a statutory approval process. Alec Finlayson is an example. Further examples are recorded at [391] above. It has not been suggested that statutory decision-making of that kind is a “core policy-making function”.

167    At J [477]–[478] importantly, the primary judge said the following about how the case was run:

[477] The Minister did not contend that her statutory task was quasi-legislative in character or a core policy-making function. Her submission relied on the “policy/operational” dichotomy but was primarily based on the inappropriateness of common law intervention into the policy-based statutory task the Minister asserted she must perform.

[478] In that respect the Minister said that her statutory duty was political or policy-based because it required choices to be made or value-laden political judgments to be made about matters of importance. For the reasons already given, the characterisation of the task as political and value-laden is not helpful in and of itself. However, the fundamental point made by the Minister was that her statutory task was steeped in policy considerations appropriately dealt with by her without intervention by the common law. In that respect the Minister contended that how to manage the competing demands of society, the economy and the environment over the short, medium and long term, is a multifaceted political challenge. In the context of climate change, measures to manage those competing demands occur within the context of evolving national and international strategies. It was said that reducing greenhouse gas emissions while simultaneously managing the demands of society and the economy is a complex and nuanced task. The Minister contended that the imposition of a common law duty of care that, by contrast, would render tortious all activities that involve generating (or allowing someone else to generate) material quantities of greenhouse gases is a blunt and inappropriate response.

168    On appeal the Minister submitted that J[477] contained a misstatement as to what was submitted to his Honour about “core policy”. The Minister submitted that it was contended before his Honour that the decision involved core policy. It may be accepted that the usual type of decision of the Minister under ss 130 and 133 would not involve core policy. But the question is and was whether the imposition of the duty would require the engagement of core policy in complying with the duty to be assessed at the point of breach. Clearly it was submitted below that that would be the case. So much is made clear by J[478].

169    At J[479], the primary judge sought to summarise the matter thus:

That contention essentially argued that the Minister is better placed to deal with the complex task of addressing climate change than the common law. The correctness of the proposition, at least in a general sense, cannot be doubted.

170    Further at J[479], the primary judge stated that the proposition contained two false premises.

First, that the imposition of a common law duty of care would be addressing the problem of climate change and thus interfere with the statutory task given to the Minister. Second, that the intervention of the common law here, would render tortious all or a multitude of activities that involve the generation of greenhouse gases.

171    At J[481]–[485], the primary judge dealt with the asserted first false premise as follows:

[481] As to the first premise, the posited duty of care will not and cannot address climate change. All that it can and will do is impose an obligation on the Minister when deciding whether or not to approve the Extension Project to take reasonable care to avoid personal injury to the Children. The imposition of a duty of care does not mandate the Minister’s decision. As already discussed, the EPBC Act itself imposes an obligation upon the Minister to take into account the personal safety of the Children.

[482] The imposition of liability for the breach of a duty of care arising from careless conduct causing personal injury is at the heart of the common law’s place in the legal system. That of itself cannot be an inappropriate intervention upon a statutory field whilst the Speirs line of authority remains good law and, in relation to the Executive, at least where “a particular exercise of power has increased the risk of harm to an individual”: Graham Barclay Oysters at [91] (McHugh J). The possibility of such an intervention being inappropriate because of its distortive impact upon the statutory task is addressed by the requirement of coherence. That is the work done by that principle. All of the potential inappropriate impacts upon the Minister’s statutory task have already been addressed and negated.

[483] The question then is what remains to sustain the idea that the imposition of a duty of care in this case would be an inappropriate intervention by the common law. The elephant in the room may well be that the Minister’s statutory task falls within the realm of a contested political issue as to, first, whether climate change is real and, secondly, if so, whose interests should take priority in addressing it.

[484] Quite correctly, the Minister did not draw my attention to that controversy. Courts are regularly required to deal with legal issues raised in the milieu of political controversy. A political controversy can never provide a principled basis for a Court declining access to justice.

[485] The Minister’s appeal to there being policy choices at play echoed that made in Brodie to which Gaudron, McHugh and Gummow JJ gave the following response at [106]:

Appeals also were made to preserve the “political choice” in matters involving shifts in “resource allocation”. However, citizens, corporations, governments and public authorities generally are obliged to order their affairs so as to meet the requirements of the rule of law in Australian civil society. Thus, it is no answer to a claim in tort against the Commonwealth under s 75(iii) of the Constitution that its wrongful acts or omissions were the product of a “policy decision” taken by the Executive Government; still less that the action is “non-justiciable” because a verdict against the Commonwealth will be adverse to that “policy decision”.

172    For the reasons set out below, this approach was inadequate to deal with the question of policy and the unsuitability of the issue for determination by the Judicial branch and the question of core policy that the Minister’s submission was addressing before the primary judge and which is addressed before us.

173    At J[489], after dealing with the second false premise, the primary judge said that he was not persuaded that the recognition of the posited duty should be declined for policy reasons.

Conclusions on duty of care: J[490]–[491]

174    The primary judge set out his summary conclusions at J[490]–[491]:

[490] ‘Coherence’, ‘control’, ‘vulnerability’ and ‘reliance’ all assume especial relevance in an assessment of whether a novel duty of care should be recognised (see [109] above). On the present facts, I regard ‘coherence’ as agnostic, but even if it is to be treated as tending against the recognition of a duty of care, ‘control’, ‘vulnerability’ and ‘reliance’ are affirmative of a duty being recognised and significantly so. ‘Indeterminacy’ and the policy considerations dealt with under the heading “Other Control Mechanisms” are also largely agnostic but if they tend in any direction it may be said that they tend against a duty being recognised. ‘Reasonable foreseeability’ strongly favours the recognition of duty of care. In totality, in my view, the relations between the Minister and the Children answer the criterion for intervention by the law of negligence.

[491] That conclusion is confirmed when re-examined through the lens of the neighbourhood principle and the criteria of reasonableness fundamental to the law of negligence. By reference to contemporary social conditions and community standards, a reasonable Minister for the Environment ought to have the Children in contemplation when facilitating the emission of 100 Mt of CO2 into the Earth’s atmosphere. It follows that the applicants have established that the Minister has a duty to take reasonable care to avoid causing personal injury to the Children when deciding, under s 130 and s 133 of the EPBC Act, to approve or not approve the Extension Project.

Whether an injunction should be issued: J[492]–[512]

175    The primary judge then considered whether an injunction ought to be issued, which his Honour answered in the negative: J[512]. Not being the subject of submissions in the appeal, it is unnecessary to address this aspect of his Honour’s reasons.

Conclusion and further steps: J[513]–[521]

176    Having concluded that the posited duty of care was established by the applicants, the primary judge requested that the parties file further submissions on the nature of the relief sought, and, in particular, on the form of the declaration.

The Minister’s grounds of appeal and submissions

177    The Minister’s appeal comprised five related grounds of appeal. As expressed in the notice of appeal the grounds commenced, in ground 1, with a conclusory ground that the primary judge erred in finding that a legal duty of care was owed and concluded, in ground 5, with five factual complaints as to the findings of the primary judge. Grounds 2–4 contain a number of specific complaints. The grounds of appeal were in the following terms:

1. The primary judge erred in finding that the Minister owed a duty to take reasonable care, in the exercise of her powers under ss 130 and 133 of the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act) in respect of referral EPBC No. 2016/7649 (Extension Project), to avoid causing personal injury or death to persons who were under 18 years of age and ordinarily resident in Australia at the time of the commencement of the proceeding (Australian children), arising from emissions of carbon dioxide into the Earth’s atmosphere.

2. Without limiting ground 1, the primary judge erred in law:

(a) in finding that, in exercising her power of approval under ss 130 and 133 of the EPBC Act, the Minister must take into account as a mandatory consideration any effect of a decision on human safety;

(b) in finding that, as a result of that mandatory consideration, the recognition of a novel duty of care would be in harmony with the EPBC Act, because that duty of care required the Minister to consider and give elevated weight to a matter that the EPBC Act already required her to consider;

(c) in failing to find that the novel duty of care alleged should not be recognised because it was incoherent with the EPBC Act, including because it would distort the capacity of the Minister, when exercising her power of approval under ss 130 and 133 in an area of highly contested public policy, to balance competing considerations and interests in the manner contemplated by the EPBC Act.

3. Without limiting ground 1, the primary judge erred in law:

(a) in failing to identify the causal requirement inherent in determining whether a risk of harm is reasonably foreseeable in the sense relevant to the recognition of a duty of care, or in treating that causal requirement as capable of being satisfied even if the contribution to the risk of the harm occurring that would result from a failure of the Minister to exercise reasonable care is assessed as “tiny” (PJ [253]);

(b) in finding that, despite the fact that the primary judge was “unable to say that the evidence itself demonstrates the extent, if any, that a fractional increase in average global temperature of the kind in question poses an additional risk of harm” (PJ [83]), a reasonable person in the Minister’s position would nevertheless foresee that a decision under the EPBC Act to approve the Extension Project would materially increase the risk:

(i) that the future world would shift from a stabilised global average surface temperature of 2°C above pre-industrial levels to a point 4°C above pre-industrial levels; or

(ii) that all Australian children would be exposed to a real risk of death or personal injury from heatwaves and bushfires induced by climate change resulting from increased CO2 in the Earth’s atmosphere;

(c) in finding that the Minister had substantial control, in the sense relevant to the consideration of salient features bearing on the recognition of a novel duty of care, over the risk of emissions of carbon dioxide into the Earth’s atmosphere causing personal injury or death to all Australian children;

(d) in finding that the Minister is in a protective relationship with the Australian children, in the exercise of executive power and founded upon the capacity of the government to protect and upon the special vulnerability of children, and in treating such a relationship as supporting the recognition of a novel duty of care;

(e) in finding that the indeterminacy of the posited novel duty of care did not tend against its recognition.

4. Without limiting ground 1, in circumstances where the identified risk of harm to Australian children arises from the emission of carbon dioxide into the Earth’s atmosphere, the primary judge erred in law:

(a) in treating the position of the Minister in exercising a power of approval under ss 130 and 133 of the EPBC Act as if it was equivalent to that of a person who chooses to engage in activities that emit carbon dioxide into the Earth’s atmosphere (PJ [79]);

(b) as a consequence, in imposing a novel duty of care the effect of which is to require the Minister to take reasonable care to prevent harm that may be caused or contributed to by the voluntary actions of other persons, in disregard to the general rule that a person is under no duty to prevent another person from doing damage to a third person.

5. Without limiting ground 1, the primary judge erred in fact in finding that:

(a) the best available outcome that climate change mitigation measures can now achieve is a stabilised global average surface temperature of 2°C above pre-industrial levels (PJ [31] and [74(ii)]);

(b) at a stabilised global average surface temperature above 2°C, there is an exponentially increasing risk of the Earth being propelled into an irreversible 4°C trajectory (PJ [31], [74(iii)] and [75]);

(c) there is a real risk that even an infinitesimal increase in global average surface temperature above 2°C above pre-industrial levels may trigger a 4oC Future World (PJ [253]);

(d) a decision under the EPBC Act to approve the Extension Project would cause an increase in CO2 emissions of 100Mt above the CO2 emissions that would otherwise occur (PJ [79], [84], [247] – [249]);

(e) if the Extension Project were to proceed, any CO2 emissions resulting from burning of coal extracted through that project would be outside the emissions contemplated by the “carbon budget” necessary to achieve a target of 2°C above pre-industrial levels (PJ [86] – [87], cf [73]).

178    The grounds are better understood and better addressed by appreciating how the appeal was put in written submissions and in oral address because of the necessarily distinct, but interrelated character of the errors contended for.

Policy

179    Broadly put the Minister contended, under the rubric of ground 1, that she was not under a duty of care in making the decision because the decision was one involving core policy considerations of financial, economic and social or political factors or constraints within a constitutional setting of a responsible Minister dealing with policy questions of the highest order, and as such the question whether it was made negligently is one that is unsuitable for judicial resolution. The primary judge erred, it was submitted, in concluding that the distinction between policy and operations was of no utility and in concluding that the decision did not involve the relevant making of policy.

Incoherence

180    Secondly, the Minister submitted that the duty was not coherent with the statutory scheme under the EPBC Act for the making of decisions and in particular this decision. Two crucial errors contributed, it was submitted, to the primary judge’s finding that the duty was conformable with the statute: first, the finding that human safety was a mandatory consideration in the making of the decision; and, secondly, by focusing on the statutory purpose of the EPBC Act as a whole and concluding that it conformed with the duty rather than by playing close attention to the nature and function of the approval power in question.

Lack of relevant legal foreseeability

181    Thirdly, the Minister submitted that the primary judge erred in concluding that in the relevant sense injury to the respondents was reasonably foreseeable. The primary judge failed, it was submitted, to have proper regard to the nature of the causal element bound up in foreseeability: that a reasonable person in the Minister’s position would foresee that carelessness on her part may be likely to cause harm to the plaintiff. It is the foreseeability of causing the harm, not making a “tiny” contribution to the condition that together with millions of other actors around the world might cause harm. The approach of the primary judge, it was submitted, was implicitly to apply a causation approach based on increasing the risk of harm (illustrated by the United Kingdom approach to causation in asbestos cases: Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22; [2003] 1 AC 32 and Sienkiewicz v Greif (UK) Ltd (Willmoe v Knowsley Metropolitan Borough Council) [2011] UKSC 10; 2 AC 229, which is not accepted in Australia), as opposed to being a material cause of the harm (illustrated by Bonnington Castings Ltd v Wardlaw [1956] AC 613, which is accepted in Australia).

Lack of relevant control

182    Fourthly, the Minister submitted that the primary judge erred, in his application of the multi-factorial salient features approach to the recognition of a duty of care in novel factual circumstances, in finding that the Minister had relevant control over the real risk of harm. It could be accepted, as the Minister did in oral argument, that she had control over such contribution as the approval and the extension of the mine made to overall greenhouse gas emissions, but that was not the correct focus or question. The relevant control was as to the harm or risk of harm: over the risk of personal injury or death from climate change events, such as heatwaves or bushfires. This lack of control in this relevant sense was said to be demonstrated by the primary judge’s finding of fact that the emissions caused by the extension were a “tiny” contribution to the identified risk of harm.

Lack of relevant vulnerability

183    Fifthly, the Minister submitted that the primary judge erred in concluding that the respondents were vulnerable, in the relevant sense. The vulnerability found by his Honour should be understood, it was submitted, to be the increased likelihood that the respondents would suffer harm. This is not, it was submitted, the vulnerability of a special kind said to be intended by cases such as Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; 179 CLR 520 at 551, and Crimmins 200 CLR 1 at 38–39 [91] and 43–44 [108]. The respondents will be, it was submitted, in the same position as everyone else alive in the second half of this century. This lack of vulnerability is reinforced, it was submitted, by the lack of control of the Minister over the harm. Nor is it legitimate, it was submitted, to call in aid the parens patriae jurisdiction to bolster a non-existent relationship of special liability. Nor is the so-called “innocence” of the respondents of relevance.

184    The overlapping nature and the interrelatedness of the salient features is illustrated at this point of the argument. The statute gives no basis, it was submitted, for the conclusion that there is a relationship based on vulnerability of the respondents or some protective relationship between the Minister and the respondents. The Minister only had power over the extension of the mine and thus some causative influence over the release of greenhouse gases because of environmental risks to threatened species or communities and a water resource unrelated to climate change and its risks to the respondents and the community as a whole.

Indeterminacy

185    Sixthly, the Minister submitted that the primary judge erred in failing to find that the potential width of the Minister’s liability was indeterminate. Not only was the class of potential claimants vast, but the variety of risks was highly variable. Again, the interconnectedness of the salient features was demonstrated by the Minister’s submission that the indeterminacy of liability was further magnified by the minimal contribution to the risk of harm of the Minister’s action and the magnitude of liability depending upon the actions of vast numbers of other global actors. The indeterminacy was reinforced upon recognition, it was submitted, that there was no principled basis to confine the duty to the respondents as children living presently, as opposed to all people who will be exposed to the hazards of climate change and global warming in the second half of the century.

186    The Minister submitted that the primary judge’s rejection of indeterminacy as relevant to a duty of care concerned with personal injury should be seen as wrong in circumstances where one does not have a finite set of physical consequences of an act or omission. Here, it was submitted, there is no meaningful limit on how many respondents (as a large class of people living in Australia) will suffer harm, when the harm will occur, and what the extent of it will be. Notions of proportionality or disproportionality intrude here. It was submitted that the duty would place an undue burden of responsibility on the Minister, well beyond her contribution by this decision to greenhouse gas emissions and the overall risk of harm.

Factual errors

187    Five factual errors were asserted as follows:

(a) the best available outcome that climate change mitigation measures can now achieve is a stabilised global average surface temperature of about 2℃ above pre-industrial levels (PJ [31] and [74(i)]);

(b) at a stabilised global average surface temperature above 2℃, there is an exponentially increasing risk of the Earth being propelled into an irreversible 4℃ trajectory (PJ [31], [74(iii)] and [75]);

(c) there is a real risk that even an infinitesimal increase in global average surface temperature above 2℃ above pre-industrial levels may trigger a 4℃ Future World (PJ [253]);

(d) a decision under the EPBC Act to approve the Extension Project would cause an increase in CO2 emissions of 100Mt above the CO2 emissions that would otherwise occur (PJ [79], [84], [247] – [249]);

(e) if the Extension Project were to proceed, any CO2 emissions resulting from burning of coal extracted through that project would be outside the emissions contemplated by the “carbon budget” necessary to achieve a target of 2℃ above pre-industrial levels (PJ [86] – [87], cf [73]).

The respondents’ submissions

188    The respondents’ submissions, in their emphasis and framing of the correct or appropriate focus for analysis, emphasised at the outset two elements of the case. First, they emphasised that the evidence was uncontested and that it revealed the real risk of catastrophic world climate change and global warming by the end of this century. From that uncontested evidence the primary judge applied accepted legal technique of using analogical reasoning upon uncontested facts, guided by relevant legal principles.

189    Secondly, the respondents submitted that the primary judge was not concerned with a political question: The international and national responses to climate change are the concern of the Executive and the Parliament. The concern of the common law and the judicature is the remedy of claims in tort. In so rejecting at the outset the Minister’s submissions of the importance of policy considerations here, the respondents emphasised what Gaudron, Gummow and Hayne JJ said in Brodie 206 CLR at 560 [106]:

Appeals also were made to preserve the “political choice” in matters involving shifts in “resource allocation”. However, citizens, corporations, governments and public authorities generally are obliged to order their affairs so as to meet the requirements of the rule of law in Australian civil society. Thus, it is no answer to a claim in tort against the Commonwealth under s 75(iii) of the Constitution that its wrongful acts or omissions were the product of a “policy decision”' taken by the Executive Government; still less that the action is “non-justiciable” because a verdict against the Commonwealth will be adverse to that “policy decision”. Local authorities are in no preferred position. Yet it is submitted that those bodies which answer the description “highway authority”, distilled from the case law, merit and require a special consideration which only statute may displace. That submission should be rejected.

190    The respondents submitted that it would be a wrongful abrogation of judicial responsibility to deny a duty of care on the basis that the matter was for the Executive and a wrongful adoption of the American “political question” doctrine: cf Baker v Carr (1962) 369 US 186, which has not been accepted in Australia: Melbourne Corporation v Commonwealth [1947] HCA 26; 74 CLR 31 at 82; Thorpe v Commonwealth (No 3) [1997] HCA 21; 144 ALR 677 at 692; and see Re Ditfort; Ex parte Deputy Commissioner of Taxation (1998) 19 FCR 347 at 370–373.

Common law and statute: inconsistency, incoherence and policy

Duties of care owed by public authorities

191    The respondents submitted that, by section 75(iii) of the Constitution, the Executive may be liable in tort unless excluded by Parliament. It was submitted that, not once in the Caledonian Collieries Ltd v Speirs [1957] HCA 14; 97 CLR 202 line of authority has the statutory origin of the power (which in its exercises causes harm) denied liability at common law. Thus, it was submitted that the imposition of a duty of care on a public authority cannot of itself be said to be an impermissible distortion of that authority’s statutory power without departing from that line of authority.

The need for exclusion by, or inconsistency with, statute

192    The respondents submitted that in circumstances of a duty of care to avoid personal injury or death the test for inconsistency or incoherence was the exclusion of the duty by the statute in express terms or necessary implication. Incoherence cannot, however, be the disguise or cloak of a principle to preclude the Executive being liable for the negligent but valid exercise of a statutory power. Such would be in conflict with Crimmins 200 CLR 1, Pyrenees 192 CLR 330 and the premise of s 75(iii).

193    The exclusion must be clear: where the duty would require a breach of statutory duty, or be irreconcilable with it, or give rise to inconsistent obligations, or require the exercise of power the repository does not possess.

194    Cases of omission to act such as Graham Barclay 211 CLR 540 or Kirkland-Veenstra 237 CLR 215 should be distinguished from powers providing for the taking of a step which creates the danger. In the latter, there is no imputation of an intention to authorise causation of physical harm in the positive exercise of power when an implication is clear.

195    Different considerations apply in cases of pure economic loss where a public power requiring a balance of a range of public and private interests may be dictated by a duty of care to have regard to private economic interests.

No incoherence here

196    The respondents placed emphasis on the preamble of the Biodiversity Convention (which was given effect to in Australia by the EPBC Act) which acknowledges the importance of “biological diversity” for “maintaining life sustaining systems of the biosphere” where such diversity is said to have been “significantly reduced by certain human activities”. Within Pt 9 of the EPBC Act, the respondents placed emphasis on the requirement that the Minister consider “economic and social matters” (s 136(1)(b)), and, in doing so, consider the principles in s 3A of the Act, including long-term “economic, environmental, social and equitable considerations”, threats of “serious or irreversible environmental damage” and the “principle of inter-generational equity”. Further telling against any incoherence, it was submitted, was the object of the protection of the environment, including (especially) matters of national environmental significance, and the inclusion of “people and communities” and “social, economic and cultural” aspects within the definition of “environment” in ss 3(1)(a) and 528.

197    It was submitted that, taken together, these provisions authorised the Minister to consider the potential for an action to engender physical or psychiatric harm to humans. The found duty was said to be consistent with ss 3(1)(a), 3A, 136(1)(b) and 136(2)(a): nothing in those provisions was said to expressly exclude the Minister from civil liability.

Policy distinguished from coherence

198    The respondents relied upon Roads and Traffic Authority of NSW v Refrigerated Roadways Pty Ltd [2009] NSWCA 263; 77 NSWLR 360 as authority for the proposition that the policy/operational distinction is unhelpful as a determinant of when the imposition of a duty of care is warranted, especially in misfeasance cases. To the extent that such a distinction is part of the common law, it was said to arise for consideration at the stage of breach. Judges have no privilege to refrain from giving an answer: questions of policy afford no excuse: Bennett v The Commonwealth [2007] HCA 18; 231 CLR 91 at 122 [82] (Kirby J). By conferral of jurisdiction to hear a “matter” under Chapter III, it was submitted that no issues of justiciability arise in determining whether a duty of care is owed by the Minister in these circumstances.

Mandatory consideration

199    The respondents did not press the mandatory consideration point, and in fact submitted that his Honour could have reached the same conclusion on duty on the basis that human safety was a permissible consideration. The characterisation of human safety as a permissive consideration was accepted by the Minister.

Reasonable foreseeability

200    The respondents placed emphasis on the statement of McHugh J in Tame 211 CLR at 351–352 [96] that the test of reasonable foreseeability is “undemanding”. The primary judge, it was said, was correct to decline to incorporate causation into reasonable foreseeability. Reliance was placed on Bonnington Castings [1956] AC 613 to support the proposition that any contribution to the harm (amid multiple conjunctive factors) above de minimis is material, although the respondents submitted that the materiality of the harm is to be determined at the causation stage, and not here. Thus, that the contribution may be described as “tiny” does not detract from the foreseeability of that harm. The respondents submitted that each contribution to the accumulation of CO2 was a necessary (but not sufficient) cause of the accumulated whole that causes the harm. This was not, it was submitted, an adoption of the Fairchild approach to causation: CO2 emissions will accumulate and contribute to the total atmospheric concentration of CO2 which drives the world towards a tipping cascade. If that cascade occurs, it was submitted that it will have been caused by all accumulated carbon, rather than any given (yet unidentifiable) emitter or emissions.

Control, vulnerability and reliance

201    The respondents submitted that control need not be exclusive: it may be shared between one or more parties. The nature of the case, being a misfeasance (as opposed to nonfeasance) case was emphasised. That Whitehaven carries out the acts contributing to the accumulation of CO2 in the atmosphere was not to the point: a positive act of the Minister approved the extraction of coal for the purpose of combustion, and thus emission. The purported control was said to be greater than that in Smith v Leurs [1945] HCA 27; 70 CLR 256, or Home Office v Dorset Yacht Co Ltd [1970] AC 1004, and in no way analogous to Modbury Triangle 205 CLR 254. The finding of the primary judge, that the Minister has “substantial and direct control over the source of harm” was reiterated: an approval was said to unlock a causal chain that would contribute to the accumulation of atmospheric CO2 which would in turn cause harm by engendering climatic events like heatwaves and fires. In making this submission, the respondents placed great weight on the vulnerability of the Children.

202    The innate vulnerability of children as a class of persons was emphasised by the respondents, drawing on the parens patriae jurisdiction and the protective aspect of the relationship between the Executive and Australian children. Even outside of this relationship, defined by the unique capacity of the executive to protect the Children, the respondents submitted that the Children were nevertheless vulnerable to the harms of climate change due to the expectation that the Children will live in the last decades of this century wherein the effects of climate change will be experienced to a much greater extent.

No indeterminacy here

203    The respondents submitted that indeterminacy does not inhere in the size of the class, nor the number of claims that may arise. Indeterminacy is more significant, it was submitted, in cases of pure economic loss than in cases of physical harm where the physical consequences of a defendant’s conduct almost always have identifiable limits. The respondents submitted it is incorrect to argue that no duty can exist where the members of the class cannot be identified before the harm eventuates. In so arguing, the respondent gave a number of examples of duties owed in circumstances where the members of the class could not be identified prospectively, such as to all pedestrians using a particular footpath (Ghantous v Hawkesbury City Council [1999] NSWCA 51; 102 LGERA 399), or to those individuals who would be exposed to asbestos on an employee’s clothing (Caltex Refineries (Qld) Pty Ltd v Stavar [2009] NSWCA 258; 75 NSWLR 649). But it may be interpolated at this point that these examples are not to the point: while the members of the class could not be identified prior to the emergence of harm, the classes in question were not so expansive as to preclude identification entirely.

204    The unprecedented scope of the duty was said to be the product of the global scale of the potential harm that may result from climate change.

No factual errors

205    As to the five errors of fact identified by the Minister, the respondent submitted that those findings disclose no error.

Consideration and determination

Introduction

206    The search for a principled approach to the imposition of a duty to take reasonable care for others and their interests has not led in Australia to any analytical formula capable of mechanical application: see Graham Barclay Oysters 211 CLR at 622–629 [229]–[244] (Kirby J). In part, that is because of the need to draw out from the detail and context of human and societal relationships a legal duty that requires the exercise of reasonable care in some fashion. Taxonomy and definition play their part, but ultimately the law of negligence concerns itself with the protection of certain human interests against certain types of human (mis)conduct where it is reasonable and in accordance with standards of the day to impose a duty that may lead to personal responsibility for compensation to a person harmed by the (mis)conduct. The notion of neighbourhood taken from Donoghue v Stevenson [1932] AC 562 at 580 is built on the human and societal relationship between the parties: Mutual Life & Citizens’ Assurance Co Ltd v Evatt [1968] HCA 74; 122 CLR 556 at 566.

207    The exposition of the proper approach in Australia in the many High Court cases such as those set out at [121] above and discussed in Stavar 75 NSWLR at 674–676 [93]–[102] and in many other intermediate courts of appeal recognises an approach necessitating a close examination, not only of all the facts attending the relationship between the parties, but also the legal relationship in which they are situated, together with legal policy and principle drawn from analogous or cognate cases. The relationship is not one built by the salient features. Rather, the salient features play their part in considering the relationship and the question of the appropriateness of the imposition of a legal duty of care as the necessary foundation for the potential personal liability of the person to the injured party for damages as a consequence of the impugned act or omission in question.

208    There is no doubt that government and statutory authorities may be liable in negligence, and as a foundation for such be subject to a legal duty to take reasonable care for another or another’s interests. Relevantly for the foundation for the duty in this case, it can be accepted that a statutory authority may be liable for the exercise of a statutory power conferred on it if exercised without reasonable care: Speirs 97 CLR at 220 (Dixon CJ, McTiernan, Kitto and Taylor JJ). It is also clear that such liability is not to be denied on the basis that the exercise of the power was, in public law terms, lawful or without vitiating error. The view in England expressed by Lord Hoffman in Stovin v Wise [1996] 2 AC 923 at 953 to the contrary was not accepted in Pyrenees 192 CLR at 373–374 [119]–[120] (McHugh J), 376 [124] (Gummow J) and 424 [253(5)] (Kirby J) and Crimmins 200 CLR at 35–36 [82]–[83] (McHugh J, with whom Gleeson CJ agreed).

209    Such does not mean, however, that the answer to any questions of duty is to be answered by analysing each so-called salient feature on the hypothesis of the necessary amenability of the putative governmental defendant to potential liability.

210    This history of the development of the proper approach in Australia to the imposition of a duty of care, in particular in novel circumstances, seen in the High Court cases referred to in Stavar at 677 [107], demonstrates the difficulties thrown up in determining the existence, scope and content of a duty of care in the many different cases of asserted tortious liability. In Sullivan v Moody 207 CLR at 579 [50] the Court said of these problems:

Different classes of case give rise to different problems in determining the existence and nature or scope, of a duty of care. Sometimes the problems may be bound up with the harm suffered by the plaintiff … Sometimes they may arise because the defendant is the repository of a statutory power or discretion. Sometimes they may reflect the difficulty of confining the class of persons to whom a duty may be owed within reasonable limits. Sometimes they may concern the need to preserve the coherence of other legal principles, or of a statutory scheme which governs certain conduct or relationships. The relevant problem will then become the focus of attention in a judicial evaluation of the factors which tend for or against a conclusion, to be arrived at as a matter of principle.

(citations omitted)

211    The importance of this passage is that it makes clear that the judicial inquiry begins with an identification of the proper focus of attention or perspective to the relationship in question to make sense of the competing factors that may bear upon the relationship and upon the question whether a legal duty should be imposed. Without such a commencement, an approach to the inquiry by reference to many diverse features within the salient features risks fragmentation and confusion by individual particular analysis of features, almost in the abstract and divorced from context, without a proper understanding of the possible interrelations between the various features attending the relationship and the legal system as a whole. The considerations enumerated in Stavar, which were taken from cases in the High Court, were intended to assist an examination of a relationship to determine whether there exist in the relationship the requisite closeness, control and vulnerability for that relationship to warrant the imposition of a duty of care by reference to the legal conception of neighbourhood and whether the relationship was suitable for the imposition of a duty capable of founding liability judged by reference to judicial or curial determination. Considerations of policy and of coherence are normative, and go to whether an individual ought to meet the description (in a legal sense) of neighbour, even if the individual’s actions may otherwise, seen through the analysis of individual salient features, seem sufficient. The features listed in Stavar are not a catalogue to be used in the manufacture of a relationship of neighbourhood piece by piece. Depending on the relationship, it may (as it does here) invite confusion and error to begin, at the granular level, augmenting and elaborating upon those features in order to fabricate a relationship warranting the imposition of the duty of care. The salient features enumerated in Stavar are the frame of reference through which existing relationships, situated within their broader social and legal context, are to be examined. With respect, the primary judge did not utilise the salient features enumerated in Stavar as an analytical tool in this way. His Honour did not examine, through this perspective, the existing relationship of the Minister to the Children, within the context of her statutory function, bearing in mind her position as Minister, and the broader constitutional and legal context within which the dispute emerged. Rather, his Honour examined the salient features as individual considerations containing elements of the construction of the duty, leaving to the end of the analysis, almost as a miscellaneous consideration, fundamental considerations of the nature of the issues raised if the duty with its identified scope and content were imposed.

212    One must begin with the relationship. This is to be found in the statute properly construed and understood in its context. The nature of the posited duty, its scope and content, must be addressed to understand the nature of the issues to be thrown up at the point of breach. It is against this context that relevant salient features are to be addressed.

213    Understanding the core concern of the law of negligence, and the nature of relationships falling within the concept of neighbourhood, is of critical importance in this case. So too, is a proper understanding of the nature of the proceeding. This proceeding, and the putative duty as framed, is not, so to speak, a “standard” or “usual” torts case. It is, in many ways, unorthodox. It involves the imposition of a personal duty of care to a broad class (all persons resident in Australia under the age of 18 at the time of the commencement of the proceeding) on the Minister for the Environment (or upon her delegate from the Department) when exercising her statutory powers (which have the object of particular environmental protections, not involving the environment as a whole, as an end in and of itself). It involves the imposition of a duty of care in the context (on the uncontested evidence) of a potential global catastrophe for the world and all humanity that has been incrementally generated, and contributed to, by generations past and present throughout the world. It involves the disaggregation of the duty, breach and causation analyses which are generally intertwined, and, to a degree, indistinguishable: Roe v Minister of Health [1954] 2 QB 66 at 85 (Denning LJ). No harm has yet materialised. No relevant causal link to the harm has yet occurred. The decision in question has the potential to make a “tiny” (J[253]) contribution to a world-wide risk of catastrophic harm not only to the Children, but also to the world and humanity itself. This proceeding involves the recognition of a novel duty of care in circumstances unfamiliar to the law of tort as it presently stands. In this context, more than ever, it is paramount that the aspects of context (being the constitutional system of government in a federation, the broader legal system, the statutory context, and the contours of the law of negligence) be borne in mind at every step of the analysis. With respect, the failure to keep an eye attuned to context, coherence and what was necessarily thrown up at the point of breach by the posited legal duty was a defect in the method deployed by the primary judge. 

Grounds 2(a) and (b): human safety is not an implied mandatory consideration in the EPBC Act

214    Given its centrality in the primary judge’s reasons, in particular as to coherence, it is convenient first to address grounds 2(a) and (b). No attempt was made on appeal by the respondents to support the finding that, for the exercise of Minister’s power under ss 130 and 133 of the EPBC Act, human safety was a “relevant mandatory consideration in relation to a controlled action which may endanger human safety” (J[404]), and moreover that it was a mandatory consideration with “at least elevated weight” (J[402]), nor did they contend for such a construction at first instance.

215    The primary judge’s finding is inconsistent with the text, purpose and context of the statute.

216    The primary judge’s finding relied upon an inversion of the proper approach to statutory construction. Rather than starting with the express terms of ss 130, 133 and 136 of the EPBC Act read in context and with proper attention to statutory purpose, the primary judge made an a priori assumption about preservation of human safety being a relevant mandatory consideration and searched for contrary intention to displace that assumption. That assumption cannot be supported. The text of s 136 expressly identifies relevant matters that the Minister must take into account in making a decision under ss 130 and 133: human safety is not expressly (or indeed impliedly) identifiable as a mandatory consideration in the terms of s 136. The primary judge acknowledged this fact, stating that “human safety” as a relevant mandatory consideration was expressly not founded on any of those specified considerations in s 136: his Honour found that “human safety” as a relevant mandatory consideration sits outside s 136(1) altogether: J[406]. Implying human safety as a relevant mandatory consideration from the “scope and purpose” of the EPBC Act as a whole is expressly contrary to s 136(5), which requires the Minister not to consider any matters that the Minister is not “required or permitted by this Division” to consider (emphasis added), that being Div 1 of Pt 9. As has already been noted, the Full Court has described the effect of s 136(5) as creating a “closed system of matters” for the Minister’s consideration under s 136 (Tarkine 233 FCR 254 at 266 [28]). It would be incongruous to imply into that closed system a relevant mandatory consideration which is not supported by the express terms of that Division.

217    The primary judge also overstated the importance of the “protection of the environment” as a relevant object of the EPBC Act (including “environment” being defined to include “people and communities”) in the construction of the express terms of ss 130, 133 and 136. The primary judge failed to give proper attention to the context of the Minister’s statutory function under ss 130 and 133: the approval or refusal of a “controlled action” having regard to the protection of particular matters of national environmental significance in Pt 3, which in this case, is species, communities and water resources. This statutory function must be understood consistently with the overarching purpose of the EPBC Act being to give effect to the Commonwealth’s sphere of responsibility over matters within the broader context of its co-ordinate, and to an extent overlapping, responsibility with the States and Territories for the environment: paras 3(1)(a) and s 3(2)(a). The primary judge was therefore wrong to extrapolate from “the protection of the environment” being a relevant purpose of the EPBC Act to the implication of human safety as a relevant mandatory consideration for the Minister’s statutory function under ss 130 and 133.

The relationship between the Minister and the applicants and the class

218    The first matter to recognise is that the relationship finds its place by reference to the decision to be made under ss 130 and 133 of the EPBC Act. Ultimately the relationship is derived from the statute. The Act and the relationship here by reference to it, however, have a context.

219    The duty said to flow from the relationship and neighbourhood is to exercise reasonable care in the making of the decision to avoid causing personal injury and death in the future to all persons resident in Australia under a certain age from emissions of CO2 into the Earth’s atmosphere. Notwithstanding the primary judge’s statement at J[481] that the posited duty of care will not and cannot address climate change, that is exactly what it does do. The posited duty requires the Minister to examine and consider in the making of this decision all relevant material, including but not limited to scientific information, concerning the proper Australian response to global warming that bears upon that question of health and safety in the future for everyone in Australia below a certain age. Such a duty at (common) law in effect creates (through the common law) a mandatory duty in making this decision to take into account matters not required to be considered by the EPBC Act.

220    For the reasons discussed below, one cannot, with respect, dismiss the matters raised by the posited duty as core policy as enunciated by the Minister in the summary in J[478]. There was no false premise in the propositions there enunciated.

221    The context for the imposition or imputation of that duty includes the following. First, the decision required of the Minister is directed to specific and limited environmental considerations, being the matters referable to ss 18, 18A, 24D and 24E of the EPBC Act (listed threatened species and communities, and water resources) that led the delegate to decide that the extension of the mine was a controlled action. Thus, the occasion for the Minister to be subjected to the imposition or imputation of the duty concerning the risks posed by greenhouse gas emissions and global warming is fortuitous, and unrelated to the subject of the duty, being the specific present concerns with identified species, communities and water resources. Conformably with the specific concern of this decision, the EPBC Act is not directed at any place to climate change other than the regulation making power in s 520(3)(k) (under which no regulations have yet been made).

222    Secondly, through the process of bilateral agreement contemplated by the EPBC Act, another government within the Federation, that of New South Wales, examined in detail the very question that is at the centre of the subject of the duty: the effect of the extension of the mine on Australia’s contribution to greenhouse gases, as the New South Wales government was specifically required by New South Wales statute to do.

223    Thirdly, that examination and consideration, by or on behalf of the New South Wales government, employed and considered policy formed by both the New South Wales government and the Commonwealth government in how to deal with greenhouse gas emissions and in the context of the proper State and Commonwealth responses to the risks of global warming. This consideration of both national and State policy was required by the New South Wales legislation.

224    Fourthly, the policies of the New South Wales and Commonwealth governments employed in the very kind of analysis called for by the duty and undertaken by or on behalf of the government of New South Wales conformed with critical aspects of the Paris Agreement, being in particular the appropriate approach to Scope 3 emissions.

225    Fifthly, the consideration of the proper approach to the impact of the extension of the mine on greenhouse gases and the risks of global warming that was carried out and arrived at by or on behalf of the government of New South Wales involved policy decisions consequent upon international agreement which must be taken by the Judiciary in this context as made by the Executive government of New South Wales in the best interest of the State of New South Wales and its people. The same can be said of the Commonwealth policy that was taken into account in the examination by or on behalf of New South Wales.

226    Sixthly, the above approach by or on behalf of the government of New South Wales to the examination of the very question to which the duty is directed was accredited by the Commonwealth government under the bilateral agreement contemplated by the EPBC Act.

227    Seventhly, the very nature of the underlying danger that gives rise to the risk of world and human catastrophe is one that can only be addressed by global co-ordinated policy and action, by countries around the world formulating and implementing effective policy measures to address the nature of the cause of the potential catastrophe, in particular, to address the accretion, incrementally, over centuries of greenhouse gases from countless individuals, corporations and governments.

228    Eighthly, the development of that policy for any nation and for nations generally involves scientific, economic, social and political considerations, often depending on the nature and character of the countries in question, their populations and economies, including but not limited to industrial development and innovation in a decarbonised world and the development of energy sources alternative to fossil fuels. That is not to say that there cannot be seen to be policy or scientific imperatives of an overwhelmingly important character.

229    Ninthly, such policy or policies, all of which could affect and bear upon a particular decision about a particular coal mine, would involve not only the Minister for the Environment, but other parts of the Executive. Considerations of wide government decision making are recognised by the EPBC Act, s 131.

230    Tenthly, the role of the Judicial branch of government is to quell controversies between citizens or the state and citizens on the basis of evidence tendered by the parties, not on the basis of policy formulation by the court.

231    Eleventhly, the duty is sought to be imposed before damage is suffered and before causal connection to damage (as distinct from risk) even exists. Such disconnection is both temporal and geographic. This is most unusual for the law of torts. The respondents might say that to wait for the damage is to wait for the catastrophe. Yet that response would only highlight what might be seen as the nature of the present political imperative or duty to act as opposed to the imposition of a duty at common law which cannot crystallise into a cause of action until well into the future, possibly not within the lifetime of the putative tortfeasor. To disaggregate the duty from causation and damage is to remove duty from the essential nature or very essence of the cause of action: damage, of which it is a necessary part; and to found it not on damage, but on contribution to risk amongst contributions of countless other unidentified actors across the world and across time.

232    That context gives content to the real nature of the relationship: the relationship being one between the governing and the governed in a democratic polity, and referable to the EPBC Act.

The EPBC Act, core policy and incoherence: grounds 1 and 2(c)

233    Whilst questions of policy, unsuitability for judicial determination, incoherence and inconsistency can be viewed as separate considerations to a degree they are all inter-related and inter-twined. All must be assessed within the context of the EPBC Act, but also in relation to the legal system as a whole.

Core policy, unsuitability for judicial determination and the policy/operational dichotomy

234    It is necessary first to say something by way of introduction to core policy and the so-called operational/policy dichotomy.

235    The distinction between policy and operational decisions was recognised by Mason J in Heyman 157 CLR at 469. The utility of the dichotomy has been questioned (sometimes generally and sometimes for the resolution of a particular case) not only because the difference in character of, or the boundary between, the two concepts can be elusive (generally or in a particular context), but also because operational decisions may sometimes be based on policy: Anns v Merton London Borough Council [1978] AC 728 at 754 (Lord Wilberforce); Rowling v Takaro Properties Ltd [1988] AC 473 at 501 (Lord Keith of Kinkel); Stovin v Wise [1996] 2 AC at 951 (Lord Hoffman); Pyrenees Shire Council 192 CLR at 358–359 [68] (Toohey J) and 393–394 [181]–[182] (Gummow J); Romeo v Conservation Commission (NT) [1992] HCA 5; 192 CLR 431 at 492 (Hayne J); and Crimmins 200 CLR at 36–38 [84]–[90] and 50 [131] (McHugh J, Gleeson CJ agreeing at 13 [3]).

236    These difficulties and the clear contestability in many cases of the assertion of immunity of decisions of public authorities (cf Brodie 206 CLR at 558–560 [102]–[106] (Gaudron, McHugh and Gummow JJ)) do not undermine or deny the legitimacy, and acceptance, of the central idea to which Mason J was referring in Heyman: that there will be in some decisions of a public authority factors that make the law of negligence an inapposite or unsuitable vehicle for examining the choices and judgements involved. Such circumstances can be described as “core area of policy-making” or “quasi-legislative or regulatory in nature”: Heyman 157 CLR at 469 (Mason J) and 500 (Deane J); Pyrenees Shire Council 192 CLR at 393–394 [180]–[182] (Gummow J); Crimmins 200 CLR at 37–38 [87]–[90] (McHugh J, with Gleeson CJ agreeing); Dorset Yacht [1970] AC at 1067–1068 (Lord Diplock); and see the discussion of the question in Bienke v Minister for Primary Industries and Energy (1996) 63 FCR 567 at 593–596 (Black CJ, Davies and Sackville JJ) referred to with approval by Gummow J in Pyrenees Shire Council at 394 [182] and by McHugh J in Crimmins at 37 [87]. I have already set out what Gleeson CJ said in Graham Barclay Oysters 211 CLR at 553–554 [6]. Regard should also be had to the following in the Chief Justice’s reasons at 556–557 [12]–[15]:

[12] [Referring to the limits of the extent to which it is possible to assimilate the tortious liability of governments to that of subjects] …They arise from the nature and responsibilities of governments. In determining the existence and content of a duty of care, there are differences between the concerns and obligations of governments, and those of citizens. Such differences led to an attempt to distinguish between matters of policy and operational matters. That distinction was never rigorous, and its validity and utility have been questioned [citing Pyrenees and Stovin v Wise]. Even so, the idea behind it remains relevant in some cases, such as the present.

[13] [After referring to Mason J in Heyman at 469] One of the reasons why matters of the first kind are inappropriate as subjects of curial judgment about reasonableness is that they involve competing public interests in circumstances where, as Lord Diplock put it, “there is no criterion by which a court can assess where the balance lies between the weight to be given to one interest and that to be given to another” [citing Dorset Yacht at 1067].

….

[14] There are forms of governmental activity, which courts in the past endeavoured to describe by the term “operational”, where there is no reason for hesitating to assimilate the position of governments to that of citizens in imposing duties and standards of care. Such activity might involve budgetary considerations, but that does not prevent such assimilation. Individuals and corporations also have to watch their budgets, and decisions about what is reasonable may have to take account of that. As the other extreme, the reasonableness of legislative or quasi-legislative activity is generally non-justiciable.

[15] [After referring to the matter before the Court]… A conclusion that such a duty of care exists necessarily implies that the reasonableness or unreasonableness of the inaction of which complaint is made is a legitimate subject for curial decision. Such legitimacy involves questions of practicality and of appropriateness. There will be no duty of care to which a government is subject if, in a given case, there is no criterion by reference to which a court can determine the reasonableness of its conduct. That negative proposition leaves open other questions as to the circumstances in which the law will treat failure on the part of a public authority to exercise a power as a breach of a private law duty of care; but it is sufficient to resolve a substantial part of the case against the State in these proceedings.

See also 211 CLR at 561 [26]–[27] (Gleeson CJ also), 579–580 [90]–[91] (McHugh J) and 606–607 [175]–[176] (Gummow and Hayne JJ, with whom Gaudron J agreed at 573 [58]).

237    It must be recognised that the field of matters not amenable to the imposition of the duty of care at common law for such reasons is narrow: Crimmins 200 CLR at 37 [87] (McHugh J, Gleeson CJ agreeing). Further, it is not a question of labels. It would be a false syllogism to say or incant: there can be no duty of care as to policy (as distinct from operations), this is policy, therefore there is no duty. The proper approach is not taxonomy or definition or labels, but to recognise that some questions of decision-making are not a legitimate or apposite or appropriate subject of curial judgement, such as where there is no criterion by reference to which a court can determine the reasonableness of the conduct. This may involve policy-making, or quasi-legislative or regulatory power. The question is the suitability of the task for judicial determination by reference to a legal standard. The question is not definitional; it is a question of institutional inappropriateness or unsuitability.

238    This is not an abrogation of judicial responsibility or the adoption of some governmental immunity. It is to recognise that decisions that involve certain types of policy and which may have important physical consequences upon the lives, health, well-being, property and economic interests of people may be made by government in its decision-making role in the interests of the polity which cannot be judged by a legal standard or the consideration of which cannot be reliably made in a curial environment of private litigation. There are choices to be made by government which may affect lives, health and property which are made by reference to expertise unavailable or less than satisfactorily available to courts, by reference to political and democratic choices involving relationships of interests incommensurable by reference to any legal standard and which are appropriate for democratic (that is political) accountability, not by reference to monetary compensation where a legal standard to judge the choice is absent or faint.

Coherence in the law

239    It is also necessary, to say something by way of introduction as to coherence.

240    In Graham Barclay Oysters 211 CLR at 617 [213], Kirby J said:

Only one unarguable principle emerges from the earlier decisions, reflected in the Federal Court’s analysis. It is the self-evident one that any duty of a public authority at common law must be compatible with the legislative powers conferred, and duties imposed, on that authority. It must conform to the apparent purpose of the legislature relating to the authority carrying out its duties according to statute.

(citations omitted)

241    In expressing the matter thus, his Honour drew on Sullivan v Moody 207 CLR at 575 [36] and 576 [41]. Relevant also is the following passage from Sullivan v Moody at 576 [42]:

… But the fact that it is foreseeable, in the sense of being a real and not far-fetched possibility, that a careless act or omission on the part of one person may cause harm to another does not mean that the first person is subject to a legal liability to compensate the second by way of damages for negligence is there is such carelessness, and harm results. If it were otherwise, at least two consequences would follow. … Secondly, the tort of negligence would subvert many other principles of law, and statutory provisions, which strike a balance of rights and obligations, duties and freedoms. A defendant will only be liable, in negligence, for failure to take reasonable care to prevent a certain kind of foreseeable harm to a plaintiff, in circumstances where the law imposes a duty to take such care.

242    This passage from Sullivan v Moody reflects that the law of negligence is but one part of the Australian legal system: as part of that system, it is not intended that the law of negligence be extended so as to undercut, obliterate, or diminish other principles which may themselves serve other important values: Perera v Genworth Financial Mortgage Insurance Pty Ltd (trading as Genworth) [2017] NSWCA 19; 94 NSWLR 83 at 94 [42]. Rather, as Gaudron and McHugh JJ put it in Breen v Williams [1996] HCA 57; 186 CLR 71 at 115: “Any changes in legal doctrine, brought about by judicial creativity, must fit within the body of accepted rules and principles”.

243    In Hall v Herbert [1993] 2 SCR 159 at 169 [17], the central concern of coherence was put thus:

It is particularly important in this context that we bear in mind that the law must aspire to be a unified institution, the parts of which – contract, tort, the criminal law – must be in essential harmony. For the courts to punish conduct with the one hand while rewarding it with the other, would be to “create an intolerable fissure in the law’s conceptually seamless web” ... We thus see that the concern, put at its most fundamental, is with the integrity of the legal system.

(citations omitted)

244    In its broadest sense, coherence pertains to the consistency and harmony (logically and normatively) of the legal system as a whole: what Gummow J referred to in Hill v Van Erp [1997] HCA 9; 188 CLR 159 at 231 as “one coherent system of law” and the joint judgment in Sullivan v Moody at 580 [50] referred to as “the need to preserve the coherence of other legal principles, or of a statutory scheme”; see also Apotex Pty Ltd v Sanofi-Aventis Australia Pty Ltd [2013] HCA 50; 253 CLR 284 at 317–318 [46].

245    Coherence is to be judged certainly by reference to the statute providing for the decision in question, but also by reference to the legal context in which the statute sits, including the “basic values which the corpus of the law promotes or protects”: Tame 211 CLR at 395 [228] (Gummow and Kirby JJ), and the law of negligence and the fundamental values therein such as neighbourhood, reasonableness and incrementalism: as to the latter, see Heyman 157 CLR at 481 (Brennan J) and Perre 198 CLR at 216217 [93] (McHugh J), 302 [333] Hayne J and 325 [405] Callinan J. Further, at an even more fundamental level the duty must be coherent with the underlying constitutional system of federally structured democratic responsible government and the domain of the Judicature therein in the quelling of controversies between subjects and subjects, and subjects and the state in matters properly justiciable and amenable to resolution by reference to legal standards and proof brought forward by the parties.

Why the posited duty calls forth core policy for its consideration that is unsuitable for judicial determination and why this cannot be brushed aside as a false premise or as the introduction of the political question doctrine

246    The policy considerations are to be appreciated assuming the imposition of the duty. Debate took place in relation to, and the primary judge considered, the so-called policy/operational divide. The question of policy and the consideration of matters unsuitable for judicial determination should not be relegated in importance because of statements made in some cases as to the difficulty in drawing a boundary line between policy and operations or in the utility of the distinction. In reality, we are not dealing here with the distinction at all. We are not dealing with decisions of authorities in the allocation of budgets or in the choice of policy as to planning and constructions of roads. We are dealing with core, indeed, high policy-making. There are many fields, as Gaudron, McHugh and Gummow JJ said in Brodie 206 CLR at 560 [106] where so-called policy cannot be used to hide a breach of requirements of the law. If a road authority wishes to rely upon policy questions and resource allocation as a justification for not repairing a plainly dangerous bridge of which it is aware, it can propound its case at the point of considering breach of duty and will be judged by reference to considerations amenable to, and suitable for, judicial determination: Brodie at 560 [106]. Here, if the duty is to be imposed, the considerations and judgements attending the decision will involve an evaluation of the adequacy of national and State policies taken within the framework of international agreement and whether such policy or policies was or were adequate and whether it or they should be persisted with or remade. Such policy or policies and such agreement draw out and are based on multi-disciplinary considerations including environmental, climatic, economic, social and political considerations. These are not only the natural province of the Executive and Parliament, but also, by their nature and character as attending the safety and wellbeing of all Australians, and (on the uncontested evidence) the wellbeing of the world and humanity, comprise the foundation of the (public or political) duty of consideration and policy formulation by at least the Executive on this subject.

247    The duty concerns matters of so-called “core” policy. Such is to appreciate the nature and character of the necessary appropriate power to deal with the problem and so the nature and extent or scope and content of the duty. The evaluation of good or bad decision-making about greenhouse gas emissions and the risks of global warming is one to which the highest considerations of the welfare of the Commonwealth attend, by reference to a range of matters that involve scientific, social and economic considerations and ultimately democratic political choice. This can be called, at the very least, core government policy. It is perhaps better described as public policy of the highest importance. The uncontested evidence makes that pellucidly incontestable.

248    The fundamental question of legality aside, it is not the function of the Judicial branch to rule upon any lack of adequacy or any lack of wisdom of government policy by reference to the law of torts. These matters of policy are not ones merely of resource allocation, funding and other such matters to which statutory authorities often must direct themselves, dealt with in cases such as Brodie. The matters of policy here involve considerations drawn from the gravity and seriousness of the international subject matter: the risk of catastrophe for the world and humanity. These considerations and the policy response involve scientific, social, economic and political factors and choices to inform the appropriate national and State responses to a world problem the subject of international discussion and agreement, conformable with the maintenance and advancement of the well-being, including health and safety, and social and economic circumstances of the people of Australia (for the Commonwealth) and of New South Wales (for New South Wales).

249     The development of that policy requires consideration of information, of judgement, of political considerations of an international and national character, many of which are incapable of being assessed and evaluated by frames of reference amenable to reduction to a legal standard or duty in private litigation by reference to evidence led by individual parties. This is not the exercise of a statutory power to construct a safe railway crossing (Speirs), or to take steps to prevent fires in buildings (Pyrenees), or to take steps and implement policy to prevent pollution of water in which oysters intended for human consumption are growing (Graham Barclay), or not to require men to work in the holds of ships showered in poisonous asbestos (Crimmins). The appropriate national and international policy here concerns mitigation of the risks of world climate disaster. It is thrown up by the duty at the point of assessment of breach.

250    The natural places for the development of such policy and the making of decisions as to the implementation of such policy is the Executive branch of government, and Parliament. Both have the power and the ability to obtain all relevant up-to-date information bearing upon policy. Both arms of government are, in a parliamentary democracy, responsible to the people of the polity. Further, Parliament has acted: Clean Energy Act 2011 (Cth), repealed in 2014 as set out at [5] above.

251    The authorities to which I have referred make clear that so-called core policy, or at least the making of it, is not, or is unlikely to be, the province of the Judiciary in its role of quelling private controversies or controversies between individuals and government.

252    The respondents submitted that the proceeding was not about climate change policy, or any political question related thereto. Politics, national and international, was the preserve of the Executive and Parliament. The Judiciary, it was submitted, applied the common law based on foreseeability of harm from the exercise of a statutory power. The respondents, as did the primary judge (J[485]), relied upon Brodie 206 CLR at 560 [106] in particular to seek to make the question one for the common law. But nothing in Brodie denies the recognition in the cases to which I have referred of the unsuitability of some policy for curial assessment, in particular when the duty throws up policy-making at the point of breach. The expressed views of the authors of that passage in Brodie (Gaudron J, McHugh J and Gummow J) in Graham Barclay Oysters that broadly conformed with the views of Gleeson CJ in the same case (see [16] above) make that clear: see in particular McHugh J at 211 CLR 580 [91], and the express distinguishing of the circumstances in Brodie by Gummow and Hayne JJ at 211 CLR 606607 [175], with whom Gaudron J agreed at 573 [58].

253    Just as there is no foundation to say that policy always denies a duty, there is no foundation to say that the common law ignores policy. It does not. Core policy or policy-making is recognised by high authority as a field in which the courts may not be the suitable governmental vehicle for resolution or determination of an issue or subject. A duty that calls forth, through the future dangers to people in Australia, the proper national response to a subject rooted in international agreement and co-operation and raising the range of considerations to which I have referred, is not to be side-stepped by saying: We have proved on the basis of looking at a body of (uncontested) scientific considerations risk of or an increased risk of physical harm in the future, therefore the common law and the Judiciary now take over as the vehicle for imposing potential monetary compensation against decision-makers (and thereby create the venue or vehicle for determining Australia’s response to climate change).

254    It can be accepted that statutes work within the common law and that the statute may exclude a duty of care: Crimmins 200 CLR at 1819 [26]–[27] (Gaudron J). That does not mean that foreseeability of harm necessarily creates a duty upon an authority unless the duty is specifically excluded by statute. Apart from the vice of creating a unitary reflex of foreseeability of harm and duty (contrary to the approach by reference to neighbourhood and salient features dictated by High Court authority), it fails to recognise that the refusal to impose a duty when such will throw up at the point of breach questions of policy-making or other considerations unsuited to judicial determination comes from principle within the common law itself.

255    I respectfully agree with the expression of the matter by the Court of Appeal of New Zealand in Smith v Fonterra Co-operative Group Limited [2021] NZCA 552 at [116] and [26] in dealing with a posited duty about greenhouse gas emissions: “Courts are … ill-equipped to address the issues that the claim raises”, which call for “a level of institutional expertise, democratic participation and democratic accountability that cannot be achieved through a court process”.

256    The application of such considerations does not depend upon the Parliament having enacted legislation to address global warning. Such action may make the conclusion plainer, but the issue is the lack of suitability of the assessment of climate change policy for judicial determination in private suits in the law of tort.

257    Nor are these considerations diminished by the fact that the suit in Smith v Fonterra was against private parties. Here, the imposition of the duty would likewise see the venue for the consideration and judgment of the suitability and adequacy of Australian climate change policy moved to the Judicial branch of government.

258    Nor are these considerations diminished by the nature of the pleaded claim in Smith v Fonterra. That there were features in that case making it even more unsuitable for judicial determination does not weaken the point.

259    Nor are these considerations diminished by the legal structure of the law of torts in New Zealand based on Anns [1978] AC 728. The question of policy arises in New Zealand from another perspective. But the question of the unsuitability of core policy of this kind to judicial decision-making is a central feature of the Australian law of torts.

260    As is clear from the above, central to the conclusion of the unsuitability of the judicial branch to judge the legitimacy of the decision in question is the nature and extent or scope and content of the duty said to attend the exercise of the statutory power. Here the duty as framed involves the Minister taking into account Scope 3 emissions and the present state of science in relation to global warming and the dangers to the world and humanity therefrom. This duty is to be imposed in the context of consideration of particular species, communities and water resources. The duty plainly calls forth policy to be judged, evaluated or re-evaluated through the prism of reasonableness at common law. The distinction between act and omission does not assist in denying the reality that the posited duty calls forth such policy questions. To the extent that “policy-making” is required for the application of these cases, that is what will be called for here. The duty requires Scope 3 emissions to be taken into account in the evaluation of reasonable conduct. This is the very matter covered by New South Wales and Commonwealth government policies and the Paris Agreement. With respect, one cannot dismiss this as a false premise (J [479]). It is not. A consideration of the proper response to the present risks, future dangers and potential harm to the world, humanity, and the Children is a matter of high public policy. To the extent that the evidence and the uncontested risks of climate catastrophe call forth a duty of the Minister or the Executive of the Commonwealth, it is a political duty: to the people of Australia.

261    That is not to say that a decision under ss 130 and 133 of the EPBC Act cannot possibly be attended by a duty to exercise reasonable care to a class of persons in the circumstances. Whilst the consideration of social and economic matters may make it difficult to adjudge negligence, a duty posited (say to exercise care in relation to the approval of a mine on land containing a large deposit of blue asbestos near a centre of population) may make the question of reasonableness of the decision properly amenable to judicial determination by reference to a legal standard: cf Crimmins at 37 [87] (McHugh J, Gleeson CJ agreeing). Any such question is to be answered, however, in a set of facts posed by a concrete problem.

262    That is not the position here. The duty calls forth a consideration of climate change policy: its making, remaking or application, bearing in mind all the complexity of that policy development. This is exacerbated by the disaggregation of consideration of duty, from breach and causation. The central consideration of policy being call forth, will likely be considered many years after the time of alleged breach with all the difficulty of searching back in time to assess the reasonableness of conduct by the Minister by reference to the long-past 2021 policy which must be said to be required to be re-evaluated for there to be a breach.

263    The conferral of the power on the Minister reinforces, though it does not determine, the proper sphere of accountability with the kind of policy considerations thrown up by the posited duty.

264    The duty, to use the words of Gleeson CJ in Graham Barclay Oysters 211 CLR at 553554 [6], invites “the judicial arm of government to pass judgment upon the reasonableness of the conduct of the … executive arm of government” in a manner that raises “issues that are inappropriate for judicial resolution, and that, in a representative democracy, are ordinarily decided through the political process.”

265    It is unnecessary to consider whether the question of policy denying a duty is limited only to so-called “policy-making”. There are powerful dicta to that effect: Heyman 157 CLR at 469 (Mason J) and 500 (Deane J), Pyrenees 192 CLR at 393394 (Gummow J), and Crimmins 200 CLR 1 at 37 [87] (McHugh J, Gleeson CJ agreeing) and cf Rowling v Takaro Properties [1988] AC at 501. But narrow characterisation is not called for. A policy may be present. Here the evidence discloses it: Scope 3 emissions were the responsibility of countries that combust the coal. It was national policy, State policy and the approach that was drawn from the Paris Agreement. The material propounded to call forth the duty was explicit and clear: the facts of global warning have outstripped the features and contours of that international response, making it inadequate. Thus the duty called forth the question to be examined at the point of breach whether the Minister, not to be negligent, must make a decision about this coal mine that reflected a change or re-evaluation of that policy. That is not only core policy, but public policy of the highest importance to the nation and the people of Australia.

266    These considerations of themselves, without more, make clear the nature of the relationship: It is the relationship of government or the governing with the governed. If the relationship and the uncontested evidence together call forth a duty, it is a political duty, not a legal duty of care.

Why the posited duty is incoherent with the text and structure of the EPBC Act, and with the legal and government framework of responsibility for the protection of the environment reflected therein

267    The duty would be incoherent and inconsistent with the EPBC Act and the legal and government framework of responsibility for the protection of the environment reflected therein.

268    The duty would create a form of mandatory consideration beyond the considerations in respect of the decision in question provided for by the EPBC Act on its proper construction. The EPBC Act does not contain as an overall purpose the safety of human life. The relevant purposes of the Act depend upon the provisions in question. Here those considerations concern the protection of species, communities and water resources. The creation of an overriding common law duty to a significant proportion of the population of Australia to consider all factors and information concerned with greenhouse gas emissions and the risks of global warming, and the proper policy response thereto is to impose upon the EPBC Act, or overlay the EPBC Act with, a responsibility and duty of the Minister personally not found within the statute and which, given its capacity for personal liability, would be to change the whole nature of the decision-making in question. That change would occur through, or be occasioned by, the unrelated fortuity of the Minister or a delegate having a concern about the effect of the action upon species, communities or water resources.

269    Fortuity, luck or happenstance sometimes enliven the obligation to act by a decision-maker or an authority: see Pyrenees 192 CLR 330. That can be accepted. But the obligation to act in such cases will conform with the underlying statutory power and duty and the relationship established by that statute.

270    That the consideration of greenhouse gas emissions may be a consideration that is permitted to be taken into account by the Minister in respect of the making of the decision does not mean that the duty should be imposed. The parties were agreed that the Minister may choose to take these matters into account as a social or economic matter (cf Sean Investments Pty Ltd v MacKellar [1981] FCA 191; 38 ALR 363). If the Minister did so she would do so under the EPBC Act in the decision as to species, communities, and water resources. But it would be to create an entirely different decision from that contemplated by the EPBC Act to require under the common law the Minister to take into account and consider, under a duty to all people in Australia below a certain age, global warming risks associated with greenhouse gases and the proper policy response thereto.

271    That broader and different decision, made not just by reference to considerations attending species, communities and water resources, and any social and economic matters of a relevantly related character to them, but by reference to the whole question of greenhouse gases and the risks of global warming would duplicate and set up the possibility of inconsistency with decision-making of a coordinate government in the field, here the government of New South Wales, in its consideration of the very same issue pursuant to the requirements of State legislation, made relevant by the bilateral agreement between the State and Commonwealth governments in the management of the EPBC Act.

272    The text, structure and context of the EPBC Act reveals a shared governmental responsibility for the environment. The EPBC Act is not directed to the proper response by the Commonwealth to the risks of climate change and global warming. That was a deliberate choice recognised in the text and structure of the EPBC Act, and in the COAG review and the COAG agreement from which the Act emerged. To impose the duty upon the Minister (effectively as a mandatory consideration) in the fortuity of circumstance to which I have referred would be inconsistent with that text and structure of the EPBC Act in its context of the wider legal and governmental framework of responsibility for the protection of the environment.

Challenge to the factual findings in ground 5

273    Before considering grounds 3 and 4 of the notice of appeal, it is necessary first to deal with the alleged five factual errors in the reasons of the primary judge. Each of the impugned findings rely, to varying degrees, upon the evidence of Professor Steffen. As will become apparent, these alleged factual errors form part of the factual matrix underlying the primary judge’s reasonable foreseeability analysis, and therefore assume some significance.

274    Before turning to the alleged factual errors, it is important to keep in mind the state of the factual “puzzle” before the primary judge: Fox v Percy [2003] HCA 22; 214 CLR 118 at 132 [41]. Professor Steffen’s evidence was unchallenged. The respondents’ evidence was led on a final basis in support of injunctive relief. The primary judge’s findings emerged after a contested factual hearing in which the Minister had every opportunity to contest or seek to contradict the evidence of Professor Steffen. The Minister did not cross-examine Professor Steffen and did not lead any responsive or supplementary evidence. Professor Steffen’s evidence was therefore left unchallenged.

275    It was in this context that the primary judge, immersed in the evidence, was left to draw inferences and arrive at the “correct and lawful conclusion to the puzzle presented to him”: Fox v Percy 214 CLR at 132 [41] (Gleeson CJ, Gummow and Kirby JJ). As the trial judge, his Honour heard and saw all the evidence in its entirety, and had an “opportunity to reflect on the evidence and to weigh particular elements against the rest of the evidence whilst the latter [was] still fresh in mind”: State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (in liq) [1999] HCA 3; 73 ALJR 306 at 330 [90] (Kirby J), cited with approval in Fox v Percy 214 CLR at 228 [28]. Respect should be afforded to the primary judge’s endeavour to draw the correct inferences. The question for the appellate court is whether the impugned factual findings disclose error, which may be demonstrated by the appellate court coming to a different view after affording the necessary respect to the primary judge’s careful endeavour.

Ground 5(a)

276    Under ground 5(a), the Minister challenged the primary judge’s finding at J[31] and [74(i)] that the best available outcome that climate change mitigation measures can now achieve is a stabilised global average surface temperature of about 2℃ above pre-industrial levels. The Minister contended this is a significant rounding up of Professor Steffen’s evidence. In particular, the Minister points to Professor Steffen’s statement that Scenario 1, being the “best possible outcome” of three possible future worlds posited (see [109] ff above), “would lead to a global average surface temperature in 2100 that would be approximately equivalent to, or slightly higher than, the upper Paris accord target of ‘well below 2°C’”. The Minister submitted that the upper Paris accord target of “well below 2°C” means approximately 1.8°C, and therefore the primary judge proceeded on a false premise that the “best available outcome” is stabilisation at 2°C: J[31] and J[74(i)].

277    I agree with the respondents’ contention that this attack on the primary judge’s finding is a quibble which seizes on one use of words at the expense of a fair reading of Professor Steffen’s evidence as a whole. While there are some differences in the way Professor Steffen described the stabilised averaged temperature under Scenario 1 (“at, or very close to, 2°C”, “around 2°C”, “approximately 2°C” and “a 2°C target”), it was open on the evidence to conclude, as the primary judge did at J[89], that Professor Steffen meant “2°C or slightly lower but not ‘well below 2°C’ and not the upper target of the Paris Agreement”. This finding is consistent with the estimates of the Intergovernmental Panel for Climate Change (IPCC) for Scenario 1 and Professor Steffen’s evidence in paragraph 11 of his 3 June 2020 report (which was before the IPC) in which he stated that the “upper Paris target”, taken to be 1.8°C warming, “is very difficult to achieve”.

278    Ultimately, Professor Steffen was engaging in a predictive exercise based on scientific data which is inherently imprecise. The primary judge’s conclusion that Professor Steffen meant “about 2°C” was clearly open on a holistic view of the evidence, which was unchallenged. The Minister did not avail herself of the opportunity to clarify, through cross-examination, what exactly Professor Steffen meant by the slightly differently expressed references to 2°C (including in the context of discussing the Paris Agreement). Nor did the Minister lead evidence to the contrary that the best possible outcome was “well below 2°C” or “1.8°C”. This finding by the primary judge discloses no appealable error.

Ground 5(b)

279    Under ground 5(b), the Minister challenged the primary judge conclusion that at a stabilised global average temperature above 2°C, there is an exponentially increasing risk of the Earth being propelled into an irreversible 4°C trajectory: J[31], read with J[74(iii)] and J[75]. The Minister submitted that Professor Steffen’s evidence, properly characterised, was that there was a small probability of initiating a tipping cascade at 2°C, but a “significant risk” that a tipping cascade would be activated by a 3°C or even lower temperature rise. This could not support a finding that the risk between 2°C and 3°C would increase on a continuum or exponentially, nor was the evidence put at a “sufficiently granular” level to permit a finding as to the nature of the relationship between 2°C and 3°C except for the fact that 2°C was “small” and “much lower” and 3°C carried with it a significant risk.

280    The Minister’s contentions should be rejected. It was accepted before the primary judge that there is an approximately linear relationship between CO2 emissions and the Earth’s global average surface temperature in the absence of non-linear feedback effects. Professor Steffen’s unchallenged evidence was that at or above 2°C warming above pre-industrial levels, there is a small but not-zero risk that a “tipping cascade” could be initiated by feedback processes accelerating global warming in a non-linear fashion. These non-linear feedback processes are of three basic types: melting ice, such as the melting of the Artic sea ice and loss of ice from Greenland and Antarctica; forest dieback, such as in the Amazon and boreal forests; and changes in Earth system circulation patterns. At 3°C or even lower, Professor Steffen concluded that there was a “significant risk” of these feedback processes occurring, resulting in the acceleration of CO2 emissions and therefore the acceleration of global warming. Indeed, Professor Steffen found that there is a “significant risk” that a 3°C future world “is not accessible” due to the risk of non-linear feedback processes preventing the stabilisation of the global average temperature if global warming approached 3°C above pre-industrial levels. While Professor Steffen did not use the word “exponential”, on a fair reading of his unchallenged evidence it was open to conclude that because of non-linear feedback processes, an increase in CO2 emissions and therefore global average temperature above 2°C will likely cause a more rapid increase in these tipping elements occurring and the risk of triggering a tipping cascade and if triggered, an irreversible 4°C trajectory.

281    This finding is also supported by the IPCC’s 2014 Synthesis Report. On p 73 of that report, a figure contained in Box 2.4 shows there is “moderate” level of additional risk of large-scale singular events (which carry the same or a substantially similar meaning as Professor Steffen’s feedback processes) occurring at 2°C, while there is a “high” level of additional risk of the same at 3°C. The figure clearly portrays the risk of large-scale singular events increasing on a continuum between 2°C and 3°C. This report, read with Professor Steffen’s evidence, leaves open a finding that there is an “exponential”, or “more rapid”, increase in the likelihood of a tipping cascade as temperature increases from 2°C to 3°C due to the increased risk of triggering non-linear feedback processes.

Ground 5(c)

282    Under ground 5(c), the Minister challenged the primary judge’s finding that even an infinitesimal increase” in global average surface temperature above 2°C warming above pre-industrial levels may trigger a 4°C future world: J[253]. The Minister contends that Professor Steffen’s evidence did not descend into sufficient detail to enable a finding that any infinitesimal increase in global surface temperature beyond 2°C carries a “real risk” of triggering the tipping cascade. This was said to be exemplified by the fact that when Professor Steffen was asked to revisit answers concerning the materiality for global warming of 336 Mt of CO2 emissions and instead consider the materiality of 100 Mt, his answer was the same. In oral address, the Minister contended that this demonstrates that Professor Steffen’s analysis was not at a level of sensitivity that suggested an infinitesimal contribution was enough to trigger a 4℃ future world.

283    It may be accepted that Professor Steffen’s evidence did not descend into detail as to precisely what amount of CO2 emissions may be necessary to trigger a tipping cascade, nor was he asked to. In that context the Minister made a forensic decision at trial not to put to Professor Steffen any question of the materiality of his analysis and the change from 336 Mt to 100 Mt of CO2. As noted above, the appellant accepted before the primary judge there is an approximately linear correlation between the accumulation of CO2 emissions in the Earth’s atmosphere and an increase in the global average temperature. The effect of Professor Steffen’s unchallenged evidence was there was a real (albeit small) risk of triggering a tipping cascade from about 2°C global warming above pre-industrial levels due to non-linear feedback processes. Indeed, Professor Steffen stated that some non-linear feedback processes (or tipping points) had been triggered at the current rise in global average surface temperature of 1.1℃ above pre-industrial levels. It is simply not possible to say with certainty which and how much CO2 emissions, and therefore at what temperature, certain non-liner feedback processes will be triggered. As stated with respect to ground 5(b), it was open for the primary judge to conclude, on Professor Steffen’s unchallenged evidence, that the risk of triggering non-linear feedback processes (and therefore the risk of a tipping cascade) increases in a more rapid fashion as the global surface temperature increases from about 2°C. In that context, it was open to the primary judge to find that there is a real (being a small but not-zero) risk that an increase in CO2 emissions which caused an even infinitesimal increase in global average temperature above 2°C could trigger a tipping cascade and consequently an irreversible 4°C future world trajectory. No appealable error has been shown.

Ground 5(d)

284    Under ground 5(d), the Minister challenged the primary judge’s finding that a decision under the EPBC Act to approve the extension of the mine would cause an increase in CO2 emissions than that which would otherwise occur: J[79], [84], [247]-[249]. This challenge can be dealt with shortly. It is true, as the appellant contended, that Professor Steffen proceeded on the assumption, pursuant to his instructions, that the project would cause an increase in 100 Mt of Scope 3 emissions. Yet the Minister did not challenge that premise underlying Professor Steffen’s report before the primary judge, and therefore must be taken to concede that fact. That premise in turn was based upon the reports of the IPC and the NSW Department which found that the extension of the mine would result in 100 Mt of Scope 3 emissions. The Minister did not seek to lead contrary evidence to those reports.

285    It is therefore uncontroversial that if the extension of the mine was approved, the coal extracted and burnt will emit 100 Mt of CO2 into the Earth’s atmosphere. Whether or not another source of coal might be extracted and burnt (such as to replace the customer’s needs from other sources) if the extension of the mine were not approved does not go to the question of fact necessary for the imposition of a duty of care: namely whether 100 Mt of CO2 emissions would result from the approval of the extension of the mine. The Minister did not seek to lead evidence before the primary judge as to a counterfactual if the extension of the mine were not approved and the respondents rightly contend that they were not required to disprove such a counterfactual. In any event, any such evidence or defence based on the notion that 100 Mt of CO2 (or more, if inferior coal were used to that which the mine could produce) would be released into the Earth’s atmosphere by the burning of an alternative source of coal if the extension of the mine were not approved, and therefore that the respondents would suffer the same harm regardless, would not preclude a duty of care: the duty of care enquiry is prospective towards the risk of harm from a negligent act or omission.

Ground 5(e)

286    Under ground 5(e), the Minister challenged the primary judge’s findings that there was:

(a)    “not sufficient evidence before me on which I could conclude there is no real prospect of the 100 Mt of CO2 being burnt outside the available fossil fuel budget necessary to meet a 2℃ target” (J[86]); and

(b)    “there is evidence before me which tends to support the proposition that 100 Mt of CO2 will not be emitted as part of the available carbon budget necessary to achieve a 2 target” (J[87]).

287    During the course of oral argument, the Minister accepted that it could not exclude the possibility that the coal from the extension of the mine may be consumed outside the Paris Agreement or in breach of it. However, the Minister maintained that the primary judge’s findings were unsupported by the evidence. This argument was advanced in two ways.

288    First, the Minister contended there was evidence before the primary judge, in the report of the NSW Department which was before the IPC, that Scope 3 emissions caused by the combustion of coal from the extension of the mine overseas would be included in the Scope 1 emissions of the likely export destinations, namely Japan, South Korea and Taiwan. The report stated that each of those countries have developed emissions reduction targets that are in line with the Paris Agreement (although Taiwan is not a signatory to that agreement, a matter the primary judge appeared to rely upon at J[86]). While it is true, as the respondents contended, that the Minister did not lead any evidence of compliance by those countries with those emissions targets, the Minister submitted that the primary judge’s treatment of the emissions as additional to the best possible outcome of Professor Steffen, that being stabilisation of the average global temperature at about 2 warming about pre-industrial levels, was erroneous.

289    Secondly, the Minister argued that the primary judge mischaracterised Professor Steffen’s evidence which contained internal consistencies. The primary judge accepted that, on the definition used in the relevant evidence (that of McGlade and Elkins referred to by Professor Steffen), the extension of the mine involved mining coal from an “existing” reserve: J[73]. Therefore, there was no basis to infer that when Professor Steffen referred to “existing” reserves he “must have meant those already being exploited” (J[87]) (and therefore used a different definition to that used by McGlade and Elkins, who defined existing reserves as those it was “economically and technically viable to exploit now”: J[73]). This was said to undermine the primary judge’s premise that the emissions associated with the extension of the mine would be additional to the emissions within the “carbon budget”, as the “carbon budget” approach contemplates that some proportion of the world’s existing fossil fuel reserves can be consumed without leading to an increase in temperatures beyond 2.

290    There is some force to the Minister’s contentions concerning internal inconsistencies in the primary judge’s analysis and the relevance of the NSW Department report. However, the factual challenge under ground 5(e) ultimately cannot succeed in light of Minister’s failure to challenge Professor Steffen’s opinion that it was “obvious” from a carbon budget analysis (that being a carbon budget, on Professor Steffen’s view, of approximately 866 Gt of CO2 emissions that can be burnt from 2021 for a 67% probability of stabilising global average temperature at 2 above pre-industrial levels) that “no new coal mines, or extensions to existing coal mines, can be allowed” and that “currently operating coal mines must be phased out as soon as possible”. I accept the respondents’ contention that it is clear that Professor Steffen’s view was that the extension of the mine would be incompatible with the carbon budget he associated with Scenario 1. The Minister did not cross-examine Professor Steffen on this conclusion as to obviousness. Had the Minister wished the primary judge to come to the conclusion that the coal from the extension of the mine would likely be burnt in line with the Paris Agreement and therefore within a “carbon budget”, she could have led evidence as to that conclusion and challenged Professor Steffen on his conclusion as to the matter being obvious. Similarly, if the Minister wished to rely upon the NSW Department report as evidence that the coal from the extension of the mine would be burnt in other countries in accordance with Paris Agreement targets, she should have put this to Professor Steffen, who expressed a directly contradictory (and unchallenged) opinion. Ground 5(e) can therefore not succeed.

Coda to ground 5

291    The above debate makes clear the fundamental policy-making element embedded within any question of breach thrown up by the posited duty. The Minister’s case on duty was that approaching the matter by reference to Scope 3 emissions (being the foundation of Commonwealth and New South Wales policies) was not shown to be impermissible. The respondents propounded Professor Steffen’s views in effect that the science (and the respondents say: the duty) requires a variation to that policy, a moving on from the Paris Agreement to avoid the risks of harm (and the respondents would say at the point of beach: not to breach the duty).

292    The unsuitability of the judicial branch deciding such questions in private litigation, dependent on whether cross-examination and evidence was adequate, is clear. As the primary judge himself stated at J[479] the correctness of the proposition that the Executive (and, I would add, Parliament) is best placed to deal with the complex task of addressing climate change than the common law cannot be doubted.

293    The focus of the complaints in ground 5 raises directly the issue of the adequacy of climate change policy based on Scope 3 emissions being a matter for countries purchasing the coal, and the appropriateness of that being decided in a framework of analysis based on common law negligence in a court to be decided by evidence chosen and led by the parties and subject to the forensic choices of advocates.

The law of negligence: reasonable foreseeability and causation, control, vulnerability and reliance, and indeterminacy: grounds 3 and 4

294    If the consideration of policy and, separately, incoherence or inconsistency with the EPBC Act is or are not determinative of the decision not to impose a duty of care, such considerations are still relevant. This is so because the ultimate question as to whether the law imposes a duty of care compensable in damages for breach involves the reasonableness and appropriateness in all the circumstances of the relationship of such imposition. Even if the questions of policy and lack of coherence and inconsistency to which I have referred are not determinative, they are relevant to considering the totality of the relationship, upon which the reasonableness and appropriateness of imposing a duty of care upon the Minister of the content and scope posited depends: Graham Barclay Oysters 211 CLR at 596 [145] (Gummow and Hayne JJ).

Disaggregation of duty of care from damage and causation

295    I have already referred to the unusualness in the law of tort of a duty of care being imposed in the absence of damage or likely threatened damage from the tortious conduct. Not only is it unusual, but also it has, in these circumstances, the vice of separating the imposition of the duty from the existence of the cause of action or damage of such immediate direct threat as to justify a conclusion of the threat of a cause of action. To a degree, this is linked to a problem of lack of causally connected foreseeable harm (as opposed to a tiny contribution to a state of overall risk for the future) to which I will come shortly.

296    Disaggregation, of itself, may not be any vice. A plaintiff may be able to show threatened harm from threatened negligent conduct. Where such conduct involves foreseeable consequences of damage causally connected to the conduct, an injunction may lie if there is a duty of care: The injunction will be to restrain breach and so restrain foreseeable caused harm.

297    Here duty is sought to be imposed decades before any foreseeable harm which could have any connection whatsoever to the act in question referable to the duty imposed.

298    A duty is being imposed decades before one knows whether there will be a cause of action. This is not to say that the risk of catastrophic global warming is only hypothetical and not real. It is, however, to say that a duty is being imposed under the law of torts not only before any damage occurs, but decades before one can even begin to assess the causal contribution, if any, of this decision to any harm.

299    This leads to the questions of foreseeability of risk and causation, control, vulnerability and reliance, and indeterminacy. For if one is to impose a duty, without damage, the foreseeable connection of act and any possible future damage will become critical. Thus the question of reasonable foreseeability of risk or harm and its place in the consideration of the salient features is framed by the attenuation brought about by this disaggregation. Likewise, questions of control, vulnerability and reliance, and indeterminacy will be affected by this disaggregation and its effect of temporal and geographic separation.

Foreseeability of contribution to risk and foreseeability of harm and causation: grounds 3(a) and (b)

300    The test of foreseeability is undemanding, but not without some demand. Whilst reasonable foreseeability is not a test of causation: Chapman v Hearse (1961) 106 CLR 112 at 122, the enquiry as to reasonable foreseeability has a causal element: The reasonable foreseeability is of the negligent act or omission causing or materially contributing to the harm.

301    In Shirt 146 CLR at 44, Mason J referred to foresight that carelessness may be likely to cause damage to the plaintiff. See also the expression of the matter by Glass JA in Shirt v Wyong Shire Council [1978] 1 NSWLR 631 at 639640; and Lord Reid in Dorset Yacht [1970] AC at 1027. As McHugh J expressed it in Crimmins 210 CLR at 32 [72]:

Basic to that determination [as to whether there was a duty of care], as always, is the question: was the harm which the plaintiff suffered a reasonably foreseeable result of the defendant’s acts or omissions?

302    The causative relationship of the act or omission and the harm is at the heart of the neighbourhood or proximity articulated by Lord Atkin in Donoghue v Stevenson [1932] AC at 580: “to avoid acts or omissions which you can reasonably foresee would be likely to injure your neighbour”.

303    That said, the question of foresight and the place of the foreseeable causation of harm is examined in a generalised inquiry at a higher level of abstraction than that of the causative relationship of breach and damage: Shirt 146 CLR at 47 and Vairy 223 CLR at 446447 [72]. Causation is relevant not because it must be proved that damage was (or will be) caused by the breach, but because it was foreseeable in the sense discussed at [131] above that damage may be caused.

304    It is, however, necessary to say something of causation. In this regard the facts challenged under ground 5 become relevant.

305    On no reading of the evidence, nor in the findings of the primary judge, could it be said that it was reasonably foreseeable that “but for” the decision and the likely release of the 100 Mt of CO2, the foreseeable harm would or may not occur. The difficulty in expressing the matter thus is brought about partly by the disaggregation of duty, breach, causation and harm and partly by the nature of the problem. The difficulty of applying the “but for” test is not however an end to any relevant causal inquiry: whether of causation itself, or its place in the relationship of foreseeable harm with act or omission. The “but for” analysis is not a comprehensive or exclusive test of causation. The decision of the High Court in March v E & M H Stramare Pty Ltd [1991] HCA 12; 171 CLR 506 and the judgment of Mason CJ (with which Toohey J and Gaudron J agreed) stands for the proposition that causation in the field of negligence is essentially a question of fact to be answered by reference to common sense and experience and one into which considerations of policy and value judgments necessarily enter and that the “but for” test is not definitive. Some aspects of Mason CJ’s judgment bear comment in the present context. First, causation is part of attribution of responsibility; it is not a philosophical or scientific inquiry: 171 CLR at 509. The rejection (at 171 CLR 509) of John Stuart Mill’s definition of cause as the sum of the conditions which are jointly sufficient to produce a given occurrence, is the foundation for holding a person responsible for damage when his or her wrongful conduct is one of a number of conditions sufficient to produce the damage. Secondly, the law recognises concurrent and successive causes in the establishment of “material contribution” of the wrongful conduct to the injury: that is any contribution that is not de minimis: March v Stramare 171 CLR at 514, citing Duyvelshaff v Cathcart & Ritchie Ltd (1973) 1 ALR 125 at 138 (Gibbs J); Tubemakers of Australia Ltd v Fernandez (1976) 10 ALR 303 at 310 (Mason J); Bonnington Castings Ltd v Wardlaw [1956] AC 613 at 620 (Lord Reid); and McGhee v National Coal Board [1973] 1 WLR 1 at 4, 6, 8 and 12. In further elaboration of “caused or materially contributed to”, Mason CJ said at 171 CLR 514:

Generally speaking, that causal connexion is established if it appears that the plaintiff would not have sustained his or her injuries had the defendant not been negligent: see ICIANZ Ltd v Murphy [(1973) 47 ALJR 122 at 127128]. But, as the decision in that case illustrates, it is often extremely difficult to demonstrate what would have happened in the absence of the defendant's negligent conduct.

306    Thirdly, Mason CJ recognised that value judgment has an important part to play in the causal question as part of attribution of liability: 171 CLR at 515. This question is echoed in the reasons of McHugh J in March v Stramare. Although McHugh J took a very different view to Mason CJ and favoured the “but for” test, the use by his Honour of a “scope of the risk” analysis after the application of the “but for” test can be seen to inform the kind of value judgment to which Mason CJ referred: whether the consequence of the conduct can be fairly regarded as within the risk created by the negligence: Roe v Minister of Health [1954] 2 QB at 85.

307    The fault lines of disagreement between the parties was as to what was reasonably foreseeable here: that a fractional increase in average global temperature merely makes contribution in a tiny way to the risk of harm, or was a prospective small, but material, contribution to the foreseeable harm.

308    One must be careful not to become tied up in language such that the debate becomes one of semantics. But the Minister essentially submitted that there could be no reasonable foreseeability of harm being caused or contributed to by the 100 Mt of emissions, rather she submitted that at its highest (from J[253], in the light of J[83]) there could only be reasonable foreseeability of a tiny contribution to the overall risk of exposure to harm.

309    The two models or analogies used by the parties in their respective arguments were: Bonnington Castings [1956] AC 613: material contribution to the harm (pressed by the respondents); and Fairchild v Glenhaven Funeral Services Ltd [2003] 1 AC 32: increasing the risk of harm (identified by the Minister as an impermissible approach because of its lack of acceptance in Australia as a legitimate causal connection).

310    It is necessary to say something of these two lines of cases. In Bonnington, Mr Wardlaw contracted pneumoconiosis while working as a dresser and fetter at Bonnington Casting’s foundry. He was exposed to silica dust from two sources – a pneumatic hammer and swing grinders. No dust extraction plant was known or practicable for use with the hammer, but such was available and actually used for the swing grinders. That equipment, however, was not kept free from obstruction in breach of factory regulations. The House of Lords held that Mr Wardlaw bore the onus of proof (overturning the earlier Court of Appeal decision in Vyner v Waldenberg Pty Ltd [1946] KB 50), but that he had discharged it. The disease was caused by the whole of the silica dust. As Lord Reid said at [1956] AC 621 “the source of his disease was the dust from both sources, and the real question is whether the dust from the swing grinders materially contributed to the disease”. Lord Reid said that any contribution that is more than de minimis is a material contribution. Lord Reid accepted (at 622) that much the greater portion of the dust came from the hammers (the innocent source); but concluded that some dust certainly, not a negligible proportion, came from the swing grinders (the impugned source). The evidence did not prove that the (innocent) hammer silica dust was substantially the sole course. The impugned dust was a material contribution: it “did help to produce the disease”.

311    One can see in the findings the place for robust common sense. But it is important to appreciate that the House of Lords rejected the proposition from Vyner that increasing the risk of harm was sufficient to reverse the onus of proof upon the putative tortfeasor.

312    The question, at least at the point of causal connection between breach and harm, cannot be left to a confluence of increase of risk and damage within the scope of risk created by the breach. The House of Lords in Bonnington overturned Vyner, which had on one view provided for the onus of proof to shift to the defendant if a breach of a safety provision were proved and the worker was injured in a way that could result from such a breach. In Betts v Whittingslowe [1945] HCA 31; 71 CLR 637 at 649, Dixon J said it was not necessary to inquire whether the Court in Vyner meant any more than that an inference of fact would be available in the circumstances of a breach and an injury within the scope of the risk provided for by the duty or statutory provision. Justice Dixon in Betts used the conjunction of breach of duty and harm that might be thereby caused as part of a process of drawing an inference. That is how Vyner has been understood in a number of Australian cases: Roads and Traffic Authority v Royal [2008] HCA 19; 245 ALR 653 at 662 [31] (Gummow, Hayne and Heydon JJ) and 689 [143] (Kiefel J); Gett v Tabet [2009] NSWCA 76; 254 ALR 504 at 556–557 [254]–[256] (Allsop P, Beazley and Basten JJA); and Flounders v Millar [2007] NSWCA 238 at [4]–[38] (Ipp JA, with whom Handley AJA agreed).

313    There are dicta in the High Court, however, that deal with the shifting of the evidential onus to a defendant in such circumstances: Bennett v Minister for Community Welfare [1992] HCA 27; 176 CLR 408 at 420–421 (Gaudron J); Chappel v Hart [1998] HCA 55; 195 CLR 232 at 238–239 [8]–[9] (Gaudron J), 257–259 [68]–[73] (Gummow J) and 273–274 [93(8)] (Kirby J); and Naxakis v Western General Hospital [1999] HCA 22; 197 CLR 269 at 278–279 [31] (Gaudron J).

314    McGhee v National Coal Board [1973] 1 WLR 1 raised this question of the difference between breach causing harm and breach increasing the risk of harm and the injury being within the scope of the risk. Mr McGhee contracted dermatitis. He worked in the defendant’s brickworks. He had worked in hot, dusty kilns. It was admitted that it was a breach of duty not to provide washing facilities for workers after they had finished work in the brick kilns. The medical evidence was to the effect that the dermatitis was caused by repeated minute abrasions of the outer skin followed by some damage to underlying cells of a kind not scientifically understood at the time. Profuse sweating over time softened the skin and made it easily injured. Dust would adhere to the skin in the kiln and exertion would cause abrasion. Washing was the only practical way of removing the danger. The man’s bicycling home while caked with sweat and dust would only exacerbate the position. Enough, however, was known about the disease to enable a conclusion that it was not like pneumoconiosis in that it was not caused by all the dust and sweat and exertion. It was clear, however, that not providing showers materially increased the risk of contracting dermatitis. The House of Lords found for Mr McGhee. There have been two interpretations placed upon the reasoning in the case. The first was as expressed by Lord Bridge speaking for the House of Lords in Wilsher v Essex Area Health Authority [1988] AC 1074 at 1090 that McGhee did not stand for any principle that the onus of proof was reversed; rather it was described as an example of a robust and pragmatic approach to inferential findings of fact. The second interpretation is reflected in the analysis of Lord Bingham, Lord Nicholls, Lord Hoffmann and Lord Rodger in Fairchild [2003] 1 AC 32 that in circumstances such as McGhee either a material increase in risk is an adequate “causal” link or no relevant distinction was to be drawn between materially increasing the risk and materially contributing to (that is causing) the injury.

315    I will come to Fairchild shortly.

316    The relationship between breach and causation and between causing an increase in risk, material contribution to harm and onus of proof, were expressly left open in Bennett 176 CLR at 416 where Mason CJ, Deane J and Toohey J said:

In order to answer that question, it might be necessary to consider the view that there is no real distinction between breach of duty and causation … as well as the question whether a failure to take steps which would bring about a material reduction of the risk amounts to a material contribution to the injury. These questions have been considered in Canada in the context of a possible shift in the onus of proof … but it seems that the problem still awaits final resolution. There is no occasion to consider it here.

(citations omitted)

317    The Court in Amaca Pty Ltd v Ellis [2010] HCA 55; 240 CLR 111 at 123 [12] noting the disavowal of the arguments by the plaintiff/respondent said:

The plaintiff expressly disavowed any argument in these appeals that demonstrating only that the exposure to asbestos increased the risk of contracting lung cancer was sufficient to establish causation. It was the plaintiff’s case in this Court, as it had been in the courts below, that she could succeed only if she showed that Mr Cotton’s exposure to asbestos had caused or contributed to (in the sense of being a necessary condition for) his developing lung cancer. This being the way in which the case was presented, it will be neither necessary nor appropriate to consider issues of the kind considered by the House of Lords in McGhee, Fairchild and Barker v Corus UK Ltd [[2006] 2 AC 572]. See also Sienkiewicz v Grief (UK) Ltd [in the Court of the Appeal] or by the Supreme Court of Canada in Resurfice Corp v Hanke [[2007] 1 SCR 333].

(citations omitted)

318    The issue in Amaca expressed at 240 CLR at 136 [68] was “whether one substance that can cause injury did cause injury”.

319    Thus, in Australia, until the matter is addressed differently by the High Court, the question can be seen to be covered by the judgment of the High Court in St George Club Ltd v Hines (1961) 35 ALJR 106 at 107:

In an action at law a plaintiff does not prove his case merely by showing that it was possible that his injury was caused by the defendant’s default.

320    Until the High Court says otherwise, causing an increase in the risk of harm occurring does not amount of itself to causing or materially contributing to the harm: Bendix Mintex Pty Ltd v Barnes (1997) 42 NSWLR 307 at 315–316; Seltsam Pty Ltd v McGuiness [2000] NSWCA 29; 49 NSWLR 262 at 278–285 [102]–[137].

321    The significant developments in causation in English law in Fairchild [2003] 1 AC 32, Barker v Corus UK Ltd [2006] 2 AC 572 and Sienkiewicz v Greif (UK) Ltd [2011] 2 AC 229 (in the United Kingdom Supreme Court) took place in the context of proof in relation to the disease mesothelioma. The disease was not like pneumoconiosis or indeed, asbestosis, which conditions or diseases can be seen to be caused by all the silica or all the asbestos dust. Mesothelioma was seen to be caused by some, perhaps even a small number of, elongated sharp blue asbestos fibres (not dust particles, often associated with the breakdown of white asbestos). The elongated sharp fibre, having been inhaled, burrowed through the lung from the inside to its outer wall where it may cause a cancerous tumour to grow in the outer cavity gradually over time (up to many years) filling the cavity with the tumour and denying the person the capacity to breathe. Exposure to asbestos, in particular blue asbestos, created and increased the risk of this occurring. The longer and the more intense the exposure, the greater the risk. But it was not possible to prove who might be responsible for what fibre, and when the inhalation took place. Such questions of proof were critical since if only one or a small number of fibres could cause the cancer, they could have been in the environment, they may not have been released into the atmosphere by any negligent conduct, there may have been more than one available actor who increased the risk, or the plaintiff himself of herself may have been responsible for some exposure.

322    Fairchild was concerned with a number of defendants each of whom had wrongfully exposed the plaintiff to asbestos. The House of Lords as a matter of policy decided that in limited circumstances, the negligent increasing of risk of harm was a sufficient proxy for material contribution. Policy based on justice and fairness was explicit in their Lordship’s reasoning: [2003] 1 AC at 66 [33] (Lord Bingham), 69 [40] (Lord Nicholls), 73 [56] (Lord Hoffmann) and 112 [155] (Lord Rodger). Chief Justice McLachlin had written on such questions in 1998 (BM McLachlin, “Negligence Law – Proving the Connection” in MJ Mullany and AM Linden (eds), Torts Tomorrow: a Tribute to John Fleming (LBC Information Services, 1998)). Chief Justice McLachlin in that paper said that the law of torts was an aspect of the rule of law and was concerned with righting wrongful conduct through rules of attribution of responsibility. If self-evident wrongs, as recognised by legitimate human perception, are incapable of reasonable attribution by the rules of causation leaving them un-remedied, people will be left with a sense of injustice. The rule of law in its application should not lead to a legitimate sense of injustice.

323    McGhee was closely examined by all their Lordships in Fairchild. All except Lord Hutton saw McGhee as standing for the proposition that materially increasing the risk of harm can be sufficient to amount to material contribution. The apparent change to principle was narrowly confined: [2003] 1 AC at 40 [2] and 55 [21] (Lord Bingham), 70 [43] (Lord Nicholls), 74 [61] (Lord Hoffmann), 91 [108] (Lord Hutton) and 118 [170] (Lord Rodger). Relevant to the confinement of the change to principle were the factors that all the risk was tortiously caused by all defendants, the causal element was singular (only exposure to asbestos) and medical science could never explain which of the negligent defendants had (in the traditional sense) caused the disease, but one of them had done so.

324    A difference of view arose in Barker v Corus as to how responsibility was to be apportioned. By majority, their Lordships found that the responsibility was several (not joint) and each defendant was liable to the plaintiff for its proportionate share of exposure to risk.

325    Fairchild was affirmed in Sienkiewicz in the Supreme Court for proof in mesothelioma cases with one defendant.

326    From the above, it is clear that it is not for an intermediate appellate court to say that for proof of causal connection in the causing of harm from global warming increasing the risk of such is a sufficient causal connection to establish liability.

327    That said, we are not dealing here with causation, we are dealing with duty: with the question at a more generalised level, or at a higher level of abstraction.

328    With the disaggregation of the inquiry in time and geography, no posited harm having yet occurred, the question is whether there should be imposed a duty. Taking the Minister’s submissions at their highest, that it can only be said that the emissions in question will increase the risk of harm by increasing in a small or tiny amount overall temperature denies, it is submitted, the imposition of a duty.

329    With respect, such does not necessarily follow. That it can never be proved that a small contribution to the risk was the contribution which caused the harm is no reason for not imposing a duty to act reasonably not to increase the risk if there is a real and not fanciful possibility that the contribution in question may cause or materially contribute to the harm. It may be an answer for an employer who negligently exposes workers to asbestos for a very short period of time to say that causation of harm cannot be proved. It is not an answer, however, to the imposition of a duty of care to say that the employer only ever intended to employ them for a short period of time in which they would be exposed to asbestos. That short exposure would increase their risk of contracting a deadly and painful disease by a tiny amount. At the level of duty, causation does not have to be proved, but some causal relationship between the act and the harm looking forward must be real and not fanciful. A contribution to the risk of harm occurring can be seen as part of that relationship. It may not be the whole of that causal relationship. At this level of generality and at this level of abstraction, the real question for the imposition of the duty or not is whether the increase in risk of the harm from this act can be seen to be so small that it is not reasonably foreseeable, that is, it is not real but is fanciful, that the act will or may have any causal relationship to harm to the Children in the future.

330    The answer to the above question depends upon the evidence and what findings can be made from them, in particular the factual errors in ground 5. In view of the answers to grounds 5(a)–(e) of the notice of appeal referred to above, the following can be said about reasonable foreseeability as to the consequences of the release of 100 Mt of emissions through combustion of the coal mined by the extension of the mine.

331    There is an approximately linear relationship between CO2 emissions and increases in the Earth’s global average surface temperature in the absence of non-linear feedback effects. On the evidence, the best possible outcome for global warming is the stabilisation of the global average surface temperature at about 2°C above pre-industrial levels. To achieve this best possible outcome, based on a carbon budget analysis by Professor Steffen, no new coal mines, or extensions of existing coal mines (and one can suppose, the extension of this mine), can be approved. The approval of the extension of the mine would therefore cause 100 Mt of Scope 3 emissions likely to be emitted outside the available carbon budget for stabilising the average global surface temperature at 2°C above pre-industrial levels. Above 2°C of global warming, there is a small (but not zero) risk of non-linear feedback processes occurring which will accelerate global warming and which could trigger a tipping cascade, resulting in an irreversible 4°C future world. The risk of these tipping elements triggering a tipping cascade increases exponentially (or more rapidly) towards a “significant risk” as the global average surface temperature rises towards 3°C (or even lower) above pre-industrial levels. The frequency and severity of bushfires and heatwaves, and therefore the risk of personal injury to the Children, increases as global average surface temperature increases.

332    From the above, one cannot say that there is no reasonable foreseeability of harm to the Children from the release of emissions caused by the combustion of the coal mined made available by the decision.

333    Thus, there is no reason to deny the duty of care by reference to there being no reasonable foreseeability of harm.

Control and conduct of third parties: grounds 3(c) and 4

334    It can be accepted that the Minister had, almost exclusive, control over any risk created by the approval of the extension of the mine: J[284]. But bearing in mind the nature of the harm as worldwide global climate catastrophe the control of the Minister of the harm is simply not present. The relevant concept of control is control over the harm, not control over the tiny increase in the risk.

335    The extent to which risk may be mitigated by countless others around the world and in particular by international agreement makes the concept of the Minister having relevant control non-existent, or extremely faint at best.

336    This makes the relationship between the Children and the Minister indirect and mediated by the intervening conduct of countless others around the world: cf Graham Barclay Oysters 211 CLR at 599600 [154] (Gummow and Hayne JJ).

337    That intermediation of others includes those who decide to buy and use the coal and their national governments who permit it within their obligations under international agreement to be responsible for Scope 3 emissions.

Vulnerability and reliance: ground 3(d)

338    The primary judge’s findings as to vulnerability at J[289][294] are in my view inadequate to characterise the Children as vulnerable in the relevant sense. Vulnerability is special vulnerability: Burnie Port Authority v General Jones Pty Ltd (1994) 179 CLR 520 at 551 (Mason CJ, Deane, Dawson, Toohey and Gaudron JJ). The Children are in the same position as everyone in the world who is or will be alive at the future times at which the harm is posited. The lack of relevant control in any relationship further undermines the notion of vulnerability.

339    The position is not affected by calling in aid an equitable notion of parens patriae. No constitutional or legal principle justified this. The Minister had powers under a statute to be exercised to take into account the protection of species, communities and water resources. There was no protective relationship between the Minister and the children of Australia and there was no foundation in the Act or in legal principle for the invocation of responsibilities affected by notions of parens patriae.

340    The Children have no reliance upon the Minister different to all other Australians. The reliance is the expectation of good government. That is or can be called general political reliance. It is not legal reliance general or otherwise.

Indeterminacy: ground 3(e)

341    The duty is owed to all children in Australia under 18 years of age and born at the time of the commencement of the proceeding. There is no reason in logic why the duty should not also be owed to the unborn.

342    The potential liability of the Minister is indeterminate in number and nature. That the duty concerns personal harm does not remove the consideration of indeterminacy. Rather it highlights the novelty of the duty.

343    There is a lack of proportionality between the tiny contribution to the increased risk of harm, the lack of control of the harm, and liability for all damage by heatwave, bushfires and rising sea level to the whole of the Australian population under 18, ongoing into the future.

Conclusion

344    Ultimately, however, all these considerations: lack of control over the harm (as opposed to the tiny contribution to the risk), the conduct of countless others around the world, the lack of any special vulnerability, and lack of reliance, are really only features or reflections of the essential problem for the respondents: the relationship that founds the duty is one between the government and the governed and lacks the relevant nearness and proximity necessary for the imposition of a duty of care. At one level of abstraction we all rely on an elected government to develop and implement wise policy in the interests of all Australians, in one sense especially the children of the country who are its future. That is not the foundation of the law of torts. It is the foundation of responsible democratic government.

345    To manufacture tortious concepts out of words of cases to transform the position of the governed into the vulnerability of the governed because of the power of the government involved in the development and application of policy is to fail to recognise the true relationship.

346    The relationship is not between neighbours in the legal sense amenable to responsibility in the private law of torts. The relationship is one of government and governed in connection with the protection of species, communities and water resources in a limited decision under the EPBC Act. The complaint made on behalf of all Australians aged under 18 is that the Minister should approach this decision by reference to the dangers of climate change and approach the decision on the basis of evidence that demands in a debate about breach the change (or justification) of government policy (Commonwealth and State) that underpinned the approval of the extension by the State of New South Wales, and that was based on earlier international agreement to which Australia was a party, and to which some, though not all, of the governments of the international buyers of the coal were parties. The litigation in effect seeks to impose a duty that will, at the point of examining breach, have investigated (in an adversarial context in a court and for the purpose of attributing potential personal liability) the question whether to act on the policy settings of the government on climate change was or would be negligent. In all the circumstances, even if my views on policy should not be seen as determinative in and of themselves, taken with the lack of coherence with the EPBC Act, the lack of control of the Minister over the harm, the tiny contribution of this decision to the overall risk of harm, the lack of relevant vulnerability, the intermediation of countless other people around the world, the indeterminacy of the liability, and recognising the true relationship for what it is: the governed and the governing, in which the latter seek to make the former personally responsible for a failure to change government policy with which the latter disagree powerfully, the duty of care should not be imposed.

Orders

347    I would allow the appeal. Before setting aside the orders 1 and 3 made by the primary judge and dismissing the application, I would hear the parties as to any other orders that may be necessary, especially ones that may involve the interests of the represented persons. It may be appropriate to ensure that they are not unduly prejudiced in the future by way of issue estoppel concerning the non-existence of a duty of care in this case. I would also hear the parties on costs. Orders should be made providing for the submission of draft orders by agreement, or failing agreement, the submission of competing draft orders.

I certify that the preceding three hundred and forty-seven (347) numbered paragraphs are a true copy of the Reasons for Judgment of the Honourable Chief Justice Allsop.

Associate:

Dated:    15 March 2022

REASONS FOR JUDGMENT

BEACH J:

348    Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540 is part of a cluster of High Court cases tending against the posited duty of care found by the primary judge. But three observations should be made at the outset.

349    First, Graham Barclay Oysters was not concerned with a scenario where it was suggested that the positive exercise of a statutory power might create or increase a risk of harm. That case involved a failure by the State of NSW and the relevant local council to act or take sufficient steps to deal with such a risk.

350    Second, the policy question that arose in Graham Barclay Oysters was of a different complexion to that with which we are concerned. I am not talking about subject matter. Rather, I am focusing on the nature of the question. Underpinning the NSW legislative regime in Graham Barclay Oysters was a policy choice of industry self-regulation. It was held to be in tension with that policy choice to find a duty of care. But that is not the type of policy question with which we are concerned.

351    Third, neither Graham Barclay Oysters nor any other High Court authority definitively rules out the posited duty of care. But the trend line is not in favour of supporting such a duty. I will return to Graham Barclay Oysters later, but I should begin by setting out my own frame of reference within which I have analysed the relevant issues.

352    In the context of considering whether a duty of care is owed with respect to the exercise of a statutory power to take reasonable care to avoid personal injury, three general scenarios may arise. Now these are by no means exhaustive. Further, one may shade into another. But they are a useful although not definitive conceptual framework within which to consider the relevant factors. Depending upon the scenario and the individual case, different emphasis may be given to different factors.

353    The first scenario is where there has been a failure to exercise a statutory power. The question that then arises is whether there was a duty of care to exercise the power.

354    The second scenario is where there has been an exercise of a statutory power but the repository of the power has not gone far enough. The question that then arises is whether there was a duty of care to exercise that or other analogous powers more than what was done.

355    Under these two scenarios, questions of policy, incoherency with the statutory regime and whether there was a special relationship between the plaintiff(s) and the repository of the statutory power may be highly significant matters in determining whether a duty of care was owed.

356    The third scenario is where there has been an exercise of a statutory power which has created or exacerbated a risk of harm. The question is whether there was a duty of care in the exercise of such a power to take into account that risk and to contemplate the interests of persons in the position of the plaintiff in that regard. In that scenario, questions such as policy and incoherency although relevant may have lesser significance. This third scenario mirrors what was recognised in Agar v Hyde (2000) 201 CLR 552 by the plurality (at [68]) that:

Further, from the earliest times, the common law has drawn a distinction between a positive act causing damage and a failure to act which results in damage…

357    To some extent such a distinction also echoes what was said by Mason J in The Council of the Shire of Sutherland v Heyman (1985) 157 CLR 424 at 458 to 460, although in the context before us it is being posited that the exercise of power to refuse approval may avoid or reduce the risk of harm.

358    Now the difficulty arises as to which scenario is applicable to the present context, which difficulty is exacerbated because we are only concerned here with whether there is a duty. Questions of breach, causation and damage have not yet arisen, let alone been determined.

359    Usually in litigation concerning whether a duty of care is owed, one has the advantage of dealing with the alleged completed tort, having identified the personal circumstances of an individual claimant, the precise harm suffered by that individual and the applicable factual causation mechanism linking the defendant’s alleged negligence with that individual’s harm. Actual factual causation can be addressed, and then compared with the appropriate retrospective hypothetical or counterfactual involved with a “but for” analysis; other more exotic mechanisms may also be more adequately addressed where the posited causal condition is neither a necessary nor sufficient condition, or is not the only sufficient condition. Further, a knowledge of and consideration of such questions feeds back into a more informed analysis concerning the questions of duty and breach. Causation cannot be assessed without identifying the act or omission said to constitute the breach. And breach depends upon identifying the duty of care and its scope.

360    Now when one is considering the completed tort, there is little difficulty in characterising the applicable scenario. But in the present case we do not have the completed tort. Nevertheless, we do know that the Minister has approved the project extension. And we do know that the respondents to the appeal contend, and hypothesised before the primary judge, that the positive act of approval would create or increase the relevant risk of harm. So, in substance one is dealing with a third scenario case. But as a matter of form, if one was just considering the statutory power in question before its exercise, one would be expressing the matter more neutrally. And in most third scenario cases, if one is formulating a duty concerning the exercise of power before it is exercised, one is usually going to express it in a neutral form. And so in the present case, the primary judge declared:

The [Minister] has a duty to take reasonable care, in the exercise of her powers under s 130 and s 133 of the Environment Protection and Biodiversity Conservation Act 1999 (Cth) in respect of referral EPBC No. 2016/7649, to avoid causing personal injury or death to persons who were under 18 years of age and ordinarily resident in Australia at the time of the commencement of this proceeding arising from emissions of carbon dioxide into the Earth’s atmosphere.

361    But in substance the characterisation that I have just given is apposite. It is the positive exercise to approve which is said to give rise to the risk of harm, rather than the failure to exercise or sufficiently exercise any statutory power.

362    There is another frame of reference question that I should deal with at the outset concerning the salient features approach to determining whether a duty of care is owed. I have no difficulty with it so long as it is appreciated that it is only a conceptual tool. But it should not distract from considering the broader questions, such as whether there is sufficient closeness and directness between the Minister and her exercise of statutory power to approve on the one hand, and any reasonably foreseeable effects on the respondents to this appeal or those that they represent (the claimant class) on the other hand by reason of such exercise. But it may be accepted that the salient features approach has useful and necessary flexibility. Such an approach recognises that the significance of particular features to a case are context dependent. Further, it implicitly accepts that there may be no bright-line between some of the features, such as questions of policy and incoherency with the statutory regime. Other examples are the overlap between the questions of sufficient closeness and directness and control of risk, and between sufficient closeness and directness and indeterminacy.

363    In summary and given the prevailing paradigm of authority binding upon me, I have reached the view that a lack of sufficient closeness and directness and its related partial inverse, namely, indeterminacy, are such as to deny the posited duty of care. Accordingly, I have allowed the appeal.

364    Now I should make two other points at the outset.

365    First, given that I have found against the primary judge concerning his identification of an implied mandatory consideration concerning human health and safety going to the exercise of the relevant statutory power, and given that such an identification underpins much of his discussion concerning various salient features, I have found it necessary to re-evaluate the relevant questions for myself as a free-standing exercise not built upon his Honour’s template.

366    Second, Allsop CJ has comprehensively set out the factual and procedural background, aspects of the primary judge’s reasons, the appeal grounds and the parties’ principal arguments. My own analysis proceeds on that foundation.

367    Let me begin with some science, and then I will turn to its contextual relevance concerning the question of reasonable foreseeability. After that I will dispose of the incoherency and policy questions, and then turn to what I see as the more difficult issues for the respondents to the appeal. Now I accept that I would normally start with the statute in question for a case of this type, but it is more convenient to leave such analysis to the incoherency and policy questions.

Climate modelling and the hypothesised tipping point

368    There were some unusual features about how the forensic case was run before the primary judge, not the least of which concerned the expert evidence before his Honour dealing with the effect of greenhouse gas (GHG) emissions on average global surface temperatures through until at least 2100. First, only one expert was called, being Professor William Steffen for the respondents. There was no witness served up by the Minister. Second, the Minister chose not to cross-examine Professor Steffen on any topic.

369    In my view, all of this was unsatisfactory given that there was one aspect of the science that appeared to be contentious concerning the tipping point hypothesis and the non-linear effects of GHG emissions when temperature reached 2°C above the relevant base line. One might have expected there to have been multiple experts produced on both sides going to this question, but the Minister chose a different strategic pathway and eschewed any meaningful technical presentation before the primary judge.

370    Of course, it may have been an appropriate course for the Minister not to challenge the central tenets of the projections of the Intergovernmental Panel on Climate Change (IPCC), particularly as they concerned various representative concentration pathways and their potential or likely consequences. Indeed many publications of the Commonwealth, its agencies and statutory authorities have confirmed or re-inforced such work and analysis. But in terms of Professor Steffen’s tipping point analysis at or around 2°C above the base line, that seems to have been, at least for the Minister, a contentious area. So much was demonstrated by the Minister’s submissions before us. But at trial, none of this was forensically challenged or the subject of a competing forensic case.

371    So, on a critical scientific question relevant to reasonable foreseeability, and an important part of the foundation upon which the posited duty of care was built, there was only one unchallenged expert. If nothing else, this all highlights the danger in just deciding a single question as to whether a duty of care was owed, rather than adjudicating upon the completed elements of the posited tort where one might expect at the trial of the latter a more complete and competing set of expert evidence. After all, a standard patent trial involving scientific complexity usually has 4 to 6 competing experts giving evidence in concurrent sessions. Was the present case as presented to the primary judge deserving of anything less given its broad-ranging consequences?

372    At all events I can only act on the evidence before the primary judge, although he had no greater advantage than do I in assessing what it reveals. After all, there was no cross-examination or indeed any need by the primary judge to adjudicate upon and resolve any competitive tension in the evidence. There was none. And nor was this a case where the primary judge had to put together a complex web of technical and voluminous evidence such that I should be reticent to assess this all for myself.

373    Before discussing Professor Steffen’s evidence, let me say something about modelling as discussed in the IPCC’s publications as it is relevant to understanding his evidence.

374    The findings of the IPCC Fifth Assessment Report were based on a set of scenarios called representative concentration pathways (RCPs), which were developed to be representative of possible future emissions and concentration scenarios published in the existing literature. They focused on the concentrations of greenhouse gases that could lead directly to a changed climate, and they included various pathways or trajectories of GHG concentrations over time to reach a particular radiative forcing at 2100. Radiative forcing is a measure of the energy absorbed and retained in the lower atmosphere. It is effectively a measure of the amount that the Earth’s energy budget is out of balance. It can be positive (heating) or negative (cooling) and is affected by GHG concentrations, aerosol concentrations, changes in land cover and natural drivers such as solar irradiance.

375    There were four pathways known as RCP2.6, RCP4.5, RCP6.0 and RCP8.5, with the numbers in each RCP referring to the amount of radiative forcing produced by greenhouse gases in 2100. It is relevant to note that the strength of drivers is quantified as radiative forcing in units of watts per square meter (W/m2), which is a measure of the change in energy flux caused by a driver.

376    In RCP2.6, radiative forcing peaks at 3 W/m2 before 2100 and then declines. In RCP4.5 and RCP6.0, radiative forcing is stabilised at 4.5 W/m2 and 6.0 W/m2 respectively after 2100. In RCP8.5, the radiative forcing is equal to or greater than 8.5 W/m2 in 2100. The RCPs have been developed using integrated assessment models as inputs to a wide range of climate model simulations to project their consequences for the climate system. These climate projections, in turn, are used for impacts and adaptation assessment. RCP2.6 is a stringent mitigation scenario. RCP4.5 and RCP6.0 are two intermediate scenarios. RCP8.5 is a scenario with very high GHG emissions. RCP2.6 is representative of a scenario that aims to keep global warming likely below 2°C above pre-industrial temperatures. As I have indicated, each RCP defines a specific emissions trajectory and subsequent radiative forcing, which trajectories are for the three main GHGs being CO2, CH4 and N2O. Of course I am focusing on CO2.

377    The CO2 concentrations in 2100 form part of the RCP data used as a model input, whilst the temperature increases are model outputs, with the likely range based on outputs using different models producing the following results:

Scenario

Atmospheric carbon dioxide concentrations in 2100 (used as input for most model simulations)

Temperature increase to 2081 to 2100 relative to 1850 to 1900 baseline

Average

Likely range

RCP2.6

421ppm

1.6 C

0.9 to 2.3 C

RCP4.5

538ppm

2.4 C

1.7 to 3.2 C

RCP6.0

670ppm

2.8 C

2.0 to 3.7 C

RCP8.5

936ppm

4.3 C

3.2    to 5.4 C

378    I note for completeness that subsequent to the hearing of the appeal, on 28 February 2022 the Working Group II contribution to the IPCC’s Sixth Assessment Report was published which reported on new pathway scenarios described as shared socio-economic pathways (SSPs) rather than RCPs. The SSPs were designed to complement the RCPs and deal with five different socio-economic development pathways. Now the Fifth Assessment Report addressed four RCPs. But these five SSP scenarios address radiative forcing of 1.9, 2.6, 4.5, 7.0 and 8.5 W/m2. The SSP scenarios corresponding to 4.5, 7.0 and 8.5 W/m2, being SSP 24.5, SSP 37.0 and SSP 58.5, show projections through to 2100 in terms of temperature increase that at the least do not detract from the force of Professor Steffen’s evidence based partly on his consideration of the RCPs published in the Fifth Assessment Report.

379    Let me now turn to Professor Steffen’s evidence and synthesise the following points before getting into the detail.

380    First, at some point in time in the future, the trajectories of increasing atmospheric CO2 concentration and increasing global average surface temperature will likely slow and then stabilise for a multi-decadal period. But the level at which this stabilisation occurs depends not only on human emissions of CO2 but also on feedbacks within the Earth system that strengthen or weaken the trajectories of CO2 and temperature. There will be a lag between the time that CO2 concentration is stabilised and the time that the global average surface temperature is stabilised. The lag time will be a minimum of multi-decadal up to a century or longer.

381    Second, what is meant by climate feedback is an interaction in which a perturbation in one climate quantity causes a change in a second quantity, which latter change leads to an additional change in the former. A negative feedback is where the initial perturbation is weakened by the consequent change. A positive feedback is where the initial perturbation is enhanced.

382    Third, there is an approximately linear relationship between human emissions of CO2 from all sources and an increase in global average surface temperature.

383    Fourth, there are non-linear effects. Now global average surface temperature is used as an indicator for the increasing heating of the Earth system including the atmosphere, ocean, land and cryosphere. Some of the impacts of this heating are approximately linear but many are not. In terms of non-linear impacts, some of these produce feedbacks that accelerate warming of the Earth system. For example, one has the melting of Arctic sea ice, which uncovers darker seawater, which absorbs more sunlight and accelerates warming. One has increasing drought in the Amazon basin, which increases fire frequency, leading to an increase in the emissions of CO2. One has the melting of the permafrost, which releases both CO2 and CH4 to the atmosphere, accelerating the warming.

384    As the global average surface temperature rises towards 2°C and beyond, the risk of such feedbacks being activated increases. Given that many of these feedback processes are linked, a global tipping cascade could form that takes the trajectory of the Earth system out of human control or influence and leads to a much hotter Earth. Professor Steffen described this as the Hothouse Earth scenario. Now there is no one definition of a tipping point. But as I understood Professor Steffen’s evidence, he was using it to describe the point where as a result of a very small change in an additive quantity a critical threshold could be reached in that quantity, beyond which a system could re-organise in an irreversible manner. So, and in the present context, a small but additional perturbation in external forcing could cause a qualitative and large change in a future state, which is irreversible. So, it is not just a bifurcation point in the equilibrium solution of a system, but a point of no return. Further, a system that is capable of reaching such a tipping point may be characterised by strong positive feedback in its dynamics, in other words the self-amplification of external forcing.

385    With these general points out of the way, let me elaborate in more detail on Professor Steffen’s evidence before further addressing the tipping point and non-linear effects question.

386    Professor Steffen was asked to choose several indicative points by reference to the level of temperature difference at the point in time when the increase in temperature difference flattens on the spectrum of possible future worlds. He was asked to start with the lowest level of temperature difference when it flattens, which is at the present date a real possibility, and to end with the highest level of temperature difference when it flattens, which is at the present date a real possibility. He proposed three possible climate futures.

387    His first future involved the stabilisation of global average surface temperature at or very close to 2°C above pre-industrial. In his view this was the best possible outcome that could be envisaged today, and is equivalent to the RCP4.5 scenario. The lowest RCP (2.6) would result in global average temperature rise of below 2°C by 2100, whilst the highest RCP (8.5) would lead to a temperature rise of 4°C or more by 2100. His second future involved the stabilisation of global average surface temperature at or very close to 3°C above pre-industrial. It is approximately equivalent to the upper end of the RCP6.0 envelope of temperature scenarios. His third future involved no stabilisation of global average surface temperature this century, with the 4°C above pre-industrial level breached late in the century with temperature continuing to rise into the 22nd century. This corresponds to the RCP8.5 scenario with its high and damaging impacts. In essence, this is the worst possible outcome that could eventuate if global cooperation on climate change breaks down and many nations continue on a pathway of high usage of fossil fuels. There is a risk that an RCP8.5 scenario could eventuate if the climate is driven into the Hothouse Earth scenario. The difference between RCP8.5 and the Hothouse Earth scenario is that in the RCP8.5 scenario, human emissions of greenhouse gases are the dominant driver of the temperature rise, whilst in the Hothouse Earth scenario, feedbacks within the Earth system play an important role in the ultimate temperature rise.

388    The characteristics and consequences of each of Professor Steffen’s futures are the following.

389    His first future has the following parameters. First, the temperature at stabilisation is approximately 2°C above the pre-industrial level. Second, cumulative emissions being the remaining carbon budget from 2021 onwards would need to be restricted to about 855 Gt CO2, assuming a 67% probability of meeting a 2°C target, and accounting for non-CO2 greenhouse gases and carbon cycle feedbacks. This equates to about 20 years of emissions at 2019 rates. Third, stabilisation would occur in the second half of this century. Fourth, stabilisation around 2°C would require a significant increase in national emission reduction targets and the corresponding policy, legislative and technological changes required to meet these targets. Achieving net-zero emissions by 2050 by all major emitting countries would be required to have a reasonable probability of stabilising the climate at a 2°C temperature level above pre-industrial. Fifth, feedbacks have been estimated for a 2°C forcing.

390    Let me at this point digress into evidence concerning carbon budgets and their relevance to the 2°C temperature target. The carbon budget framework has significant implications for fossil fuel reserves and resources, particularly for coal. In that respect, Professor Steffen adopted the following definitions of “reserves” and “resources” that had been used in McGlade, C and Ekins, P, “The geographical distribution of fossil fuels unused when limiting global warming to 2°C” (2015) Nature 517: 187 to 190. Reserves are defined as a subset of resources that are recoverable under current economic conditions and have a specific probability of being produced. Resources are the remaining ultimately recoverable deposits of fossil fuels that are recoverable over all time with both current and future technologies, irrespective of economic conditions. So, resources are all of the fossil fuels that are known to exist, and reserves are the subset of resources that are economically and technologically viable to exploit now.

391    McGlade and Ekins used a global carbon budget framework to assess the amount of fossil fuel reserves that could be exploited without transgressing a particular temperature target. For example, based on a 50% probability of meeting the 2°C temperature target, they estimated the global carbon budget for the 2011-2050 period to be 1,100 Gt CO2. This was somewhat higher than the budget of 855 Gt CO2 in the second parameter for the first future used by Professor Steffen. I should note that if a higher probability were adopted, say 67%, the remaining carbon budget would be much less than 1,100 Gt CO2, and even less coal and other fossil fuel reserves could be exploited.

392    McGlade and Ekins showed that if all of the worlds existing fossil fuel reserves were burned, about 2,860 Gt CO2 would be emitted, and about 2,000 Gt of these emissions would come from the combustion of coal. This level of emissions is about 2.5 times greater than the allowable budget for the 2°C temperature target. It was estimated that globally, 62% of the worlds existing fossil fuel reserves need to be left in the ground, unburned, to remain within the carbon budget. Meeting the 2°C carbon budget therefore means that not only must currently operating mines and gas wells be closed before their economic lifetime is completed, but also that no approved (but not yet operating) and no proposed fossil fuel projects or extensions of existing fossil fuel projects based on existing reserves could be implemented.

393    McGlade and Ekins also applied an economic analysis individually to the three types of fossil fuels being coal, oil and gas, and also to the various regions of the world that are major producers of fossil fuels. Again, the overall goal was to meet the 2°C temperature target. Under their regional analysis they concluded that over 90% of Australias existing coal reserves could not be burned to be consistent with a 2°C target, and no new coal resources could be developed. Furthermore, many existing coal extraction facilities would need to be closed before the end of their economic lifetimes.

394    Professor Steffen’s conclusion from his carbon budget analysis, based in part on the analysis of McGlade and Ekins, was that currently operating coal mines must be phased out as soon as possible and preferably no later than 2030, and that no new coal mines or extensions to existing coal mines can be allowed.

395    His second future involved the following elements. First, the temperature at stabilisation is approximately 3°C above the pre-industrial level. Second, implementation of current national climate policies around the world would lead to stabilisation around 3°C. Third, stabilisation would be achieved, at the earliest, late in this century but more likely early in the 22nd century. Fourth, the future emissions of CO2 consistent with a 3°C temperature are about 2,600 Gt from 2021 until net zero emissions are achieved.

396    Now I note Professor Steffen’s caveat that there is a significant risk that his second future is not accessible in that it may not be possible to stabilise the Earth system at a 3°C level above pre-industrial. The reason is that many feedback processes will be activated by a 3°C or even lower temperature rise, with a consequent significant risk that a tipping cascade will be activated, taking the global average surface temperature beyond 3°C and into a Hothouse Earth scenario. According to Professor Steffen, there is a very significant risk that strong non-linear feedbacks will be activated by a 3°C warming, leading to his assessment that the 3°C stabilisation scenario may not be possible. As support for this assessment, the IPCC in 2018 estimated that there is a moderate risk of triggering these feedbacks already at a 2°C temperature rise, and this risk will undoubtedly rise with a 3°C temperature forcing on the Earth system. I should also note that Professor Steffen gave evidence that there was a risk that a 2°C temperature rise could trigger a Hothouse Earth trajectory, but the probability of such a scenario is much lower for a 2°C temperature rise than for a 3°C temperature rise.

397    His third future involved the following elements. First, the global average surface temperature continues to rise throughout the 21st century with no stabilisation until sometime in the 22nd century and at a temperature of at least 4°C above pre-industrial and probably higher. Second, stabilisation would be dictated by Earth system processes and not by human actions. Third, the time of stabilisation is difficult to predict but would occur sometime in the 22nd century or beyond. Fourth, human CO2 emissions are less relevant for this scenario, as once a tipping cascade is initiated, the internal dynamics of the Earth system comprise the controlling factor, with CO2 emissions from feedbacks such as permafrost melt and forest dieback becoming an important source of CO2. Feedbacks are the key feature of this third future, with a set of interacting Earth system feedbacks driving a cascade that would drive the Earth system to a much hotter state.

398    His Honour accepted Professor Steffen’s evidence and that his first future is the best possible outcome that can be envisaged today.

399    Now the risk of triggering a tipping cascade increases with the rise in global average surface temperature. More specifically, a cascade of tipping elements could be initiated at a rise in global average surface temperature of around 2°C. According to Professor Steffen, his assessment of tipping point behaviour is that there is a small (but non-zero) probability of initiating a tipping cascade at or around a 2°C temperature rise.

400    Professor Steffen variously referred to the temperature increase associated with his first future as being “at, or very close to, 2°C”, “approximately 2°C above the pre-industrial level”, “a 2°C target”, “the 2°C temperature target”, “around 2°C”, “a 2°C temperature level”, “about 2°C”, as well as “approximately equivalent to, or slightly higher than, the upper Paris accord target of “well below 2°C””. So much was acknowledged by the primary judge in concluding (at [89]) that:

Read in context, the better view is that when Professor Steffen was referring to the stabilised average global temperature for his “Scenario 1” he meant 2°C or slightly lower but not “well below 2°C” and not the upper target of the Paris Agreement.

401    Further, the primary judge found that at a stabilised global average surface temperature above 2°C, there is an exponentially increasing risk of the Earth being propelled into an irreversible 4°C trajectory. In my view, this finding was supported by the evidence of Professor Steffen. Now he also gave evidence about the approximately linear relationship between human emissions of CO2 from all sources and the increase in global average surface temperature, but as I have noted there are also non-linear impacts of this relationship whereby the activation of feedback processes may accelerate warming of the Earth system. It was these non-linear impacts to which the primary judge referred when reference was made to an exponentially increasing risk of the Earth being propelled into an irreversible 4°C trajectory.

402    Consistently, the evidence was that there is a small non-zero probability of initiating a tipping cascade at or around a 2°C temperature rise, but that whilst there is a risk that a 2°C temperature rise could trigger a Hothouse Earth trajectory the probability of such a scenario is much lower for a 2°C temperature rise than for a 3°C temperature rise. Further, as the global surface temperature rises towards 2°C and beyond, the risk of such feedbacks being activated increases.

403    In summary, the primary judge found that there is a real risk that even an increase in global average surface temperature of around or more than 2°C above pre-industrial levels may trigger a 4°C future world, based upon the risk of initiating a tipping cascade, which could in turn trigger the Hothouse Earth scenario. In my view it was open to the primary judge to construe the risk in the manner that his Honour did.

404    Now the Minister says that the fact that the risk of triggering a tipping cascade increases beyond 2°C says nothing about the sensitivity of the tipping cascade mechanism, and she says that the evidence relied upon by the respondents does not justify the primary judge’s finding that even a very small increase in global average surface temperature may trigger a 4°C future world. I disagree. Professor Steffen gave evidence which supported such a finding.

405    Further, the Minister says that the primary judge erred in finding that a decision to approve the extension project would cause an increase in CO2 emissions of 100 Mt above the CO2 emissions that would otherwise occur. Further, the Minister says that the primary judge erred in finding that if the extension project were to proceed, any CO2 emissions resulting from burning of coal extracted through that project would be outside the emissions contemplated by the carbon budget necessary to achieve a target of 2°C above pre-industrial levels. But the respondents were not required to establish a counterfactual regarding what would happen to global CO2 emissions if the extension project were not approved. Further, the Minister chose to lead no evidence at trial in support of her posited counterfactual.

406    Further, the Minister contended that it should be inferred that the 100 Mt of CO2 would likely be emitted in accordance with the Paris Agreement. But there is no sufficient basis for that inference. The Minister relied upon little else than speculation, in circumstances where the evidence showed that at least one of the potential consumers of the coal is not even a signatory to the Paris Agreement.

407    Now Professor Steffen gave uncontroverted evidence that in order to stabilise global average surface temperatures at or around 2°C, over 90% of Australia’s existing coal reserves should not be burned, currently operating coal mines should be quickly phased out, and no new coal mines or extensions to existing coal mines should be allowed. And in that context it is clear from Professor Steffen’s evidence that he considered the scope 3 emissions expected to result from the extension project to be incompatible with the carbon budget associated with his first future. Tellingly, the Minister led no evidence to suggest that the 100 Mt of CO2 from the extension project would likely be burnt within the carbon budget that Professor Steffen identified as being compatible with his first future. Indeed, the Minister did not even cross-examine Professor Steffen. I have no sympathy for her belated stone throwing before us.

408    Quite correctly, the primary judge observed that there was not sufficient evidence on which he could conclude that there was not a real prospect of the 100 Mt of CO2 being burnt outside the available fossil fuel budget necessary to meet a 2°C target.

409    Let me make one other point concerning the factual significance of the scope 3 emissions. In the NSW Department’s assessment report, the context and detail of which I will discuss later, the significance of the scope 3 emissions was discussed. Now putting to one side that the scope 3 emissions would not contribute to Australia’s nationally determined contribution under the Paris Agreement, conveniently because the coal was to be combusted off-shore and by the importing countries or entities based there, it was nevertheless said (at [692] and also in the introduction at p xiv):

The Department acknowledges that the Scope 3 emissions from the combustion of product coal is a significant contributor to anthropological climate change and the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

410    The Independent Planning Commission agreed with this assessment. It said (at [221]):

The Commission agrees with the Department’s statement in paragraph 209 above and acknowledges that Scope 3 emissions from the combustion of product coal are a significant contributor to anthropological climate change and that the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

411    So in terms of the contribution of the scope 3 emissions from the extension project, which is after-all what was being discussed, their significance as a contributor was not being denied. Now admittedly these paragraphs were expressed a little vaguely in terms of whether they were referring to just the extension or the entire project (as extended), but they are not unhelpful to the respondents to the appeal.

412    I need not linger further on his Honour’s evidence based findings. Apart from what I have said, I agree with Allsop CJ’s analysis concerning appeal ground 5.

413    Let me then turn more directly to the question of reasonable foreseeability. In this area there are two principal but related themes of reasonable foreseeability and causation. But as the present debate concerns the issue of duty only, extensive discussion of causation is not necessary.

Reasonable foreseeability

414    It is convenient to begin with some general observations, recognising of course that reasonable foreseeability is not a free-standing issue in the sense that it does relate to and must also take its context from other questions such as sufficient closeness and directness (Graham Barclay Oysters at [86] per McHugh J).

415    First, the requirement or condition of reasonable foreseeability is not and cannot be a test of causation. After all, it is forward looking and serves a different function depending upon the element of the tort that one is considering. But to the extent that one is ever looking forward on likely causation, say, in the context of seeking an injunction for a threatened breach of duty, as distinct from a backward looking analysis once damage has occurred, there are areas of overlap with reasonable foreseeability to the extent that one is hypothesising on the risk of injury if hypothesised action or inaction eventuates.

416    But as was said in Chapman v Hearse (1961) 106 CLR 112 at 122 per Dixon CJ et al:

As we understand the term “reasonably foreseeable” is not, in itself, a test of “causation”; it marks the limits beyond which a wrongdoer will not be held responsible for damage resulting from his wrongful act.

417    Second, it is important to bear in mind that reasonable foreseeability involves a more general inquiry at the duty stage than at the breach stage. At the duty stage, in addressing reasonable foreseeability, one is considering whether it is reasonably foreseeable as a possibility that careless conduct of any kind on the part of the defendant may result in damage of some kind to the person or property of the plaintiff. But at the breach stage, the lens is narrower. One is considering whether it is reasonably foreseeable as a possibility that the kind of carelessness that the defendant is charged with may result in damage of some kind to the person or property of the plaintiff. And as part of that calculus one should also consider whether there were or are means of obviating that possibility available and which would have been adopted by a reasonable defendant.

418    Now the primary judge said (at [257]):

In sum, this is a case where the foreseeability of the probability of harm from the defendant’s conduct may be small, but where the foreseeable harm, should the risk of harm crystallise, is catastrophic…

419    But considerations such as the foreseeability of the probability of harm and magnitude of the consequences are really for the breach stage. At that stage the calculus described in Wyong Shire Council v Shirt (1980) 146 CLR 40 at 47 and 48 per Mason J is applied, which, although undemanding in an absolute sense, is more demanding than at the duty stage. But the fact that his Honour went further than he needed to in no way undermines his principal conclusion that reasonable foreseeability was shown in the present case in the duty context.

420    Third, it might be said here that the Minister had actual knowledge of the risk, not just that it was reasonably foreseeable. She knew the likely consequences of the grant of approval, that is, that coal was likely to be mined, sold, exported and burnt overseas thereby producing the scope 3 emissions. And flowing from her knowledge of these likely consequences, she knew of the risk of harm that I have previously identified. Moreover, even if it be accepted that the posited harm to the claimant class was indirect, in the sense that there were many hypothesised links in the causal chain, that does not assist the Minister on the question of reasonable foreseeability. After all, she did foresee or could reasonably foresee all realistic intervening events after the approval through until likely harm.

421    Fourth, I have referred to the two dimensions of reasonable foreseeability in the context of duty and breach. But of course remoteness is a slightly different concept. Remoteness concerns whether the kind of damage suffered was foreseeable as a possible outcome of the kind of carelessness charged against the defendant. In the present context this can be put to one side.

422    Let me turn then to the primary judge’s consideration.

423    In my view, his Honour was not incorrect to find that reasonable foreseeability had been established.

424    Now the primary judge acknowledged that it was necessary to consider whether the foreseeable steps in the relevant chain of events, individually or collectively, by reason of the complexity of their interactions were such as to deny reasonable foreseeability. But the Minister says that what needed to be foreseeable was that the Minister’s act was likely to cause the identified harm, as opposed to making a small contribution to the conditions that, together with acts of other actors, might cause that harm unless mitigated by countervailing action. I disagree with that oppositional perspective. The latter perspective, if made out, is sufficient to establish reasonable foreseeability.

425    A risk of injury flowing from a defendant’s action need only be real to be reasonably foreseeable, and it is no barrier to the existence of a duty of care that the probability of injury occurring is remote (Overseas Tankship (UK) Ltd v Miller Steamship Co Pty (The Wagon Mound (No 2)) [1967] 1 AC 617 at 642 per Lord Reid), provided that it is “not far-fetched or fanciful” (Shirt at 46).

426    Now the Minister says that the primary judge ignored the difference between a case where it is possible to foresee that a particular action or inaction may combine with other circumstances to cause harm, and a case where it is not possible to predict whether a defendant’s action or inaction will make any difference at all to whether harm will arise.

427    The Minister says that in the former case it is possible to foresee a causal chain that is contingent on the occurrence of other events in addition to the defendant’s action or inaction, although if those contingencies become too complex, the risk of harm may be too remote to be real. But contrastingly, in the latter case, so the Minister says, there is no foreseeable causal connection between a defendant’s action or inaction and the harm.

428    But in my view, if this distinction be apposite, the former case was established on the evidence in any event.

429    Further, the Minister says that reasonable foreseeability does not embrace a risk of harm that exists whether or not the extension project proceeds, such that the element of “but for” causality is missing, being a risk of harm that the evidence does not establish would be increased at all by any additional GHG emissions from the extension project. But I disagree. This is all too narrow and a conflation. We are looking forward, rather than addressing causation and accepted factual causation tests such as the strong necessary condition formulation.

430    I should say that the Minister may not have been alone in blending reasonable foreseeability and causation.

431    His Honour accepted that the risk of harm to the claimant class caused by climate change does not just increase linearly as CO2 emissions accumulate, rather there is a risk that as the global average surface temperature increases from 2°C to 3°C above the pre-industrial level, that a tipping cascade may trigger a 4°C trajectory. And having regard to this risk, an additional 100 Mt of CO2 emissions is capable of constituting a material contribution to the risk that the claimant class will be exposed to harm.

432    And his Honour seems to have implicitly reasoned that in deciding whether it is reasonably foreseeable that harm of a particular kind will be caused by a defendant’s action or inaction, causation is to be assessed not on the usual “but for” basis, but by one of the less stringent tests that have been adopted for limited purposes in the tort of negligence. So, one has the test adopted in Bonnington Castings Ltd v Wardlaw [1956] AC 613 at 621 per Lord Reid, where causation can be proved in certain circumstances where negligent conduct has materially contributed to the occurrence of the harm. And one has the test adopted in Fairchild v Glenhaven Funeral Services Ltd [2003] 1 AC 32, where causation can be proved where negligent conduct has resulted in a material increase in the risk of harm occurring.

433    Now putting to one side for the moment that one should not conflate reasonable foreseeability and causation, the Bonnington Castings approach has no application. The present case is more apposite for a Fairchild analysis. Let me say something more about Bonnington Castings. I will do so first by addressing the tipping point thesis. It is said that the CO2 from the scope 3 emissions creates the risk of reaching the tipping point. If the tipping point is reached, there is a risk that the non-linear effects will cause the temperature to proceed from 2°C to 4°C above the base line by 2100. If that occurs, the personal injury to members of the claimant class or some of them may occur. But once one appreciates these elements one can see that the Bonnington Castings scenario has little to do with causation concerning the tipping point thesis.

434    The temperature of 4°C above the base line may be compared with the totality of the noxious dust inhaled by the worker. After all, it is that condition, that is, 4°C above the base line or the totality of the dust, which is said to cause the injury, although even that is simplistic; there are additional steps between temperature increase and harm, particularly if one is talking about intermediate events such as bush-fires or floods.

435    In Bonnington Castings the noxious dust was made up of two component sources, the totality of which was inhaled and caused harm. And in that context, the dust from the swing grinders was a material contribution to the disease. But in the case before us the indivisible condition which is said to cause harm is the temperature, not the CO2 emissions. Here the scope 3 emissions do not, with other emissions, directly cause the 4°C above the base line. Rather, the scope 3 emissions increase the likelihood or risk of producing the tipping point. And if that risk occurs, then there is a risk that the 4°C above the base line will occur.

436    That is no analogue with the Bonnington Castings scenario. One is only dealing with an increase in risk. Moreover, the additional CO2 molecules caused by the scope 3 emissions cannot be equated with the dust in Bonnington Castings. The CO2 molecules themselves do not directly cause or contribute to harm. It is rather their effect on increased temperature, which temperature ultimately causes the harm. The more appropriate analogue is Fairchild dealing with a material increase in risk from the scope 3 emissions. But that has not yet been accepted as a test for causation in Australia. But in the present context that does not matter. We are presently only concerned with reasonable foreseeability and not causation. Further, we are not here concerned with apportionability, so Barker v Corus (UK) Plc [2006] 2 AC 572 may be put to one side.

437    One can make similar points concerning the non-tipping point causation thesis dealing with the linear correlation between increasing CO2 emissions and temperature. But again, this is still not a Bonnington Castings scenario. Of course, many players contribute to the total CO2 emissions (like the dust), but it is the temperature that is then directly produced from the combined CO2 emissions, not the harm. Contrastingly, in Bonnington Castings the harm was produced from the combined dust, with no intermediate step like the temperature as in our case. Again, even the linear correlation thesis is more a Fairchild analogue. I note that the NZ Court of Appeal was referred to Fairchild rather than Bonnington Castings in Smith v Fonterra Co-Operative Group Ltd [2021] NZCA 552 at [105] to [113] in the context of an argument concerning material contribution to risk rather than material contribution to harm.

438    But I do not need to linger on which of Bonnington Castings or Fairchild is the appropriate analogue. The present context concerns the existence of a duty rather than causation. But whilst I am here, let me make a suggestion. No doubt it might be said that the “but for” test concerning factual causation is inapposite in the present context and that the NESS (necessary element of a sufficient set) test advocated by Professor Richard Wright and others is more suitable, particularly when dealing with a posited causal condition that is neither necessary nor sufficient. So, Wright said in his chapter “The NESS Account of Natural Causation: A Response to Criticisms” (Perspectives on Causation (R. Goldberg (Ed)), Hart Publishing, 2011) at 304 and 305:

Moore acknowledges the validity and usefulness of the NESS account’s ability to identify as causes conditions that were neither strongly necessary nor independently strongly sufficient, while also noting and criticising the consequent increased proliferation of causes. David Fischer also objects to the proliferation of causes and questions the validity of recognising trivial contributions as causes – eg, a teaspoon of water added to a flooding river or a match added to a raging forest fire. Yet the teaspoon of water and the match contributed to and are part of the flood and forest fire, respectively. What if the same flood or fire were caused by a million (or many more) different people all contributing a teaspoonful of water or a single match? Denying that any of the teaspoonfuls or matches contributed to the destruction of the property that was destroyed by the flood or fire would leave its destruction as an unexplained, non-caused miracle. As a pure matter of causation, it cannot possibly matter whose hands supplied the different bits of water, flame or fuel. What is driving the intuition of no causation is the judgment regarding attributable responsibility, which is especially brought to mind if the question is posed as ‘Did the teaspoon of water or match destroy the property?’ rather than ‘Did the teaspoon of water or match contribute, even if only extremely minimally, to the flood or fire that destroyed the property?’ What is generally agreed upon is that the trivial contributor should not be held liable when her contribution was trivial in comparison to the other contributing conditions and was neither strongly necessary nor independently strongly sufficient for the injury at issue, but this is a normative issue of attributable responsibility rather than causal contribution.

439    Professor Jane Stapleton has also written on the problem of dealing with a non-necessary and non-sufficient factor and has posited a test that “a factor is a factual cause if it contributes in any way to the existence of the phenomenon in issue” (J. Stapleton, “Factual Causation” (2010) 38 Federal Law Review 467 at 475 to 477). She said:

There are many other instances where the contribution of the defendant’s breach of duty to the outcome in issue is neither necessary nor sufficient. Moreover, there may be situations where the extent of a factor’s contribution to an outcome is hard for the honest plaintiff to quantify so that, while he can show it contributed, he is unable to show that this contribution was necessary or sufficient for the outcome. This is typically the situation in pollution cases and in cases where the plaintiff has made a decision after taking into account a number of considerations. For example, consider the following:

    Three factories each independently and in breach of duty discharge oil into a bay. By a regulatory standard, fishing in the bay is forbidden if the concentration of oil is greater than a particular level. By the time the pollution is detected the concentration level far exceeds this regulatory standard. When the ban is triggered this causes grave economic injury to local commercial fishermen who are unable to quantify the contribution each factory’s discharge made.

...

What [this example] show[s] is that a non-necessary non-sufficient factor may contribute to the existence of a phenomenon. It does so by forming part of an undifferentiated whole that operates to bring about the existence of the phenomenon: … the discharge by one factory formed part of the total concentration that exceeded the regulatory standard; … The examples also illustrate how the law is interested in the possibility of imposing liability on such a non-necessary non-sufficient factor. To do so the law needs to designate a notion of factual causation that is wide enough to accommodate these contributions. This is why the ‘but-for’ test of factual cause is under-inclusive and why courts grasp at vague undefined labels such as ‘substantial factor’ and ‘material contribution’ in their attempt to recognise a non-necessary non-sufficient factor as a ‘cause’

It is only the seductive simplicity of the ‘but-for’ test that distracts courts from enunciating an appropriately wide statement of the relation of ‘factual cause’: namely, that a factor is a factual cause if it contributes in any way to the existence of the phenomenon in issue.

Two associated points need to be mentioned here. First, notice that the … examples given so far involve an indivisible outcome and the discussion has been about a non-necessary non-sufficient contribution to the existence of that indivisible outcome: if the breach did make such a contribution and if all other elements of the cause of action are established, the defendants are jointly and severally liable for the entire indivisible outcome. This should be sharply distinguished from cases where the issue is whether a factor is a factual cause of only part of a divisible outcome, say pneumoconiosis or deafness: here if all elements of the cause of action are established, the defendant is only liable for that part of the divisible outcome in relation to which the breach was shown to be a factual cause (unless a special proof rule is available…).

440    It seems to me that the common law is going to have to evolve to deal with scenarios such as the present, including adopting such considered suggestions to deal with factual causation. But all of this analysis concerns causation. It has little to do with reasonable foreseeability in the context of whether a duty is owed.

441    In my view, the primary judge’s conclusion on reasonable foreseeability is sustainable for the reasons that I have given. And the preponderance of authority does not demand that a legally acceptable pathway to ultimately demonstrating causation must be used in any reasonable foreseeability analysis. Moreover, who knows what the legally acceptable factual causation test will be in 80 years when a fully formed tort is likely to arise, if at all? And indeed, who knows what the science will show in 80 years in terms of factual causation?

442    It is now convenient to turn to the questions of incoherency and policy. But before getting into the detail it is necessary to say something concerning the legislative framework and the processes envisaged by and that have been engaged in under such frameworks.

The legislative framework

443    It is appropriate to begin with the Commonwealth regime before turning to the NSW regime.

The Commonwealth regime

444    An object of the Environment Protection and Biodiversity Conservation Act 1999 (Cth) (the Act) is “to provide for the protection of the environment…” (s 3(1)(a)), especially concerning matters of national environmental significance. In s 528, “environment” is defined such that:

environment includes:

(a)    ecosystems and their constituent parts, including people and communities; and

(b)    natural and physical resources; and

(c)    the qualities and characteristics of locations, places and areas; and

(d)    heritage values of places; and

(e)    the social, economic and cultural aspects of a thing mentioned in paragraph (a), (b), (c) or (d).

445    Limb (e) of the definition is of significance in the present context, particularly reference to “social” or “economic” aspects. Clearly, human health and safety is not expressly referred to, although it is impliedly embraced by limb (e).

446    Another object of the Act is “to promote ecologically sustainable development…” (s 3(1)(b)). And in terms of the cognate principles, s 3A provides:

The following principles are principles of ecologically sustainable development:

(c)    the principle of inter-generational equity—that the present generation should ensure that the health, diversity and productivity of the environment is maintained or enhanced for the benefit of future generations;

(d)    the conservation of biological diversity and ecological integrity should be a fundamental consideration in decision-making;

447    The reference to “inter-generational equity” is not unhelpful to the respondents. Further and interestingly, there is reference to “the health…of the environment”. And environment is defined to include “ecosystems and their constituent parts, including people…”.

448    Part 3 deals with the requirements for environmental approvals, although it does not deal with the environment or environmental concerns generally. Part 3 is divided into Divisions 1 and 2.

449    Division 1 deals with 9 enumerated matters which deal with requirements relating to “matters of national environmental significance”; there is no definition of that latter phrase. I do not need to discuss these, save to note “Subdivision C – Listed threatened species and communities” (ss 18, 18A and 19), which is relevant to the approval of the extension project. Further, “Subdivision FB - Protection of water resources from coal seam gas development and large coal mining development” (ss 24D and 24E) is also relevant to the approval of the extension project. These provisions are in the form of prohibitions on action, with carve-outs or exclusions, inter-alia, if an approval under Pt 9 has been given. Clearly, neither Subdivision C nor Subdivision FB either expressly or directly address climate change or CO2 emissions.

450    Division 2 deals with the protection of the environment from proposals involving the Commonwealth, including actions involving Commonwealth land, activities of Commonwealth agencies and analogous matters. I do not need to discuss these further.

451    Part 4 should be noted. It deals with cases in which environmental approvals are not needed. Division 1 deals with actions covered by a bilateral agreement between the Commonwealth and a State or Territory, where the relevant action is declared by such an agreement not to need approval. I can put this to one side. Divisions 4, 5 and 6 can also be put to one side. But Divisions 2 and 3 are of interest, principally because of the declaratory powers given to the Minister that can be characterised as either quasi-legislative or involving policy questions to such an extent as might negate the imposition of a duty of care with respect to their exercise or non-exercise. They may be contrasted with the powers under ss 130, 133 and 136 that I will come to later.

452    So, s 33(1) provides:

The Minister may declare in writing that actions in a class of actions specified in the declaration wholly or partly by reference to the fact that their taking has been approved by the Commonwealth or a specified Commonwealth agency, in accordance with a management arrangement or authorisation process that is an accredited management arrangement or an accredited authorisation process for the purposes of the declaration, do not require approval under Part 9 for the purposes of a specified provision of Part 3.

453    This must be read in the context of s 32 which provides:

A person may take an action described in a provision of Part 3 without an approval under Part 9 for the purposes of the provision if:

(a)    the action is one of a class of actions declared by the Minister under section 33 not to require approval under Part 9 for the purposes of the provision (because the action is approved in accordance with an accredited management arrangement or an accredited authorisation process for the purposes of the declaration); and

(b)    the declaration is in operation when the action is taken; and

(c)    one of the following applies:

(i)    in the case of an accredited management arrangement—the management arrangement is in operation under a law of the Commonwealth identified in or under the declaration;

(ii)    in the case of an accredited authorisation process—the authorisation process is set out in a law of the Commonwealth, and the law and the authorisation process are identified in or under the declaration; and

(d)    the action is taken in accordance with the accredited management arrangement or accredited authorisation process.

454    Further, s 34A provides:

The Minister may make a declaration under section 33 only if the Minister is satisfied that the declaration:

(a)    accords with the objects of this Act; and

(b)    meets the requirements (if any) prescribed by the regulations.

455    Clearly, such powers are quasi-legislative and involve matters of policy to such a degree as would be inconsistent with a co-existent duty of care.

456    Further, s 37A and 37B provide:

37A Making declarations that actions do not need approval under Part 9

Subject to Subdivisions C and D, the Minister may, by legislative instrument, declare that an action or class of actions specified in the declaration, wholly or partly by reference to the fact that the taking of the action or class of actions is in accordance with a bioregional plan, do not require approval under Part 9 for the purposes of a specified provision of Part 3.

37B General considerations

(1)    In deciding whether to make a declaration under section 37A, the Minister must consider the following, so far as they are not inconsistent with any other requirements of this Subdivision:

(a)    matters relevant to any matter protected by a provision of Part 3 that the Minister considers is relevant to the action or class of actions to which the declaration relates;

(b)    economic and social matters.

(2)    In considering those matters, the Minister must take into account the principles of ecologically sustainable development.

(3)    The Minister must not make a declaration under section 37A in relation to an action or class of actions and a provision of Part 3 if the Minister considers that the action, or an action in the class, if taken, would have unacceptable or unsustainable impacts on a matter protected by the provision.

457    This is also to be read in the context of s 37 which provides:

A person may take an action described in a provision of Part 3 without an approval under Part 9 for the purposes of the provision if:

(a)    the action is an action, or one of a class of actions, declared by the Minister under section 37A not to require approval under Part 9 for the purposes of the provision (because the taking of the action is in accordance with a particular bioregional plan); and

(b)    the declaration is in operation when the action is taken; and

(c)    the action is taken:

(i)    in the bioregion to which the plan applies; and

(ii)    in accordance with the plan.

458    Clearly, the s 37A power is legislative or quasi-legislative. Interestingly, the phrase in s 37B(1)(b) being “economic and social matters” is similar to s 136. Now the Minister sought to leverage off this by saying that this suggested that s 136 was also in the policy basket such as to deny a duty. But this argument went nowhere. They are quite different powers. Moreover, true it is that such a phrase might bring within it policy questions, but for the s 136 power they can appropriately be dealt with at the breach stage.

459    Generally, the powers given in ss 33, 37A and 37B may be contrasted with the quite different ss 130, 133 and 136.

460    For convenience I should also refer to s 158. An analogous point can be made concerning the exemption power under s 158, sub-sections (1) to (5) of which provide:

(1)    A person proposing to take a controlled action, or the designated proponent of an action, may apply in writing to the Minister for an exemption from a specified provision of Part 3 or of this Chapter.

(2)    The Minister must decide within 20 business days of receiving the application whether or not to grant the exemption.

(3)    The Minister may, by written notice, exempt a specified person from the application of a specified provision of Part 3 or of this Chapter in relation to a specified action.

(4)    The Minister may do so only if he or she is satisfied that it is in the national interest that the provision not apply in relation to the person or the action.

(5)    In determining the national interest, the Minister may consider Australia’s defence or security or a national emergency, including an emergency to which a national emergency declaration (within the meaning of the National Emergency Declaration Act 2020) relates. This does not limit the matters the Minister may consider.

461    So far I have just discussed Pts 1 to 4.

462    Part 5 deals with bilateral agreements between the Commonwealth and a State. It is relevant to note s 47(4) that I will say something about in a moment. Relevant statutory provisions have been identified and set out in the reasons of others.

463    Parts 6 and 7 deal, inter-alia, with the question of whether approval of actions is needed. It addresses the concepts of “controlled action” (ss 67, 67A and 68), reference of a proposal to the Minister, and decisions made by the Minister as to whether an action the subject of a proposal needs approval (s 75). Again, the relevant provisions have been identified and set out in the reasons of others.

464    Clearly, these provisions and, if I might say so, the gateway, are not concerned with, let alone focused on, GHG emissions or human health and safety. They are tied back to the controlling provisions under Pt 3 and are, relatively speaking, narrow in their operation.

465    But there is one matter that it is convenient to note at this point. One of the Minister’s points on incoherency is that the Minister has no general power to impose conditions, but could only approve or not approve rather than to approve subject to conditions that could accommodate or be tailored to any duty of care.

466    But of course s 72 provides for the making of alternative proposals. It says:

(1)    A referral of a proposal to take an action must be made in a way prescribed by the regulations.

(2)    A referral of a proposal to take an action must include the information prescribed by the regulations.

(3)    A referral of a proposal to take an action may include alternative proposals relating to any of the following:

(a)    the location where the action is to be taken;

(b)    the time frames within which the action is to be taken;

(c)    the activities that are to be carried out in taking the action.

467    And s 133 accommodates such alternatives. It provides in s 133(1A):

If the referral of the proposal to take the action included alternative proposals relating to any of the matters referred to in subsection 72(3), the Minister may approve, for the purposes of subsection (1), one or more of the alternative proposals in relation to the taking of the action.

468    A person seeking approval might well anticipate that any exercise of power under s 133 might well need to accommodate a duty of care and so submit alternative or even cascading proposals to address that eventuality. And if that be so, the limitations concerning the power to impose conditions may not carry the Minister as far as she would wish. I will identify the conditions provisions later.

469    Part 8 dealing with assessment does not apply in the present context given that s 83(1)(b) has been triggered.

470    In relation to assessment reports, in the present context, s 47(4) applied, which provides:

If a bilateral agreement has (or could have) the effect that an action need not be assessed under Part 8 but the action must still be approved under Part 9, the agreement must provide for the Minister to receive a report including, or accompanied by, enough information about the relevant impacts of the action to let the Minister make an informed decision whether or not to approve under Part 9 (for the purposes of each controlling provision) the taking of the action.

471    Now s 47(4) refers to the “relevant impacts” of the action. What is meant by that expression?

472    Before saying something about the relevancy question, I should note s 527E concerning “impacts” which provides:

(1)    For the purposes of this Act, an event or circumstance is an impact of an action taken by a person if:

(a)    the event or circumstance is a direct consequence of the action; or

(b)    for an event or circumstance that is an indirect consequence of the action—subject to subsection (2), the action is a substantial cause of that event or circumstance.

(2)    For the purposes of paragraph (1)(b), if:

(a)    a person (the primary person) takes an action (the primary action); and

(b)    as a consequence of the primary action, another person (the secondary person) takes another action (the secondary action); and

(c)    the secondary action is not taken at the direction or request of the primary person; and

(d)    an event or circumstance is a consequence of the secondary action;

then that event or circumstance is an impact of the primary action only if:

(e)    the primary action facilitates, to a major extent, the secondary action; and

(f)    the secondary action is:

(i)    within the contemplation of the primary person; or

(ii)    a reasonably foreseeable consequence of the primary action; and

(g)    the event or circumstance is:

(i)    within the contemplation of the primary person; or

(ii)    a reasonably foreseeable consequence of the secondary action.

473    I should flag here that the Minister contends that the common law concept of reasonable foreseeability is incoherent with this statutory concept, a contention with which I disagree.

474    And as to the relevancy question, this is not unconstrained. What is being considered here is impacts that the action has or will have or is likely to have on the relevant matter protected by a provision of Part 3; see s 82 (1) under Pt 8, although the definition of “relevant impacts” (s 528) applies the s 82(1) definition more broadly. It is not dealing, for example, with climate change impacts as such. That is not a relevant matter protected by a provision of Part 3. Relevancy is not a broad or open-ended question in the present context.

475    Let me now turn more directly to Pt 9, which embodies the statutory powers relevant to the present context.

476    Sections 130(1), (1A), (1B)(a) and (2)(a) provide:

(1)    The Minister must decide whether or not to approve, for the purposes of each controlling provision for a controlled action, the taking of the action.

(1A)    The Minister must make the decision within the relevant period specified in subsection (1B) that relates to the controlled action, or such longer period as the Minister specifies in writing.

(1B)    The relevant period, in relation to a controlled action, is as follows:

(a)    if the action is the subject of an assessment report—the period of 30 business days beginning on the first business day after the Minister receives the assessment report;

(2)    An assessment report is a report given to the Minister as described in:

(a)    subsection 47(4) (about assessments under a bilateral agreement); or

477    Section 131 then provides:

(1)    Before the Minister (the Environment Minister) decides whether or not to approve, for the purposes of a controlling provision, the taking of an action, and what conditions (if any) to attach to an approval, he or she must:

(a)    inform any other Minister whom the Environment Minister believes has administrative responsibilities relating to the action of the decision the Environment Minister proposes to make; and

(b)    invite the other Minister to give the Environment Minister comments on the proposed decision within 10 business days.

(2)    A Minister invited to comment may make comments that:

(a)    relate to economic and social matters relating to the action; and

(b)    may be considered by the Environment Minister consistently with the principles of ecologically sustainable development.

This does not limit the comments such a Minister may give.

478    As contemplated under s 131(2)(a), a Minister invited to comment can comment on economic and social matters.

479    Sections 131AA(1) and (2) then provide:

(1)    Before the Minister decides whether or not to approve, for the purposes of a controlling provision, the taking of an action, and what conditions (if any) to attach to an approval, he or she must:

(a)    inform the person proposing to take the action, and the designated proponent of the action (if the designated proponent is not the person proposing to take the action), of:

(i)    the decision the Minister proposes to make; and

(ii)    if the Minister proposes to approve the taking of the action—any conditions the Minister proposes to attach to the approval; and

(b)    invite each person informed under paragraph (a) to give the Minister, within 10 business days (measured in Canberra), comments in writing on the proposed decision and any conditions.

(2)    If the Minister proposes not to approve, for the purposes of a controlling provision, the taking of the action, the Minister must provide to each person informed under paragraph (1)(a), with the invitation given under paragraph (1)(b):

(a)    a copy of whichever of the following documents applies to the action:

(i)    an assessment report;

(ii)    a finalised recommendation report given to the Minister under subsection 93(5);

(iii)    a recommendation report given to the Minister under section 95C, 100 or 105; and

(b)    any information relating to economic and social matters that the Minister has considered; and

(c)    any information relating to the history of a person in relation to environmental matters that the Minister has considered under subsection 136(4); and

(d)    a copy of any document, or part of a document, containing information of a kind referred to in paragraph 136(2)(e) that the Minister has considered.

480    Section 131A provides:

Before the Minister decides whether or not to approve, for the purposes of a controlling provision, the taking of an action, and what conditions (if any) to attach to an approval, he or she may publish on the internet:

(a)    the proposed decision and, if the proposed decision is to approve the taking of the action, any conditions that the Minister proposes to attach to the approval; and

(b)    an invitation for anyone to give the Minister, within 10 business days (measured in Canberra), comments in writing on the proposed decision and any conditions.

481    Section 133(1), (1A), (2) and (8)(a) provide:

(1)    After receiving the assessment documentation relating to a controlled action, or the report of a commission that has conducted an inquiry relating to a controlled action, the Minister may approve for the purposes of a controlling provision the taking of the action by a person.

(1A)    If the referral of the proposal to take the action included alternative proposals relating to any of the matters referred to in subsection 72(3), the Minister may approve, for the purposes of subsection (1), one or more of the alternative proposals in relation to the taking of the action.

(2)    An approval must:

(a)    be in writing; and

(b)    specify the action (including any alternative proposals approved under subsection (1A)) that may be taken; and

(c)    name the person to whom the approval is granted; and

(d)    specify each provision of Part 3 for which the approval has effect; and

(e)    specify the period for which the approval has effect; and

(f)    set out the conditions attached to the approval.

(8)    In this section:

assessment documentation, in relation to a controlled action, means:

(a)    if the action is the subject of an assessment report—that report; or

482    I will deal with conditions in a moment, but let me at this point go to s 136.

483    Section 136(1), (2) and (5) provides:

(1)    In deciding whether or not to approve the taking of an action, and what conditions to attach to an approval, the Minister must consider the following, so far as they are not inconsistent with any other requirement of this Subdivision:

(a)    matters relevant to any matter protected by a provision of Part 3 that the Minister has decided is a controlling provision for the action;

(b)    economic and social matters.

(2)    In considering those matters, the Minister must take into account:

(a)    the principles of ecologically sustainable development; and

(b)    the assessment report (if any) relating to the action; and

(ba)    if Division 3A of Part 8 (assessment on referral information) applies to the action—the finalised recommendation report relating to the action given to the Minister under subsection 93(5); and

(bc)    if Division 4 of Part 8 (assessment on preliminary documentation) applies to the action:

(i)    the documents given to the Minister under subsection 95B(1), or the statement given to the Minister under subsection 95B(3), as the case requires, relating to the action; and

(ii)    the recommendation report relating to the action given to the Minister under section 95C; and

(c)    if Division 5 (public environment reports) of Part 8 applies to the action:

(i)    the finalised public environment report relating to the action given to the Minister under section 99; and

(ii)    the recommendation report relating to the action given to the Minister under section 100; and

(ca)    if Division 6 (environmental impact statements) of Part 8 applies to the action:

(i)    the finalised environmental impact statement relating to the action given to the Minister under section 104; and

(ii)    the recommendation report relating to the action given to the Minister under section 105; and

(d)    if an inquiry was conducted under Division 7 of Part 8 in relation to the action the report of the commissioners; and

(e)    any other information the Minister has on the relevant impacts of the action (including information in a report on the impacts of actions taken under a policy, plan or program under which the action is to be taken that was given to the Minister under an agreement under Part 10 (about strategic assessments)); and

(f)    any relevant comments given to the Minister in accordance with an invitation under section 131 or 131A; and

(fa)    any relevant advice obtained by the Minister from the Independent Expert Scientific Committee on Coal Seam Gas and Large Coal Mining Development in accordance with section 131AB; and

(g)    if a notice relating to the action was given to the Minister under subsection 132A(3) — the information in the notice.

(5)    In deciding whether or not to approve the taking of an action, and what conditions to attach to an approval, the Minister must not consider any matters that the Minister is not required or permitted by this Division to consider.

484    A number of observations can be made.

485    First, the matters in s 136(1) are mandatory considerations.

486    Second, the subject areas in s 136(2) are subject areas, whether documents or information, that must be taken into account in considering the mandatory factors.

487    Third, s 136(1)(a) and (b) deal with different subject matter.

488    Fourth, as to the second limb, being “economic and social matters”, an interesting question arises. For s 136(2)(b), can you take into account any aspect of the s 47(4) assessment report which is relevant to “economic and social matters”? Or are ss 136(2)(a) to (d) for example referable to s 136(1)(a), and s 136(2)(f) for example referable to s 136(1)(b)? After all, the s 47(4) report was never procured for economic and social matters. And it does not seem justifiable to make it mandatory to take into account all aspects of the s 47(4) report, but only so much of that report referable to considering the “relevant impacts” which are of narrower compass.

489    Fifth, let me at this point say something about s 136(5). In terms of “any matters….not required…”, clearly ss 136(1) and (2) refer to required matters. So, anything outside of ss 136(1) and (2) are not required. But the phrase is “not required or permitted”. Could you consider something beyond ss 136(1) and (2) provided that it was permitted? In my view you can. And there is nothing in Division 1 of Part 9 that indicates that it is not permitted to consider human health and the posited duty. But of course they would have been non-mandatory considerations, and whether or not under s 136(1)(b).

490    But let it be assumed that the two limbs of ss 136(1)(a) and (b) are exhaustive of the possibilities. Then perhaps I do not need to consider such questions, if it be accepted that such matters fall under s 136(1)(b) anyway. And for this purpose it does not matter whether they fall under s 136(2) also, as the listing thereunder does not purport to be exhaustive. So, they could be a non-mandatory or permitted aspect falling under the mandatory subject matter of s 136(1)(b).

491    Now for completeness I should note that I have considered the four observations made by Jessup J in Tarkine National Coalition Inc v Minister for the Environment (2015) 233 FCR 254 at [25] to [28]. I agree with his first and second observations (at [25] and [26]). But let me say something about his third and fourth observations and deal with some aspects that he did not consider; I should say that in his context he may not have needed to.

492    As to his third observation (at [27]), in its generality it is not objectionable. However, in relation to the document referred to in s 136(2)(b), I am not convinced that the legislature intended that the entire contents of the assessment report had to be taken into account. Rather, and as s 47(4) makes plain, this report was to be obtained so as to assess “the relevant impacts of the action”, which of course is then focusing on the matters in s 136(1)(a). So, only that part of the report was mandated to be taken into account, not all of it. But admittedly, a literal meaning of s 136(2)(b) would require the entire contents of the assessment report to be considered (whatever the topic or scope) in relation to the matters in either ss 136(1)(a) or (b). But it seems to me that the literal approach suffers from taking a non-purposive and decontextualized reading of s 136(2)(b). And as I have touched on, why would the legislature mandate that the entire contents of the assessment report, outside the scope of s 47(4), had to be considered?

493    The choices for the possible construction of s 136(2)(b) seem to be the following. First, you must take into account the entire contents of the assessment report insofar as it is relevant to ss 136(1)(a) or (b). Second, you must take into account that part of the assessment report relevant to and contemplated by s 47(4). But as to this second possibility there are two colours. Should that part of the assessment report only be taken into account with the analogue s 136(1)(a) matter? Or should it be taken into account in relation to either of the matters under ss 136(1)(a) or (b)?

494    The literal view favours the first possibility described. But then this would mandate taking into account matters that are extraneous. But then perhaps s 136(1) is the constraining lens. A purposive and contextual approach would favour the second possibility, and of the two colours of the second possibility, arguably only the first colour. But I do not need to resolve these questions. The parties before us proceeded on the first possibility involving the literal approach.

495    Now why am I making these points? The reason is that the Minister is seeking to leverage off the NSW discussion of GHG emissions and policy, which appears in the NSW Department’s assessment report by reason of different NSW statutory provisions that I will come to in a moment, to say that her decision must be similarly policy based. This is a non-sequitur. There is nothing in the Act which mandates any consideration of such matters in terms of policy. Now she may be able to take these matters into account under s 136(1)(b). But that is a different question. In my view the Act does not elevate policy questions concerning GHG emissions, as the Minister would have it, to such a level as would deny any duty of care. But of course to conclude in essence that the Act has little to do with GHG emissions is a two-edged sword for the respondents for other reasons that I will come to concerning a lack of sufficient closeness and directness.

496    As to Jessup J’s fourth observation (at [28]), I doubt that it is wholly accurate to say that s 136(5) confirms “the impression that Subdiv B established a closed system of the matters that the Minister was to consider in making his decision, and the things that should be taken into account”. That cannot be literally true, as s 391 makes apparent. Further, it understates the work that the second limb of the disjunction “…or permitted” can do. That limb has considerable flexibility as compared with the first limb. One cannot say confidently what “is not … permitted by this Division to consider” in the absence of an express provision saying that something should not be taken into account. In any event I do not need to linger on this aspect, particularly as his Honour left himself with some wriggle room with phraseology such as “the impression”.

497    Let me now identify the provisions concerning conditions, as they are relevant to one of the Minister’s points concerning incoherency.

498    On the question of conditions, ss 134(1), (2), (3) and (3A) provide:

(1)    The Minister may attach a condition to the approval of the action if he or she is satisfied that the condition is necessary or convenient for:

(a)    protecting a matter protected by a provision of Part 3 for which the approval has effect (whether or not the protection is protection from the action); or

(b)    repairing or mitigating damage to a matter protected by a provision of Part 3 for which the approval has effect (whether or not the damage has been, will be or is likely to be caused by the action).

(2)    The Minister may attach a condition to the approval of the action if he or she is satisfied that the condition is necessary or convenient for:

(a)    protecting from the action any matter protected by a provision of Part 3 for which the approval has effect; or

(b)    repairing or mitigating damage that may or will be, or has been, caused by the action to any matter protected by a provision of Part 3 for which the approval has effect.

This subsection does not limit subsection (1).

(3)    The conditions that may be attached to an approval include:

(aa)    conditions requiring specified activities to be undertaken for:

(i)    protecting a matter protected by a provision of Part 3 for which the approval has effect (whether or not the protection is protection from the action); or

(ii)    repairing or mitigating damage to a matter protected by a provision of Part 3 for which the approval has effect (whether or not the damage may or will be, or has been, caused by the action); and

(f)    conditions requiring specified environmental monitoring or testing to be carried out; and

(h)    conditions relating to any alternative proposals in relation to the taking of the action covered by the approval (as permitted by subsection 133(1A)).

This subsection does not limit the kinds of conditions that may be attached to an approval.

(3A)    The following kinds of condition cannot be attached to the approval of an action unless the holder of the approval has consented to the attachment of the condition:

(a)    a condition referred to in paragraph (3)(aa), if the activities specified in the condition are not reasonably related to the action;

499    Clearly the conditions that can be imposed are quite limited in the sense that they are constrained by the ambit of ss 134(1) and (2).

500    In terms of varying conditions, ss 143(1)(a), (b), (ba) and (2) provide:

(1)    The Minister may, by written instrument, revoke, vary or add to any conditions (other than the condition referred to in subsection 134(1A)) attached to an approval under this Part of an action if:

(a)    any condition attached to the approval has been contravened; or

(b)    both of the following conditions are satisfied:

(i)    the action has had a significant impact that was not identified in assessing the action on any matter protected by a provision of Part 3 for which the approval has effect, or the Minister believes the action will have such an impact;

(ii)    the Minister believes it is necessary to revoke, vary or add a condition to protect the matter from the impact; or

(ba)    all of the following conditions are satisfied:

(i)    the action has had a significant impact on a matter protected by a provision of Part 3 for which the approval has effect, or the Minister believes the action will have such an impact;

(ii)    the Minister is satisfied that the impact is substantially greater than the impact that was identified in assessing the action;

(iii)    the Minister believes it is necessary to revoke, vary or add a condition to protect the matter from the impact; or

(2)    The Minister may, by written instrument, revoke any condition (other than the condition referred to in subsection 134(1A)) attached to an approval under this Part of an action if the Minister is satisfied that the condition is not needed to protect any matter protected by a provision of Part 3 for which the approval has effect.

501    Sections 144 and 145 deal with the suspension and revocation of approvals which it is not necessary to set out. I should also make reference to some other provisions.

502    First, there are provisions in the Act such as ss 193, 212, 236 and 255 which address specific aspects relating to human health, but they add little to the present debate. It might be said that such limited references are not consistent with superimposing the duty of care in question. Equally it might be said that such limited references demonstrate the point that the Act has principally a different focus. But that does not entail incoherence or inconsistency with the posited duty of care. Rather, they can sit side by side.

503    Second, ss 391(1) and (2) provide:

(1)    The Minister must take account of the precautionary principle in making a decision listed in the table in subsection (3), to the extent he or she can do so consistently with the other provisions of this Act.

(2)    The precautionary principle is that lack of full scientific certainty should not be used as a reason for postponing a measure to prevent degradation of the environment where there are threats of serious or irreversible environmental damage.

504    Further, in s 391(3) in item 2 of the table there is a reference to an approval decision under s 133.

505    Third, there are various enforcement provisions contained in the Act dealing with contraventions of the Act; this is not of course the present context, but such provisions are not irrelevant to the incoherence question. Sections 475 to 480 deal with injunctions and standing questions. Sections 480A to 480N deal with remediation orders and determinations. Section 500 deals with compensation. Section 501 provides:

This Division does not affect any other powers or rights under this Act, the regulations or any other law.

506    Fourth, there is no express or necessarily implied statutory immunity or limitation of liability conferred on or operating for the benefit of the Minister in relation to any common law tort claim.

507    Fifth, the Minister is able to delegate her ss 130 and 133 powers. She is not personally indispensable to the process. Section 515(1) provides:

The Minister may, by signed instrument, delegate all or any of his or her powers or functions under this Act to an officer or employee in the Department or to the Director. The delegate is, in the exercise or performance of a delegated power or function, subject to the directions of the Minister.

The NSW regime

508    I should now say something about the NSW legislation under which the assessment report was given.

509    Section 4.2 of the Environmental Planning and Assessment Act 1979 (NSW) (the NSW Act) relevantly provided:

Development that needs consent

(1)    General If an environmental planning instrument provides that specified development may not be carried out except with development consent, a person must not carry the development out on land to which the provision applies unless—

(a)    such a consent has been obtained and is in force, and

(b)    the development is carried out in accordance with the consent and the instrument.

(2)    For the purposes of subsection (1), development consent may be obtained—

(a)    by the making of a determination by a consent authority to grant development consent, or

(b)    in the case of complying development, by the issue of a complying development certificate.

(5)    Complying development An environmental planning instrument may provide that development, or a class of development, that can be addressed by specified predetermined development standards is complying development.

510    The relevant consent authority for present purposes was the Independent Planning Commission (IPC) (see s 4.5(a)).

511    In terms of matters that had to be considered in determining a development application, s 4.15(1) provided:

Evaluation

(1)    Matters for consideration—general In determining a development application, a consent authority is to take into consideration such of the following matters as are of relevance to the development the subject of the development application—

(a)    the provisions of—

(i)    any environmental planning instrument, and

(ii)    any proposed instrument that is or has been the subject of public consultation under this Act and that has been notified to the consent authority (unless the Planning Secretary has notified the consent authority that the making of the proposed instrument has been deferred indefinitely or has not been approved), and

(iii)    any development control plan, and

(iiia)    any planning agreement that has been entered into under section 7.4, or any draft planning agreement that a developer has offered to enter into under section 7.4, and

(iv)    the regulations (to the extent that they prescribe matters for the purposes of this paragraph),

(v)    (Repealed)

that apply to the land to which the development application relates,

(b)    the likely impacts of that development, including environmental impacts on both the natural and built environments, and social and economic impacts in the locality,

(c)    the suitability of the site for the development,

(d)    any submissions made in accordance with this Act or the regulations,

(e)    the public interest.

512    So, unlike the Commonwealth legislation under present discussion, the public interest was required to be expressly considered.

513    I should also note s 4.16(1) and (2) which provided:

Determination

(1)    General A consent authority is to determine a development application by—

(a)    granting consent to the application, either unconditionally or subject to conditions, or

(b)    refusing consent to the application.

(2)    Despite subsection (1), the consent authority must refuse an application for development, being the subdivision of land, that would, if carried out, result in a contravention of this Act, an environmental planning instrument or the regulations, whether arising in relation to that or any other development.

514    Further, s 4.40 provided:

Evaluation of development application (s 4.15)

Section 4.15 applies, subject to this Division, to the determination of the development application.

515    An assessment report was published in May 2020 and prepared by the NSW Department of Planning, Industry and Environment for consideration by the IPC. The report considered GHG emissions. But importantly, it did so as part of considering the public interest, which was an express statutory requirement under the NSW regime. A summary of the position concerning GHG emissions was included up front (p xiv) and is set out in the reasons of Allsop CJ. I need not repeat it.

516    Social and economic factors were also considered (see pp xiii to xiv) and [576] to [643]). Social and economic factors and the public interest were mandatory factors to consider under s 4.40 of the NSW Act (see [71]).

517    In section 6.10 ([670] to [715]) there was a detailed discussion of the mandatory “public interest” consideration. In particular, [672], [673], [675], [676] and [678] discussed GHG emissions and separately addressed scope 1, scope 2 and scope 3 emissions. It is not necessary to set these paragraphs out.

518    Further, [683] to [688], [690], [692] to [701], [704], [707], [708], [711] to [713] and [715] discussed policy considerations concerning climate change. I would note some paragraphs dealing with the context for the discussion being the following:

Climate Change Policy Consideration

Under Clause 14(1) of the [State Environmental Planning Policy (Mining, Petroleum Production and Extractive Industries 2007 (Mining SEPP)], the consent authority is required to consider whether conditions should be attached to consents to ensure that the development is undertaken in an environmentally responsible manner, including conditions to ensure that GHG emissions are minimised to the greatest extent possible.

Under Clause 14(2), the consent authority, in determining a development application, must also consider an assessment of GHG emissions (including downstream emissions) from the development, and must do so having regard to any applicable State or national policies, programs or guidelines concerning GHG emissions.

There are two key documents of relevance for the assessment of the Project:

    the NSW Government’s NSW Climate Change Policy Framework (CCPF); and

    the Commonwealth Government’s commitments to the United Nations Framework Convention on Climate Change (UNFCCC) Paris Agreement 2015 (Paris Agreement).

It is noted that more recently (i.e. March 2020), the Government announced a new 10- year plan to put the State on track to achieve net-zero emissions by 2050, the Net Zero Plan Stage 1: 2020-2030. The Plan builds on the CCPF and sets out a number of initiatives to deliver a 35% cut in emissions by 2030, compared to 2005 levels.

Whitehaven in Section 6.15 of its Submissions Report provided further advice on the application of the CCPF and the Paris Agreement commitments to the Project.

The Department has considered the CCPF, which outlines the State’s long-term aspirational objectives of achieving net-zero emissions by 2050 and making NSW more resilient to a changing climate. The CCPF does not set prescriptive emission reduction targets and sets policy directions for government action, for example, to improve opportunities for private sector investment in low emissions technology in the energy industry, which is needed for a transition to a net-zero emissions inventory.

Under the Paris Agreement, the Australian Government committed to a nationally determined contribution (NDC) to reduce national GHGEs by between 26 and 28 percent from 2005 levels by 2030. Australia has committed to meeting this target through initiatives that focus on expanding renewable energy sources, supporting low emissions technologies, improving energy efficiencies and incentivising companies to reduce their emissions without compromising economic growth and driving up energy prices.

A regular 5-yearly review of NDCs is required under the Paris Agreement with the next review to be submitted by signatories in 2020. The Department acknowledges that ongoing review to meet emission targets by signatories may affect export markets for coal and that the UNFCCC global approach to nationally determined emission reduction targets is the appropriate mechanism for managing Australia’s Scope 3 emissions, rather than regulating Scope 3 emissions on a project by project basis in Australia.

The Department also notes that the Department’s ‘Guidelines for Economic Assessment of Mining and Coal Seam Gas Proposals’ and the associated 2018 technical notes do not require the social cost of Scope 3 emissions to be incorporated into the economic evaluation when determining the net benefits to NSW or Australia of the development. This approach, where both the costs and benefits of consumption and use of the coal is considered by the country/development where the coal is being used, is consistent with the global accounting framework for GHG emissions under the UNFCCC.

Importantly, the NSW or Commonwealth Government’s current policy frameworks do not promote restricting private development as a means for Australia to meet its commitments under the Paris Agreement or the long-term aspirational objective of the CCPF guidelines. Neither do they require any action to taken by the private sector in Australia to minimise or offset the GHG emissions of any parties outside of Australia, including the emissions that may be generated in transporting or using goods that are produced in Australia.

In November 2019, the Commonwealth Government wrote to Minister Stokes (see Appendix G2-3) about the consideration of GHG emissions advising that “any requirement to consider scope three emissions within a sub-national or state jurisdiction is inconsistent with long accepted international carbon accounting principles and Australia’s international commitments.”

The Department also notes that it is not the NSW Government’s policy that planning conditions should seek to regulate directly or indirectly matters of international trade which are appropriately regulated by the Commonwealth Government.

On this basis, the Department does not consider the Project is inconsistent with Australia’s commitments to the Paris Agreement.

In regard to Clause 14(1) of the Mining SEPP, the focus should be on the impacts that can be reasonably controlled by the applicant of a development (such as the Scope 1 and relevant Scope 2 emissions) and not Scope 3 or downstream emissions, as these would be the Scope 1 or 2 emissions of another development. Again, this is consistent with the global accounting framework for GHG emissions under the UNFCCC.

There is no NSW or Commonwealth policy that supports placing conditions on an applicant to minimise the Scope 3 emissions of its development. Any such policy is likely to result in significant implications for the NSW and Australian economy and it is not clear it would have any effect on reducing GHG emissions generated by parties in other jurisdictions outside Australia. Further, conditions must be for a proper planning purpose, must fairly and reasonably relate to the subject development, and must not be manifestly unreasonable.

6.10.3 Conclusion

The assessments undertaken by Whitehaven for the Project have indicated that the potential GHG emissions (total and annual average forecast) generated over the life of the Project would be small in comparison to Australia’s current total contribution and its NDC target for 2030, under the Paris Agreement. Further compared to the Approved Mine there is a reduction in Scope 1 emissions due to the reduction in truck haulage by road replaced by trains once the rail spur line is commissioned.

The Department notes that Scope 3 emissions for the Project have been adequately accounted for in the assessment, considering that these indirect emissions would occur at sources not owned or controlled by Whitehaven. Further, the Department recognises that the agreed accounting framework for measuring GHG emissions (as also adopted by the Paris Agreement) is focused on a production-based methodology (ie. Scope 1 and Scope 2 emissions) and as such, Scope 3 emissions from the Project would be accounted for by the consuming country’s NDCs.

The Department considers that the Project has been adequately assessed against the relevant State or national policies, programs or guidelines concerning GHG emissions and would be measured and reported as required under the Commonwealth Government’s National Greenhouse and Energy Report Scheme (NGERS) requirements. Importantly, neither the NSW or Commonwealth policy frameworks require the private sector in Australia to minimise or offset the GHG emissions of any parties outside of Australia.

Overall, the Department considers that the GHG emissions for the Project have been adequately considered and that, with the Department’s recommended conditions, are acceptable when weighed against the relevant climate change policy framework, objects of the EP&A Act (including the principles of ESD) and socio-economic benefits of the Project.

519    The IPC made its decision and published its reasons on 12 August 2020. But the IPC’s decision and reasons has no separate status under the Commonwealth legislation. The relevant report for the purposes of the Act was the NSW Department’s assessment report, not the IPC’s reasons. Having said that, it should be noted that the IPC at [211] to [216], [220], [221] and [223] discussed GHG emissions, the context for which is explained by the following extracts:

Commission’s Findings

Clause 14(1)(c) of the Mining SEPP requires the Commission to consider whether or not the consent should be issued subject to conditions aimed at ensuring that the development is undertaken in an environmentally responsible manner, including conditions to ensure that greenhouse gases emissions are minimised to the greatest extent practicable. Clause 14(2) requires the Commission to consider an assessment of the greenhouse gas emissions (including downstream emissions) of the development, and to do so having regard to any applicable State or national policies, programs or guidelines concerning greenhouse gas emissions…

The Commission has considered the public’s written comments and presentations at the Public Hearing in relation to GHG emissions and the potential contribution of the Project to climate change

The Commission acknowledges that the aim of the NSW Climate Change Policy Framework (CCPF) is to “maximise the economic, social and environmental wellbeing of NSW in the context of a changing climate and current and emerging international and national policy settings and actions to address climate change” with the aim to achieve net-zero emissions by 2050 and to ensure NSW is more resilient to a changing climate. The Commission notes that the CCPF does not set prescriptive emission reduction targets and sets policy directions for government action as stated by the Department in paragraph 207 above. The Commission also notes that the NSW Government released the Net Zero Plan Stage 1: 2020-2030 (Net Zero Plan) in March 2020 as referenced by the Department in paragraph 208 above. The Commission notes that the Net Zero Plan builds on the CCPF and sets out a number of initiatives to deliver a 35% cut in emissions by 2030, compared to 2005 levels. The Commission agrees with the Department’s assessment in paragraph 207 above that the Project is not inconsistent with the CCPF and that the Applicant has committed to minimising its Scope 1 emissions over which it has direct control.

The Commission notes that, under the Paris Agreement, the Australian Government committed to a nationally determined contribution (NDC) to reduce national GHG emissions by between 26 and 28 percent from 2005 levels by 2030. The Commission also notes that Australia does not require monitoring or reporting of Scope 3 emissions under the NGERS and they are not counted in Australia’s national inventory of GHG emissions under the Paris Agreement. The Commission agrees with the Department’s statement in paragraph 209 above that the Project’s Scope 3 emissions would not contribute to Australia’s NDC, as product coal would be exported overseas. The Commission notes that these Scope 3 emissions become the consumer countries’ Scope 1 and 2 emissions and would be accounted for under the Paris Agreement in their respective national inventories.

The Commission agrees with the Department’s statement in paragraph 209 above and acknowledges that Scope 3 emissions from the combustion of product coal are a significant contributor to anthropological climate change and that the contribution of the Project to the potential impacts of climate change in NSW must be considered in assessing the overall merits of the development application.

For the reasons set out above, the Commission is of the view that the GHG emissions for the Project have been adequately considered. The Commission finds that on balance, and when weighed against the relevant climate change policy framework, objects of the EP&A Act, ESD principles (section 4.10) and socio-economic benefits (section 4.9.6), the impacts associated with the GHG emissions of the Project are acceptable and consistent with the public interest. The Commission therefore imposes the Conditions B35, B36 and B37 as recommended by the Department.

520    Let me deal with the topic of bio-diversity, which was after all one of the central themes sought to be protected under the Act.

521    The NSW Department’s assessment report dealt with the Commonwealth topics concerning bio-diversity at [85], [86], and Section 6.4 generally ([399] to [467]). I do not need to linger on the discussion concerning threatened species such as the Regent Honeyeater, Swift Parrot, South-eastern Long-eared Bat (Corben’s Long-eared Bat), Large-eared Pied Bat and, surprisingly, the Murray Cod. Moreover, Appendix J to that report was headed “Consideration of Commonwealth Matters” and dealt in more detail with the bio-diversity questions. It did not deal with the “public interest” questions or the GHG emissions issues. And for completeness, the IPC in its decision also discussed the bio-diversity questions (see at [349] to [384]).

522    I have set out more than is usual from the NSW Department’s assessment report and the IPC report. And admittedly there is some repetition with some of the material extracted and set out in the reasons of Allsop CJ. But I have done so to establish the following points.

523    First, the discussion in such material concerning GHG emissions necessarily arises because of the different statutory regime and mandatory factors under the NSW Act. And in such a context, it is well understandable why one might characterise such material as raising core policy questions under the NSW regime given its boundaries and content. Now the Minister asserts that these matters are core policy, and I will deal with that assertion later. But there is nothing under the Act making them so. And if they are, it will not be by way of a side-wind by reference to the NSW Department’s assessment report.

524    Second, s 47(4) of the Act is not addressed to making relevant or picking up such matters. The utility of the NSW Department’s assessment report to the Commonwealth regime concerns the discussion on bio-diversity questions or Part 3 questions. These are the matters that are concerned with “relevant impacts of the action” (s 47(4)), which are impacts that the action in question has or is likely to have on the relevant matter protected by a provision of Part 3.

525    Third, it is an open question under s 136 as to whether the Minister can pick up everything in the NSW Department’s assessment report on GHG emissions and make it relevant to “economic and social matters” (s 136(1)(b)). But assume for the sake of argument that she can. That does not make it core policy under the Act. True it is that these are policy questions. But if they are not core policy questions under the Commonwealth regime, then there is no good reason why they cannot be dealt with at the breach stage, rather than the duty stage. They should not foreclose the existence of a duty. But as policy considerations, and assuming that they are taken into account when considering whether or not to approve, there may be no breach of duty if approval is given. Hence my enquiry of the parties during the course of the hearing as to whether the Minister’s approval decision and reasons should be tendered before us, a proposal objected to by the respondents’ counsel. I will return to this in a moment.

526    Now before getting into the detail of the incoherency and policy arguments, it is convenient to briefly address some other legislation, noting that there was little regard paid by the parties or the primary judge to relevant State legislation that may have impacted on the existence or ambit of any common law duty of care, such as the Civil Liability Act 2002 (NSW), applying choice of law principles and ss 79 and 80 of the Judiciary Act 1903 (Cth). But let me make the following brief points.

527    First, in terms of s 42 dealing with whether a duty of care is owed by “a public or other authority”, it seems to me that it is not applicable to a Commonwealth Minister. The s 41 definition of “public or other authority” and the definition of “Crown” under s 3 of the Crown Proceedings Act 1988 (NSW) do not embrace the Crown in right of the Commonwealth or a Commonwealth Minister acting in such a capacity. And I say that notwithstanding the extension referred to in s 4(1). On the face of ss 41 and 42 the earlier provided extension does not apply. Further, ss 41 and 42 cannot be picked up under the Judiciary Act.

528    Second, even assuming the application of the more general provisions such as ss 5B to 5D, none of them cut across what I have said concerning any common law duty that may have been owed by the Minister.

529    Third, whatever State provisions may operate concerning proportionate liability (see for example ss 34 to 39), they do not apply to claims arising out of personal injury (s 34(1)(a)). I will return later to discuss another dimension of proportionality when I address its more conspicuous relative, indeterminacy.

530    I will not linger on any of these localised State related questions, particularly given that there may be multiple possibilities for the appropriate lex loci delicti given the breadth and location of the members of the claimant class; in other words, other States’ and Territories’ legislation may need to be considered.

Incoherency and policy questions

531    Before dealing with the incoherency and policy questions, I should deal with some preliminary matters.

Preliminary matters

532    Subsequent to the primary judge’s decision but before the hearing of the present appeal, the Minister on 15 September 2021 granted her approval to the extension project. So, notwithstanding that the primary judge had found that she owed a duty of care, that does not seem to have precluded or significantly inhibited her exercise of statutory power.

533    Now during the hearing of the appeal, I raised the question of whether the Minister’s approval decision and her reasons should be tendered. The Minister did not oppose this course, but the respondent objected. In those circumstances, I joined in the decision not to require that they be tendered. Having said that, let me make the following points.

534    First, the fact that the Minister proceeded as she did rather highlights the incoherency of one of the Minister’s arguments before the primary judge concerning whether the existence of any duty of care would be incoherent with the statutory regime under the Act.

535    Second, and notwithstanding that I do not have the Minister’s reasons, it would seem that it can reasonably be inferred that any policy questions were and could adequately be taken into account at the breach stage, namely, at the time she considered and then decided to give her approval.

536    Third, this whole exercise shows how unsatisfactory it is to split off the consideration of the existence of the duty from the questions of breach, causation and damage. Such matters should usually be decided as a composite exercise and when damage has occurred. After all, all of these levels interact or at the least inform each other. Now some might consider each question to be conceptually discrete. In one sense that is true. But at a practical and forensic level, it is much more desirable to consider them together. I suspect that Viscount Simonds would have been underwhelmed by the notion that a duty could be declared in the ether. In Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd [1961] AC 388 he said (at 425):

It is, no doubt, proper when considering tortious liability for negligence to analyse its elements and to say that the plaintiff must prove a duty owed to him by the defendant, a breach of that duty by the defendant, and consequent damage. But there can be no liability until the damage has been done. It is not the act but the consequences on which tortious liability is founded. Just as (as it has been said) there is no such thing as negligence in the air, so there is no such thing as liability in the air. Suppose an action brought by A for damage caused by the carelessness (a neutral word) of B, for example, a fire caused by the careless spillage of oil. It may, of course, become relevant to know what duty B owed to A, but the only liability that is in question is the liability for damage by fire. It is vain to isolate the liability from its context and to say that B is or is not liable, and then to ask for what damage he is liable. For his liability is in respect of that damage and no other. If, as admittedly it is, B’s liability (culpability) depends on the reasonable foreseeability of the consequent damage, how is that to be determined except by the foreseeability of the damage which in fact happened — the damage in suit? And, if that damage is unforeseeable so as to displace liability at large, how can the liability be restored so as to make compensation payable?

537    Moreover, one has to be careful in considering whether a duty of care is owed, separately from the other elements of the tort of negligence. It is not a legally enforceable duty as such but one of four elements of a common law construct, the combination of which gives rise to a cause of action. One does not take proceedings to enforce it like other contexts where specific performance or mandamus may be appropriate. Now true it is that an injunction was sought before the primary judge, which was not granted. But that concerned anticipated conduct, namely, approval, which was said to amount to a threatened breach leading to likely injury. But one was not enforcing a legal duty per se, such as a statutory or contractual duty, but rather seeking to restrain a threatened breach.

538    I must say that I have difficulty with declaring in the ether the existence of a duty of care, divorced from any question of breach, causation and damage. But that is what the primary judge did. I should say that there were three possibilities open to the judge. First, he could, as he did, declare there to be a duty. Second, he could have declared there to be no duty. Third, he could have refused to make any declaration, saying that whether there was a duty should be decided in the known world of breach, causation and damage, which was not before him. In my view that third course was the more appropriate option open to his Honour, although he did not take it. He took the first course that was not challenged by the Minister in terms of giving effect to his determination as to the existence of a duty. Of course, the Minister challenges the underlying determination. The course taken by his Honour was strictly open to him as an exercise of discretion to grant declaratory relief. And given the Minister’s acquiescence to his Honour’s formalization, I need not chart the outer reaches of any jurisdiction or discretion to grant a declaration in a context such as the present save to otherwise say that I agree with Wheelahan J’s exposition in his preliminary remarks supporting his conclusion that there was a real question as to whether the application for the declaration should have been entertained. But as I say, that question was foreclosed by the Minister’s stance both before the primary judge and before us.

Is there any incoherency? And if so, to what degree?

539    Let me turn then to deal with incoherency more directly, recognising of course that this question and the policy question interact, and on one view should be considered together given that one dimension of incoherency is incompatibility.

540    There are various levels at which one can approach the incoherency question.

541    First, one can look at it from the perspective solely of considering just the Act and whether the posited duty of care is inconsistent, incompatible or otherwise incoherent with the Act.

542    Second, one can look at it from the perspective of considering both the Act and the NSW Act and whether the posited duty is incoherent with such legislation seen in combination; after all, elements of the output of the NSW regime are utilised under the Act. The Minister focused on the first perspective and said little concerning the second perspective.

543    As to the first perspective, let me begin by making the following general points.

544    First, clearly the purpose of the Act and its operation has little expressly to do with climate change, GHG emissions or human safety. Contrastingly, the posited duty of care focuses on such matters. But to say that each is dealing with different subject matters does not establish inconsistency, incompatibility or other incoherency with the Act; incoherency is not established by a “two ships passing in the night” scenario.

545    Second, the posited duty does not undermine or substantially interfere with the Minister’s exercise of power. Indeed, the very approval by the Minister in the present case is at odds with such a suggestion. To accept such a duty did not in any way pre-empt or foreclose the Minister’s exercise of power in the present case, let alone dictate or require that the approval be refused.

546    Third and relatedly, it may be accepted that, to some extent, the recognition of a duty of care may skew or distort the exercise of a statutory discretionary power. But that has never been a bar to finding a common law duty that might affect the exercise of statutory power. Such a point could be made in respect of all cases falling within this class. The real question is the degree of distortion rather than the binary question of whether or not there is distortion.

547    Fourth, to find that there is such a duty does not as a matter of form impermissibly impose a mandatory consideration to be considered as such on the exercise of statutory power. Whether and how the Minister chooses to consider the duty is a matter for her. But of course if she chooses not to consider it and there is later a breach with consequential damage, then liability in tort may be imposed. But to impose a duty is not imposing a mandatory consideration, such that not to take it into account will invalidate the exercise of statutory power. And the validity or otherwise of the exercise of power has little to do with the tortious duty or liability questions in any event. Her exercise of statutory power could be perfectly valid, yet if she was under a duty, she could be liable in tort. Conversely, her exercise of power could be invalid, but she may not be liable in tort.

548    Fifth, the Act contains no exclusion or limitation of liability relating to common law negligence concerning the Minister’s exercise of power.

549    Sixth, to the extent that the posited duty might give rise to policy questions concerning GHG emissions, that is not a policy question that the Act deals with. It is silent on the question. No incoherency arises. Moreover and as I will come to, policy questions can be dealt with at the level of breach. Now of course the Minister may have to weigh various competing factors that may involve policy. And it might be said that under public law, that is for the decision-maker to weigh. Now that is all true. But such public law questions go to the validity of the exercise of the statutory power. They relate to the standards of administrative conduct relevant to validity. But we are here not dealing with validity questions. As the primary judge pointed out, neither the common law duty nor its content is directed to the making of a valid decision. Whatever policy questions need to be weighed for a valid decision is one thing. But we are here concerned with policy questions that may need to be weighed at the breach level. Of course they may overlap. But it is a conflation of these different dimensions to say that incoherence arises. Moreover, it is obvious that the standard of administrative conduct for a valid decision is not to be equated with a standard of reasonable care in terms of a common law duty of care.

550    Seventh, I accept that the limited power to impose conditions on any statutory approval gives rise to a small element of incompatibility as between the Act and any common law duty. There is limited scope to tailor conditions to match the duty or how it might be discharged.

551    Eighth, the common law duty coheres with the Act to the extent that the Minister can take into account as a non-mandatory consideration human health and safety.

552    Ninth, the present context does not involve an omission or failure to act, where incoherency may more easily be established. Here it is said that the positive act of the Minister in approving the extension is likely to give rise to the risk of harm.

553    Tenth, the statutory causation test under s 527E does not give rise to substantial incoherence.

554    Eleventh, the fortuitous nature of the need to get approval is not an incoherency question, particularly where the positive exercise gives rise to the risk of harm.

555    Twelfth, we are not dealing with a conflicting duties scenario of the type discussed by Beach JA in Regent Holdings Pty Ltd v State of Victoria [2013] VSC 601 at [56], [213] to [217] and [223], which in any event was a pure economic loss case.

556    Now I will address some of the authorities and elaborate on some of these points in a moment. But let me address the second level of perspective that I noted at the outset of this discussion, where it might be said that the posited duty is incoherent with both the NSW Act and the Act in combination.

557    The following argument could conceivably have been put involving the following elements. First, the IPC under the NSW Act was required to consider GHG emissions and policy questions including human health and safety. Second, the IPC gave approval to the extension after considering such matters. Third, the IPC owed no common law duty of care because of incoherency with the NSW Act or policy questions. Fourth and contrastingly, the Minister under the Act was not required to address the matters dealt with by the IPC except only indirectly as “economic and social matters”; the Act principally dealt with discrete subject matter. With these four elements, it might be said that incoherency arises in imposing the posited duty on the Minister, perhaps requiring or leading to the result of the non-approval of the extension, that would not by analogy be imposed on the IPC, in circumstances where it was the IPC and not the Minister that was required to address GHG emissions and policy questions including human health and safety, and where nevertheless the IPC had approved the extension. But such an argument was not put by the Minister in such terms, and she should not be given the benefit of it if it is otherwise sound. Moreover, the third element is problematic in any event. Indeed, if you negate it, the strength of this combined incoherency argument is diminished, albeit not removed. But even if you negate it, it still might be said that it is no business of the Minister to reconsider or take a different view on what the IPC was bound to consider but the Minister not, and then take a different position from the IPC on approval relevant to such questions. Of course though, such questions could be left to the breach stage rather than give rise to any incoherency such as to deny the posited duty. Anyway, I will not linger on arguments not put in such terms.

558    Let me begin with some authorities that have considered the question of incoherency.

559    Sullivan v Moody (2001) 207 CLR 562 was a clear case of incoherency between the posited duty of care and the statutory regime. A common law duty to persons suspected of harming children was inconsistent with a statutory scheme that made the interests of the children paramount. The Court said (at [60]):

The circumstance that a defendant owes a duty of care to a third party, or is subject to statutory obligations which constrain the manner in which powers or discretions may be exercised, does not of itself rule out the possibility that a duty of care is owed to a plaintiff. People may be subject to a number of duties, at least provided they are not irreconcilable. A medical practitioner who examines, and reports upon the condition of, an individual, might owe a duty of care to more than one person. But if a suggested duty of care would give rise to inconsistent obligations, that would ordinarily be a reason for denying that the duty exists. Similarly, when public authorities, or their officers, are charged with the responsibility of conducting investigations, or exercising powers, in the public interest, or in the interests of a specified class of persons, the law would not ordinarily subject them to a duty to have regard to the interests of another class of persons where that would impose upon them conflicting claims or obligations.

560    But no such problem concerning the question of inconsistency of obligation arises before us.

561    The Court said (at [62]):

The statutory scheme that formed the background to the activities of the present respondents was, relevantly, a scheme for the protection of children. It required the respondents to treat the interests of the children as paramount. Their professional or statutory responsibilities involved investigating and reporting upon, allegations that the children had suffered, and were under threat of, serious harm. It would be inconsistent with the proper and effective discharge of those responsibilities that they should be subjected to a legal duty, breach of which would sound in damages, to take care to protect persons who were suspected of being the sources of that harm. The duty for which the appellants contend cannot be reconciled satisfactorily, either with the nature of the functions being exercised by the respondents, or with their statutory obligation to treat the interests of the children as paramount. As to the former, the functions of examination, and reporting, require, for their effective discharge, an investigation into the facts without apprehension as to possible adverse consequences for people in the position of the appellants or legal liability to such persons. As to the latter, the interests of the children, and those suspected of causing their harm, are diverse, and irreconcilable. That they are irreconcilable is evident when regard is had to the case in which examination of a child alleged to be a victim of abuse does not allow the examiner to form a definite opinion about whether the child has been abused, only a suspicion that it may have happened. The interests of the child, in such a case, would favour reporting that the suspicion of abuse has not been dispelled; the interests of a person suspected of the abuse would be to the opposite effect.

562    But such a scenario dealt with in Sullivan is no analogue for the present case. And nor are we in the realm of the type of case considered in Tame v New South Wales (2002) 211 CLR 317, where Gleeson CJ said (at [26]):

The primary duty of a police officer filling out such a report is to make available to his or her superiors, honestly and frankly, the results of the observations, inquiries and tests that were made. It would be inconsistent with such a duty to require the police officer to take care to protect from emotional disturbance and possible psychiatric illness a person whose conduct was the subject of investigation and report.

563    McHugh J (at [122] to [126]) made a similar point.

564    Further, discussions of inconsistency or incoherence in failure to act cases such as Graham Barclay Oysters and Stuart v Kirkland-Veenstra (2009) 237 CLR 215, cannot be automatically translated to a third scenario case of a type that I discussed at the outset asserting a duty to take reasonable care in the exercise of statutory powers, where there is less scope for inconsistency. The Minister’s positive steps are taken under the statute, and it is those steps which are said to create a danger.

565    In my view, to recognise the posited duty does not create substantial incoherence with the Act.

566    Now the Minister focused her submission on three principal areas in challenging the primary judge’s conclusion that a duty limited to an obligation to take reasonable care to avoid personal injury or death to the claimant class was not incoherent with the Act, or to the extent that it was, that incoherence was not determinative.

567    The first area concerns his Honour’s analysis and his paramountcy given to statutory purpose. It is said that his Honour gave insufficient attention to the nature and function of the approval power and its place within the scheme of the Act. I tend to agree.

568    As the Minister correctly contended, ss 130 and 133 confer power to approve the taking of an action that is a controlled action and thus prohibited unless approved, but only because it has been determined to significantly impact one or more of the matters of national environmental significance identified in Pt 3. And s 136(1) makes clear that in determining whether to approve a controlled action, the Minister must weigh competing interests, including matters relevant to a matter of national environmental significance and “economic and social matters”. So, the Minister is empowered to approve a controlled action notwithstanding the impact that action will have on the relevant matters of national environmental significance.

569    Further, as the Minister also correctly contended, the approval power recognises that the Minister may need to address considerations that are in tension with each other and to reconcile competing objectives.

570    Further, as the Minister correctly said, the scheme for approval of controlled actions confirms that the focus of Commonwealth involvement in relation to the environment is not all aspects of the “environment” as defined in s 528, but the limited matters of national environmental significance identified in Pt 3, and the impacts of Commonwealth actions and of actions affecting Commonwealth areas. If a person proposes to take an action that they think may be a controlled action, they must refer the proposal to the Minister for a decision as to whether it is a controlled action (s 68(1)), who then decides whether an action the subject of a proposed referral is a controlled action, and which provisions of Pt 3 are controlling provisions for the action (s 75(1)). Section 75(2) provides that at the stage of determining whether a particular action is a controlled action, the Minister is obliged to consider only adverse impacts of the action, and not beneficial impacts. The task of bringing to bear both adverse and beneficial impacts is left to the final step of approval.

571    If an action is a controlled action, the relevant controlling provisions in Pt 3 prohibit the taking of the action unless an approval under Pt 9 is in operation. And to inform the Minister’s decision whether to approve the taking of a controlled action under Pt 9, the Act provides for various assessment processes. Each of the assessment processes is directed towards assessing the “relevant impacts” of a controlled action (see s 47(4)). Those are the impacts of the action on the matters protected by the controlling provisions.

572    Accordingly, as the Minister correctly contended, the scheme for assessment of referrals in relation to controlled actions confirms that the focus of the Act is not on all aspects of the environment, but the impacts upon particular matters of national environmental significance identified in Pt 3.

573    According to the Minister, the full terms of s 3 and the scheme of the Act as a whole do not support the primary judge’s finding that protection of people and communities is a purpose of the Act. She says that the approval provisions in Pt 9 are not engaged at all unless a proposal involves the taking of action that is prohibited by a provision of Pt 3, which concerns harm to matters of national environmental significance or harm to the environment in a Commonwealth area or from Commonwealth action. It is only where harm of that kind will occur that the Minister is called upon under ss 130 and 133 to make a decision that strikes a balance between competing interests and objectives. Accordingly, it is said that the primary judge’s approach of identifying a single statutory purpose, being one that was not tied to a matter protected by Pt 3, was in error.

574    Now I agree that his Honour overstated the matter. But the Minister too over-states her position.

575    In deciding under Pt 9 whether to approve the extension project with or without conditions, the Minister had to consider matters relevant to any matter protected by the “controlling provisions” for the action and “economic and social matters” (s 136(1)). In considering those matters, the Minister could take into account the principles in s 3A.

576    An object of the Act is to provide for the protection of the environment (s 3(1)(a)) including matters of national environmental significance. But the “environment” includes “people and communities” and the “social, economic and cultural aspects of” the environment (s 528). Such provisions, given the scope of “economic and social matters”, permit the Minister to consider whether an action may cause physical harm to human beings. This falls within the matters that the Minister may consider in making her decision.

577    Further, the Act is replete with references to protecting ecologies and ecological communities, and seeking to promote ecologically sustainable use. And one idea permeating the Act is inter-generational equity. In that context, is it seriously doubted that humans and their communities depend upon ecologies and ecological communities? Indeed we are in a symbiotic relationship with them. The environment is not just there to admire and objectify like a pretty butterfly.

578    Moreover, the Act is replete with references to the Convention on Biological Diversity done at Rio de Janeiro on 5 June 1992 as amended; see ss 15B(5) & (6), 15C(9), (10) & (14), 34D(1) & (2), 37G(a)(i), 49A(c), 53(1)(a) & (2)(a), 139(1)(a), 146K(2)(a), 171(4), 303BA(1)(a), 303ER(a)(i), 303GN(1)(a)(i), 305(6)(a) to (c), 324Y(2)(b) & (4), 520(3)(i) and 528. The pre-amble to that Convention notes that the contracting parties:

(a)    are conscious of “the importance of biological diversity for evolution and for maintaining life sustaining systems of the biosphere”, which of course includes us;

(b)    have affirmed that “the conservation of biological diversity is a common concern of humankind”; and

(c)    are aware “that conservation and sustainable use of biological diversity is of critical importance for meeting the food, health and other needs of the growing world population, for which purpose access to and sharing of both genetic resources and technologies are essential”.

579    So, biodiversity is not just some intellectual fancy or aesthetic pleasure. It is important to the sustainability of humans. Now it is true that the Act does not say much about human health and safety, which are dependent upon biodiversity and the protection of the environment. But its omission hardly tells against its relevance.

580    Further, I would make the point that even if his Honour was incorrect on his focus or identification of statutory purpose, that does not take the Minister far. Indeed, let it be assumed that the protection of people and communities is not a purpose of the Act. To so assume does not entail incoherence between the posited duty of care and the statutory framework. Both can sit together, even if they have different purposes and perspectives. To demonstrate different purposes and perspectives does not entail direct or indirect inconsistency. In other words, it is not sufficient to show that they are different. It must be shown that they cannot sit together in any coherent fashion. Test it this way. Assume everything the Minister says about statutory function and purpose is correct. Is it seriously suggested that a duty to take reasonable care to avoid injury, say to those in the physical vicinity of a project, could not sit by side with the function of deciding whether to approve a project for asbestos mining or chemical manufacture? That is not our case of course. But this hypothetical example points out a flaw in the Minister’s analysis on this aspect.

581    Generally, one can accept everything the Minister says, but respond: so what? To me, the Minister’s point is better placed in another context. So, it may be said that the relevant statutory powers here were not conferred, unlike the statutory powers in Pyrenees Shire Council v Day (1998) 192 CLR 330 and Crimmins v Stevedoring Industry Finance Committee (1999) 200 CLR 1, to protect persons or communities in terms of safety and avoiding harm. But that observation really goes to other questions such as sufficient closeness and directness, the relationship question and also the control of risk question. The Minister gains little leverage from such an observation in the context of incoherency which I am presently considering.

582    Let me deal with the second area concerning his Honour’s analysis of the nature, scope and purpose of the approval power.

583    His Honour dismissed the contention that the nature of the power exercised, the subject matter of that power and the statutory context required rejection of the alleged duty of care, because to distinguish this case on that basis involved descending into the policy/operational dichotomy. The Minister says that contrary to the primary judge’s approach, consideration of the nature of the approval power, including the requirement to balance competing public and private interests, was an important aspect of the coherence analysis which could not be avoided on the basis that it involved impermissible reliance on the policy/operational dichotomy. That follows, so it is said, because s 136(1) having identified the mandatory (and permissible) considerations relevant to the exercise of the approval power then leaves the balance to be struck to the Minister, who must weigh the competing interests in the exercise of public power. The Minister says that the primary judge should have found that the imposition of a duty of care to avoid causing personal injury to the claimant class would tend to skew resolution of the balancing exercise conferred by the legislature on the Minister, and distort the focus of the Minister’s statutory task.

584    I would say now that I reject this aspect. But it is more convenient to return to this question later when I deal with the policy question. Questions of policy and potential incoherency with the statute are related questions.

585    Let me deal with the third area concerning his Honour’s treatment of human safety as a mandatory consideration. His Honour reasoned that the preservation of human safety and the avoidance of personal injury were societal priorities of such magnitude that it was unlikely they would not be reflected and accommodated in the Act. But as the Minister rightly points out, no authority suggests, as his Honour appears to have concluded, that the protection of human safety is to be implied as a mandatory consideration in every case where the exercise of a power may have even an indirect effect on human safety even if the legislative scheme is not directed to preventing any such effect, unless that implication is positively excluded.

586    Further, his Honour’s finding that human safety was a relevant mandatory consideration in relation to a controlled action which may endanger human safety was not founded on any of the specified considerations in s 136. So, whilst the Act mandated consideration of “social matters” (s 136(1)(b)), his Honour held that human safety sits outside this concept and outside s 136(1) altogether. Nevertheless, his Honour concluded that, faced with a controlled action which poses a real risk to the safety of members of the Australian community, the Minister may be expected to give at least elevated weight to the need to take reasonable care to avoid that risk of harm. That is so even though the reason that the relevant action is controlled action may have nothing whatsoever to do with human safety.

587    Now the Minister says that such reasoning was significant to the primary judge’s conclusion that the posited duty of care “would be in harmony with the statutory scheme in relation to the need to protect the safety of humans” ([408]), to his conclusion that such a duty did not give rise to any process-based impairment upon the exercise of the statutory power under ss 130 and 133 ([409]), and to his conclusion that even if such impairment was made out, it was “outweighed by the consistency between statutory purpose and the duty of care in relation to the avoidance of personal injury to the [claimant class]” ([409]).

588    But the Minister says that such conclusions were erroneous because the primary judge should not have found that human safety was a mandatory relevant consideration under ss 130 and 133, let alone a consideration to which special priority or elevated weight must be given.

589    Now I agree that human safety was not a mandatory relevant consideration.

590    First, the mandatory considerations for decisions under ss 130 and 133 were to be identified from the text, context and purpose of the Act. Yet those matters provide no basis for the conclusion that human safety is a mandatory consideration. Moreover, and as the Minister points out, the only reason that there is any occasion for the Minister to make an approval decision under ss 130 and 133 with respect to the extension project is because it has been determined to be a controlled action under s 75. But that determination was based upon impacts that the project may have had on listed threatened species and ecological communities and on water resources. So, the need for approval of the extension project was unrelated to any potential impact of GHG emissions or any risk of personal injury to the claimant class.

591    Second, s 136 identifies the considerations that are relevant under ss 130 and 133. And to imply a mandatory requirement to consider human safety is contrary to the express prohibition in s 136(5). It may be a permitted consideration to take into account, but it was not mandatory.

592    Third and relatedly, the primary judge ought not to have construed the expression “economic and social matters” in s 136(1)(b) as excluding matters relating to human safety. The ordinary meaning of “social” includes matters relating to society, which can encompass matters of human safety. So, human safety is a matter that the Minister may choose to consider when deciding whether to approve the taking of a controlled action. But this is not an unstated mandatory consideration.

593    Now the Minister says that once it is appreciated that human safety is not a mandatory consideration, the primary judge’s conclusion that the alleged duty of care was in harmony with the statutory scheme and would not impermissibly distort the Minister’s discretion or skew the intended statutory balance cannot stand. I disagree. To so argue, as the Minister does, is a non-sequitur. The conclusion does not follow from the premise. Let me deal with some other arguments raised by the Minister.

594    The Minister says that the imposition of a duty of care on the Minister in exercising the power of approval would also distort the statutory scheme in other respects.

595    First, the Minister says that recognition of a duty would introduce common law standards of foreseeability. But that is in tension with the statutory standards of causation in characterising whether actions have an impact of the relevant kind (ss 82, 136(2)(e) and 527E). According to s 527E(1)(b), an event that is an “indirect consequence” of an action is only an “impact” of that action if the action is a “substantial cause” of that event. But I would reject this point. The definition of “impact” in s 527E, which affects s 136(1)(a), does not create any potential inconsistency. That definition applies in substance to controlled actions and the relevant impacts on matters of national environmental significance under the Act. The duty of care does not affect those matters.

596    Second, the Minister points out, correctly, that there is only a limited capacity to impose conditions on approval. The Minister may impose a condition of approval if it is necessary or convenient for, inter alia, protecting a matter protected by a provision of Pt 3 for which the approval has effect (s 134), but she has no power to impose conditions directed to the prevention of physical harm being caused to the claimant class. There are also cognate limits on the Minister’s powers under ss 143 (variation of conditions), 144 (suspension of approval) and 145 (revocation of approval).

597    The Minister says that to find that there was a duty of care would involve a distortion of the task that the legislature intended the Minister to perform, which was to use conditions on approval to protect matters for which an approval has effect under Pt 3, rather than to attempt to comply with a duty to take reasonable care not to cause harm to the claimant class.

598    Now it may be accepted that the Minister’s condition-making power correlates to s 136(1)(a) rather than s 136(1)(b). But the legislature clearly intended that the Minister could refuse to approve an action that threatened to cause economic and social harm, which was something that the Minister could not take into account under s 75. There is nothing in s 136(1)(b) to indicate that this consideration can only support approval. It may be inferred that in such a case, it was intended that the Minister could unconditionally refuse, and could invite the proponent to refer a proposed action without the consequent harm.

599    I accept that the Minister has no power to tailor conditions on her approval which may seek to ameliorate or deal with the potential circumstances within the ambit of the duty concerning GHG emissions. She lacks power to specifically target conditions to control or ameliorate the risk. Relevantly to such a subject matter she only has a binary choice: approve the project or refuse it. But does this limited optionality entail incoherency? I do not think so. If the risk is so serious, not to approve may be the appropriate course. But if she would have approved if the proponent was subject to the requirement to take ameliorating steps, how can this be dealt with? Perhaps the proponent could have put forward cascading alternative proposals, each conditioned by steps that the proponent would agree to take to reduce risk. Indeed, practically speaking, some of these may be discussed in preliminary discussions with the Minister or her officers before the formal step of approval. Before the Minister could be alternative and fall-back proposals, thereby allowing the Minister to choose the appropriate course from the menu. The Act clearly envisages such an alternatively layered approach.

600    But I do agree that there is some tension between the posited duty of care and the limited optionality concerning the imposing of conditions. But of course this perceived difficulty may only arise at the margin. In most cases, the choice as to whether to approve or not will be clear, assuming the posited duty, without needing to anguish over the lack of an option to approve with conditions targeted at GHG emissions flowing from the project.

601    Third, the Minister says that the posited duty distorts the discretionary evaluation required to be undertaken by the Minister under s 136. The Minister must weigh adverse and beneficial impacts of a proposed action by reference to specified criteria that may be in tension. The Act does not dictate the outcome, except where it does so explicitly in, say, ss 137 to 140A or give any preference to particular considerations. It proceeds on the implicit basis that, except in the case of particular actions that it expressly provides cannot be approved, an action which has a significant impact on matters protected by Pt 3 may nevertheless be allowed to proceed if the Minister decides that is appropriate. The Minister says that to impose on that decision-making a duty to take reasonable care not to approve an action that may contribute to a risk of physical harm to the claimant class is to give decisive priority to a factor not even mentioned in the statute. But in my view the Minister’s submission is an exaggeration to the extent that she suggests that decisive priority must be given.

602    Fourth, the Minister says that the common law duty of care puts pressure on the Minister to refuse approval so as to avoid the risk of exposure to a damages suit. It is said that this distorts the decision-making process under the Act. In my view, the Minister’s approval decision in the present case belies such a suggestion. She gave her approval before the present appeal at a time when his Honour’s declaration was in place.

603    Fifth, and contrary to the Minister’s position, there is no competing duty imposed on the Minister by the Act, such as would require the approval of the extension project regardless of the harm caused to humans. There is no incoherence. The Act does not stand against the Minister giving appropriate weight to a risk of harm to human safety from an action, and refusing approval if that risk, together with any residual impacts on matters of national environmental significance, outweighs the economic and social benefits of the action. Is it seriously suggested that there would be no duty of care regarding the approval of a mining project concerning contamination of ground water or indeed any water resource insofar as it concerned human health? Take another example. If the Minister approved the building and operation of a factory, up-river from a town, knowing that it would leak cyanide that would poison the residents’ water supply and might cause them harm, is it seriously suggested that the legislature intended to deny the residents relief at common law? And is it seriously suggested that there would be no duty of care concerning dust or air pollution that might be caused to those in the vicinity as a result of the approval of a mining project? And is it seriously suggested that there would be no duty of care concerning approvals relating to nuclear actions which might result in the diffusion of radio-active products or waste into the environment?

604    Sixth, it was unclear at one point whether the Minister was saying that in the exercise of her approval power, any duty of care to avoid personal injury was incoherent with the Act or whether she was just saying that in respect of the present posited duty of care to the claimant class. If the former, then such a position is quite untenable. But if the latter only, then the situation is a little more nuanced but nevertheless still substantially unpersuasive on the incoherency ground.

605    Seventh, in favour of her incoherence thesis, the Minister argued that it was merely fortuitous that the extension project came before the Minister for approval. At first blush, this point sounded quite plausible. But on reflection, I consider there to be nothing to it.

606    Now by way of background I note that the then proposed action to construct and operate the original open cut coal mine was determined on 17 May 2012 not to be a controlled action, provided that, inter-alia, fencing and signposting was installed for a patch of winged peppercress plants and there was a translocation of about 50 such plants into that fenced area. So, as the then proposed action was not substantively considered to be a controlled action, no substantial approval was required for the project. Contrastingly, on 14 August 2016 the extension project was determined to be a controlled action, and required assessment under the bilateral agreement between the Commonwealth and New South Wales. The Minister says that it would be incoherent to find a duty concerning GHG emissions for some projects which required approval, whereas there would be no duty for other projects, indeed the original project here, where no approval was required even though they may produce problematic GHG emissions.

607    Now in a general sense this argument, although intuitively appealing, is a distraction. One is considering the exercise of power, which has been enlivened, to approve a particular project before the Minister. The question is, in that context, whether there is a duty. Test the matter this way. Say before the Minister is an application to approve an asbestos mine on Commonwealth land and the question arises as to whether a duty is owed to persons in the nearby township concerning asbestos contamination in the air and ground water. Is it seriously suggested that no duty is owed because there may be other asbestos mine projects that may not require the Minister’s approval? Now of course we have a different case where the project here is one of numerous “feed-stocks” of GHG emissions to the overall risk. And of course, projects involving such other feed-stocks may not come across the Minister’s desk so to speak for approval. But such a difference does not support any incoherency argument. Rather such a difference feeds into questions concerning reasonable foreseeability, whether at the duty or breach stage, breach more generally, or causation.

608    Whether a duty is owed by the Minister concerning the approval of a particular project is to be looked at when her power has been enlivened in the individual circumstances and context before her, including what she knows or ought to have known or could reasonably contemplate. It is not to be considered by reference to projects not before her or which she has no power to approve, except to the extent that they may provide context for the particular approval process that has been invoked before her.

609    In summary, whilst there is a modicum of incoherency, I do not consider that this is a strong feature against recognising a duty of care, particularly in a third scenario case as I identified at the outset.

Is high policy involved? And should it trump the duty?

610    The Minister says that the decision as to whether the extension project should be approved is a discretionary decision involving high policy considerations made according to a process of evaluation that is prescribed by the Act. Further, she says that the responsibility for exercising the discretion personally lies with her. These two features are said to trump the primary judge’s assessment that a duty of care should be recognised.

611    Now it is not in doubt that the exercise of the Minister’s discretionary power to approve is likely to involve policy questions. But the posited duty of care is not to be denied by the policy points proffered by the Minister. Before turning to the principal authority that she has prayed in aid, namely, Graham Barclay Oysters, let me make the following general points.

612    First, the present context is not in the realm of a scenario where the legislation explicitly or implicitly enshrines one policy and the posited duty of care is inconsistent or in tension with such a policy; Graham Barclay Oysters provides such an example, and in any event is not a third scenario type case.

613    Second, the present context does not deal with a scenario where the relevant statutory power and its exercise could be characterised as being quasi-legislative. In that regard, ss 130, 133 and 136 may be contrasted with provisions such as ss 33, 34A, 37A and 37B of the Act.

614    Third, the Minister could delegate her statutory powers. There was no need for her to personally exercise them.

615    Fourth, the present context is not in the realm of what some authorities label as “core” policy, which in my view is an unhelpful characterisation in any event. Reference has been made to the notion that a core policy decision is protected from negligence liability or is not justiciable. But what is meant by “core”? Is this to be in contrast with “operational”? If so, the label adds little. One is back to the policy/operational dichotomy. Further, to say that something is non-operational still does not tell you what a policy decision is, let alone a core policy decision. And are we talking about “core” by reference to subject matter, the level of the decision-maker or the degree of significance of the policy to the question of whether the statutory power should be exercised or not?

616    Of course, an anterior question also arises. What do we mean by policy? Now policy is to be distinguished from discretion. A discretion connotes a choice as to whether to act or not. But to have such a choice may have nothing to do with policy. So, a discretion involves a broader class. A policy choice is a subset of the broader class. But again, what do we mean by policy? Now a policy decision may be one informed by social, economic or political considerations. Such a decision may involve the balancing of those considerations and having some public aspect to them. But is something more required? Does a reference to policy connote generality, that is, something of general application albeit to be applied to the specific case in question? Not necessarily. If a policy may be thought to be a principle or course of action adopted or proposed as necessary or desirable to achieve a particular end, this need not be something of general application. If it is, it is a policy. But it may be a policy decision even if one is not applying a principle or course of action of general application. For present purposes I will accept that the taking into account and weighing of social, economic or political considerations may make something a policy decision, without some general principle or tenet being applied. But to accept that this is so still provides no meaningful content to the label “core”.

617    Now does the governmental level at which a decision is being made tell you whether it is a core policy decision? Not necessarily, although the higher the level the more likely it may approach or be suitable for such an epithet. The requirement or responsibility to balance and weigh public policy considerations, including those that may be in tension, are usually reposed in those with higher and broader authority. So, the Minister here was making the decision, although it could have been delegated. But in any event the fact that the Minister was making the decision does not define away the inquiry.

618    Why do I say that we are not in the realm of what some label as “core” policy? There are various points.

619    The first point is that a quasi-legislative decision can be thought of as a core policy decision, but on no view in the present case could it be said that the Minister’s exercise of power was quasi-legislative.

620    The second point is that even if it be accepted that we are not on the operational side of the so-called policy/operational dichotomy or we are not concerned with budgetary or expenditure questions, that hardly assists the Minister. Of course we are not concerned with such matters because we are not dealing with a failure to act scenario. The present context is a third scenario type case rather than a first or second scenario type case.

621    The third point is that to say that the Minister has to take into account economic and social matters does not establish that one is in a core policy area as such. Section 136(1)(b) comes after s 136(1)(a) and does involve policy, although not expressly GHG emissions. Moreover, GHG emissions are hardly central to or the very essence of the statutory inquiry concerning approval. They are not core to the Act. Moreover, the Minister can delegate her power.

622    The fourth point involves a shift in emphasis. Let it be assumed that the concept of core policy is useful. But let it be assumed that one is not dealing with a core policy enshrined in or referable to some aspect of the Act, whether consistently with various statutory provisions or in tension with them. Can you say, nevertheless, that by reason of how the posited duty has been framed, core policy questions will arise? Now here I have several difficulties. How do you assess what “core” is once you have moved away from the statutory provisions? And how will you know what core policy questions to consider and weigh until the time for the exercise of the statutory power arises and you are engaged in that very exercise? Perhaps you can confidently project in advance. But why would you as the basis for denying a duty, particularly in a third scenario type case? Such policy questions, if they arise, can adequately be addressed at the stage of breach. Indeed, the present case so demonstrates. The Minister has given her approval to the extension project notwithstanding the primary judge’s posited duty. No doubt she had little difficulty in weighing up the relevant policy considerations and giving her approval irrespective of the posited duty.

623    Let me now turn to Graham Barclay Oysters. Contrary to the Minister’s submissions, Graham Barclay Oysters does not strongly support the Minister’s contentions. Gummow and Hayne JJ (Gaudron J agreeing) provided the following important context (at [175] and [176]):

Lindgren J explained in his reasons for judgment that Div 4 of Pt 2 of the Management Regulation reflects a political decision by the State to enlist shellfish industry participants in a system of industry-funded self-regulation or co-regulation, rather than to impose on that industry a publicly funded regulatory regime. In particular, the State decided not to adopt the approach of some other Australian and foreign jurisdictions which require regular sanitary surveys of oyster growing regions pursuant to a classification structure based on water pollution levels. This decision was reached after much consideration and was based in part on budgetary concerns. In accordance with that decision by the Executive Government of New South Wales, which found partial expression in the Regulations referred to above, the State neither required regular sanitary surveys of oyster growing areas (whether as a condition of aquaculture leases or permits or otherwise) nor undertook to conduct such surveys itself. A decision of that nature involves a fundamental governmental choice as to the nature and extent of regulation of a particular industry. It is in a different category to those public resource allocation decisions which, in the manner described in Brodie v Singleton Shire Council, may be considered in determining the existence and breach of a duty of care by a public authority.

Once the nature of the decision by the State is appreciated, its observance by agents of the State in respect of any particular region falls outside the scope of any common law duty of care that may otherwise arise…The scope of any common law duty that may arise in those circumstances necessarily accommodates itself to, and is controlled by, the insusceptibility of that decision by the State to curial review under the rubric of the tort of negligence. It follows that the State was under no common law duty to conduct sanitary surveys of Wallis Lake.

624    Clearly, what Gummow and Hayne JJ were discussing as the policy question is quite different to the present case. Further, what Gleeson CJ said must be seen in the context of what he was required to consider. He was considering scenarios concerning non-feasance or a failure to act. Indeed, he stressed this point at the outset by saying (at [5]):

…The allegations now pressed against the State, and the Council, do not involve allegations of carelessness in the exercise of a statutory power. The complaint is not about acts, but about omissions. In the particular circumstances of the case, the issues, raised by this assertion of direct governmental liability in negligence, include what are, in the final analysis, issues of justiciability.

625    That explains the context for what he said at [6] concerning public funds, scarce resources etc. And again at [7] concerning the setting of priorities. Broader statements in [6] and [7] concerning courts passing judgment on the reasonableness of government action or inaction need to be read in context. Complaints may be “political in nature” where it is said that the Government has not acted or has not done enough. But the situation is different where the allegation is one of carelessness in the positive exercise of a statutory power.

626    Further, at [9], he said:

…In the case of a governmental authority, it may be a very large step from foreseeability of harm to the imposition of a legal duty, breach of which sounds in damages, to take steps to prevent the occurrence of harm. And there may also be a large step from the existence of power to take action to the recognition of a duty to exercise the power. Issues as to the proper role of government in society, personal autonomy, and policies as to taxation and expenditure may intrude…

627    At [15], he said:

Here we are concerned with the problem of deciding, in a case where the government had certain powers, whether it is accountable, through the law of negligence, for not exercising its powers, or for not exercising them sufficiently. To apply that form of legal accountability requires the identification, not merely of a power, but also a duty; a duty of care owed to a citizen or a class of citizens. A conclusion that such a duty of care exists necessarily implies that the reasonableness or unreasonableness of the inaction of which complaint is made is a legitimate subject for curial decision. Such legitimacy involves questions of practicality and of appropriateness. There will be no duty of care to which a government is subject if, in a given case, there is no criterion by reference to which a court can determine the reasonableness of its conduct. That negative proposition leaves open other questions as to the circumstances in which the law will treat failure on the part of a public authority to exercise a power as a breach of a private law duty of care; but it is sufficient to resolve a substantial part of the case against the State in these proceedings.

628    The context in terms of the policy question is all important. Clearly, Gleeson CJ was dealing with a first or second scenario case that I described at the outset of my reasons. But a failure to act scenario and the policy questions arising in that context are quite different from the scenario where the repository of the statutory power by its positive exercise has given rise to a foreseeable risk of harm. But in any event, an even narrower policy context was being considered by Gleeson CJ as he went on to explain. Relevantly to the context that he had to decide, at [26] and [27] he said:

Other evidence to like effect is set out in the reasons of Lindgren J. It demonstrates that the nature and extent of State government involvement in oyster quality control was a matter of policy, that it received attention at the highest levels, that it had substantial budgetary implications, and that it involved government concern to encourage an important primary industry.

This demonstrates two things. First, the proposition that the State government had substantial managerial control over the oyster industry is, at best, an over-simplification. Secondly, the proposition that the State had a legal duty of care, owed to oyster consumers, obliging it to exercise greater control (and, presumably, to permit less industry self-regulation) takes the debate into the area of political judgment. By what criterion can a court determine the reasonableness of a government’s decision to allow an industry a substantial measure of self-regulation?

629    So, what Gleeson CJ was discussing as the policy question that he had to address is quite different to the present case. The posited duty of care was in tension with the policy choice made by the NSW legislature concerning industry self-regulation of the oyster industry.

630    Further, in Brodie v Singleton Shire Council (2001) 206 CLR 512, it was said by Gaudron, McHugh and Gummow JJ (at [106]):

Appeals also were made to preserve the “political choice” in matters involving shifts in “resource allocation”. However, citizens, corporations, governments and public authorities generally are obliged to order their affairs so as to meet the requirements of the rule of law in Australian civil society. Thus, it is no answer to a claim in tort against the Commonwealth under s 75(iii) of the Constitution that its wrongful acts or omissions were the product of a “policy decision” taken by the Executive Government; still less that the action is “non-justiciable” because a verdict against the Commonwealth will be adverse to that “policy decision”…

631    These observations were also accepted by Kirby J.

632    So, apart from some general statements by Gleeson CJ which were broader than the context dealt with in Graham Barclay Oysters and not expressly endorsed by others at that level, that case does not strongly point against the posited duty because of matters of high policy.

633    In summary, I accept that policy questions are involved. But whatever they may be, they can adequately be dealt with at the breach stage. In my view, where the Minister may create a danger by exercising her statutory power in favour of approving a project, policy is no answer to denying the duty unless the Act itself makes such policy questions so fundamental to the exercise of statutory power that such a conclusion is compelling. I am not so compelled to find here.

Control of risk

634    The Minister made much of the point that she did not control the relevant risk. But before dealing with her contentions, it is necessary to begin with some general observations.

635    First, the feature concerning control of the relevant risk is usually discussed where there has been a failure to exercise a statutory power or an inadequate exercise of that power. What is in essence being said is that because you controlled the risk, you should have done something about it or something more than you did to avoid or minimise it. But where the relevant statutory entity or officer created the risk of harm by the positive exercise of power, to talk of control of risk is a little artificial. The risk was not just controlled. It was in fact brought about by the entity or officer. What is then said, and of course more directly, is that you should not have created the risk or at least had the plaintiff in contemplation when you created the risk.

636    Second, and following on from the first point, the real question is what risk was created by or would be created by the Minister in approving the extension.

637    Third, the Minister spent much time in taking us to cases such as Agar and Graham Barclay Oysters concerning the control of risk question for the purpose of establishing, by analogy, that the Minister did not control the relevant risk here. But we are dealing here with a hypothesised positive act, here approval of the extension project, potentially leading to possible harm. As the plurality in Agar said (at [68]), part of which I have already set out:

Further, from the earliest times, the common law has drawn a distinction between a positive act causing damage and a failure to act which results in damage. The common law does not ordinarily impose a duty on a person to take action where no positive conduct of that person has created a risk of injury to another person.

638    As they went on to say (at [69]):

Here the appellants were members of the IRFB, an institution which “saw itself as the law-giver for the sport of rugby”. But they have done nothing that increased the risk of harm to either of the respondents. The complaint is that they failed to alter the status quo, failed to alter the rules under which the respondents voluntarily played the game…

639    That was important context for what was then said (at [81]):

It follows that in no relevant sense did the Board of the IRFB, or those who attended its meetings as delegates, control what happened in the matches in which the respondents were injured. The IRFB did not organise either of these matches. It did not decide whether the laws of the game which it promulgated would be adopted in these matches. The highest point to which the respondents’ contentions could rise was to assert that the IRFB “influenced” the way in which rugby football would be played in Australia. But it is not arguable that the influence amounted to control over the sport: at least at the level at which the respondents played. In particular, they were not subject to any legal control by the IRFB or the delegates to its meetings. Nor can it be argued that they were subject to control in any practical sense. There were too many intervening levels of decision-making between the promulgation by the IRFB of laws of the game and the conduct of the individual matches in which the respondents were injured. What happened depended to a greater or lesser extent upon the several decisions of the national union, the local union and the association which organised the competition and on the decisions of the referees who acted in those matches.

640    But here we are dealing with a hypothesised positive act being the start of a causal chain leading, it is hypothesised, to scope 3 emissions, temperature increase and potential harm. Of course other actors would be involved, indeed are necessary for this causal chain. But to have in a hypothesised causal chain numerous actors each of which control only a very small part of the risk referable to their particular conduct, does not entail that none of them owe any duty for the foreseeable consequences of their conduct because no one actor controls the risk.

641    Further, there is no analogy with Agar concerning the Board and the conclusion that they did not control matches; moreover, no one member of the Board could control anything on any view. Here, the Minister’s act of approval is a positive step in the causal chain giving rise to the relevant risk. Indeed, the need to obtain approval controlled whether the project went ahead. The Minister knew that, and of the likely consequences that scope 3 emissions would be produced by her act of approval. She controlled the trigger. Of course there were “intervening levels of decision-making” between approval and emission, but not of the variability and uncertainty that was contemplated in Agar between the Board and a rugby match. The likely decision-making at each of the main levels was within her contemplation if not expectation.

642    Likewise, the present context is not analogous with Graham Barclay Oysters. Again, that was in essence a failure to act case. But it was contended in that case that the State exercised managerial control over the Wallis Lake oyster industry, and that the local council also controlled the risk of harm. Now Gummow and Hayne JJ said (at [150]):

The factor of control is of fundamental importance in discerning a common law duty of care on the part of a public authority. It assumes particular significance in this appeal. This is because a form of control over the relevant risk of harm, which, as exemplified by Agar v Hyde, is remote, in a legal and practical sense, does not suffice to found a duty of care.

643    And in terms of the council they said (at [152] to [154]):

The Council in the present appeal, by contrast, exercised a much less significant degree of control over the risk of the harm that eventuated. At no stage did the Council exercise control, let alone significant or exclusive control, over the direct source of harm to consumers, that is, the oysters themselves. It may be that the predominantly land-based sources of pollution were all ultimately subject to Council control. That, however, is the start, not the end, of the inquiry. Control over some aspect of a relevant physical environment is unlikely to found a duty of care where the relevant harm results from the conduct of a third party beyond the defendant’s control. Modbury Triangle Shopping Centre Pty Ltd v Anzil illustrates the point. What is significant here is the extent of control which the Council had over the risk of contaminated oysters causing harm to the ultimate consumer; control in that sense is not established by noting the Council’s powers in respect of some or most of the sources of faecal pollution.

As Lindgren J observed in the Full Court, the relationship between the Council and the oyster consumers is indirect; it is mediated by intervening conduct on the part of others. Between the Council on the one hand and the oyster consumers on the other, there stands, in the present case, an entire oyster-growing industry comprising numerous commercial enterprises, each of which, in pursuit of profit, engages in conduct that presents an inherent threat to public safety. That threat arises from the insusceptibility of oysters to effective and reliable tests to identify contamination of the type that eventuated here.

In broad terms, the Council’s statutory powers enabled it to monitor and, where necessary, to intervene in order to protect, the physical environment of areas under its administration. However, the conferral on a local authority of statutory powers in respect of activities occurring within its boundaries does not itself establish in that authority control over all risks of harm which may eventuate from the conduct therein of independent commercial enterprises. As the course of this litigation itself indicates, control over the safety of the Wallis Lake oysters for human consumption has been fragmented. The conduct of the Council did not “so closely and directly [affect]” oyster consumers so as to warrant the imposition of a duty of care owed by the former to the latter. There were “too many intervening levels of decision-making” between the conduct of the Council and the harm suffered by the consumers. As the trial judge noted, the Council had no direct responsibility for the operation of the oyster industry or the quality or safety of Wallis Lake oysters. It did not control the process by which commercial oyster growers cultivated, harvested and supplied oysters, nor the times or locations at which they did so. The Council has not been given, by virtue of its statutory powers, such a significant and special measure of control over the risk of danger that ultimately injured the oyster consumers so as to impose upon it a duty of care the breach of which may sound in damages at the suit of any one or more of those consumers.

644    But again, in the case before us it is hypothesised that the Minister will create the risk or increase the risk of harm to the respondents by the positive act of approval, an important distinction (see McHugh J at [81] and [91]).

645    Further, Brodie is another example where the significance of the “control of risk” was coupled with a scenario where it was said that there was a duty to act. So, Gaudron, McHugh and Gummow JJ explained (at [102]):

…Whatever may be the general significance today in tort law of the distinction between misfeasance and non-feasance, it has become more clearly understood that, on occasions, the powers vested by statute in a public authority may give it such a significant and special measure of control over the safety of the person or property of citizens as to impose upon the authority a duty of care. This may oblige the particular authority to exercise those powers to avert a danger to safety or to bring the danger to the knowledge of citizens otherwise at hazard from the danger. In this regard, the factor of control is of fundamental importance.

646    Further, the scenario in Pyrenees is quite different from the present context. Gummow J laid out the smorgasbord of statutory powers that the local council had to address fire risks and to prevent, deal with or ameliorate fire hazards (at [146] to [154]). Clearly they had a purpose or function involving, inter-alia, human safety and the protection of property. He then said (at [168]):

…In May 1990, the situation occupied in relation to this litigation by the Shire as the arm of local government gave it a significant and special measure of control over the safety from fire of persons and property in Neill Street. Such a situation of control is indicative of a duty of care. The Shire had statutory powers, exercisable from time to time, to pursue the prevention of fire at No 70. This statutory enablement of the Shire “facilitate[d] the existence of a common law duty of care”, but the touchstone of what I would hold to be its duty was the Shire’s measure of control of the situation including its knowledge, not shared by Mr and Mrs Stamatopoulos or by the Days, that, if the situation were not remedied, the possibility of fire was great and damage to the whole row of shops might ensue. The Shire had a duty of care “to safeguard others from a grave danger of serious harm”, in circumstances where it was “responsible for its continued existence and [was] aware of the likelihood of others coming into proximity of the danger and [had] the means of preventing it or of averting the danger or of bringing it to their knowledge”.

647    Of course, the “statutory enablement of the Shire” in that case strongly supported the duty. But the stronger point was that the council had “a significant and special measure of control over the safety from fire of persons and property in Neill Street”.

648    Now of course this “control of risk” aspect was not in the context where the positive act of the council had created the risk, which by analogy is our case. In other words, “control of risk” is usually coupled with the concept of a failure to act or negligently acting, say an omission in the course of a positive act but not going far enough to address the risk. Contrastingly, you don’t usually refer to or require “control of risk” where the putative tortfeasor has in fact created the posited risk by a positive act.

649    I also note that the question of control of risk dealing with the criminal behaviour of third parties was discussed in Modbury Triangle Shopping Centre Pty Ltd v Anzil (2000) 205 CLR 254. But the very context of that case concerned whether liability should be imposed for an omission. The case was all about whether there was a duty to act to prevent injury being done to another by the conduct of a third party. It was not about whether the defendant owed a duty to take reasonable care to avoid doing what might cause or increase the risk of harm to another.

650    Let me turn to the Minister’s specific arguments.

651    The primary judge found that the Minister had substantial control over the real risk of harm to the claimant class that would flow from her approval of the extension project. But the Minister says that that finding significantly overstates the extent of the Minister’s relevant control.

652    Further, the Minister says that a form of control that is remote or indirect will not suffice. Control is not simply the capacity to affect in some way the risk that harm arises. Such a diffuse notion of control would be little more than a description of the test of reasonable foreseeability. The Minister says that what is significant is a measure of control over the relevant harm that is, by its nature and extent, indicative of a relationship where it is normatively appropriate to recognise a duty of care arising by reason of that relationship.

653    The Minister says that she has little, if any, control over the risk of death and personal injury from heatwaves or bushfires that may be caused by climate change. She says that the extent to which that risk can be mitigated or eliminated depends on many different actions or factors, including the extent of co-ordinated global action to address climate change over the coming years and decades.

654    The Minister says that the limited form of control that she has over any future harm from GHG emissions is too remote and indirect to support the recognition of a duty of care. She says that in exercising the power to approve or not approve the extension project, she has power to prevent an action that will make at most a tiny contribution to the risk of harm. Further, the risk of harm relied upon by the respondents will not be a product of the Minister’s conduct in approving the extension project per se, but the accumulation of atmospheric CO2 from all sources.

655    The Minister says that she has control only in the limited sense that she has power to grant an approval for the extension project, which even if it proceeds will make a tiny contribution to the identified risk of harm. And even if approval is granted, the Minister thereafter has no capacity to control the way in which the activity is carried out, subject to a confined power to impose conditions of approval and the power to enforce compliance. Nor does she have exclusive control. And she says that whether the coal from the extension project is consumed, and in what circumstances, depends on the downstream decisions and actions of other people.

656    Generally, the Minister says that the primary judge ought to have found that there were too many intervening levels of decision-making between the conduct of the Minister in approving the extension project and any physical harm that may be suffered by the claimant class at an unknown time in the future to impose a duty of care on the Minister to try to avoid those harms. And it is said that control of the remote kind that the Minister has is insufficient to justify an exception to the usual principle that a person is not responsible in negligence for the conduct of a third party, even where that conduct is foreseeable.

657    I do not accept many of the Minister’s arguments concerning control of risk.

658    In the present case, it is the posited and positive act of approval by the Minister that it is said will ultimately lead to harm. Now of course the acts of third parties will be part of the causal chain. But it is the very act of the Minister that ignites the hypothesised causal chain. Rather than hindering, controlling or preventing the acts of third parties, the Minister is facilitating and intentionally or at least knowingly fostering such acts.

659    Now for a proposed duty to act, which is not the present context, the authority’s control over the risk must be very significant. And in such non-feasance cases, where statutory powers exist which may be exercised to minimise or increase the risk, this may be indicative of a duty. An example is where a public authority has statutory powers that enable it to prohibit or regulate activity in a particular location. But in cases such as the present where the positive exercise of statutory power is said to create the danger, control is more readily seen as sufficient to impose a duty of care to exercise that power with reasonable care.

660    Further, control need not be exclusive. One or more entities may have control to varying degrees over the risk of harm. And a duty may be imposed on a public authority even if a private entity has more control than such an authority.

661    Now the autonomy of actors in a causal chain may point against the existence of a duty. For example, in Agar, injured rugby players were unable to establish that the defendants owed them a duty of care to alter the rules of play so as to reduce their exposure to unnecessary risk, because they and those who injured them autonomously engaged in the risky activity. But unlike adult rugby players, many of the claimant class in the present case will have no control in respect of the foreseeable harm that is coming.

662    Further, that a third party, the owner of the mine, also has control over the risk does not deny the Minister’s control. This is a positive act case. The Minister has authorised the owner to extract the coal for the very purpose of exporting it for combustion and emission as CO2. I agree with the respondents that the Minister cannot deny control because the owner is likely to do exactly what she has authorised it to do. The owner will extract and export the coal for burning, precisely as it proposes, with the Minister’s knowing and necessary approval. Without it, it will not.

663    Generally, the Minister has made much play of her “many actors” point and the observation of Gummow and Hayne JJ in Graham Barclay Oysters (at [145]):

…As will appear, the common law should be particularly hesitant to recognise such a duty where the relevant authority is empowered to regulate conduct relating to or impacting on a risk-laden field of endeavour which is populated by self-interested commercial actors who themselves possess some power to avert those risks.

664    Decontextualised from its setting, this observation had a superficial attraction. But its setting was what Gummow and Hayne JJ described at [153], which I have set out earlier. This was all in the context of a failure by the council to take any or any sufficient steps. The case was not about positive action of the council that gave rise to a risk of harm. Further, in the present case, the Minister would well know of the commercial steps that were likely to occur following approval. Indeed, her approval was predicated on an expectation of such steps. Of course, she had no control or say over other emissions from other players, but that is a separate question going to reasonable foreseeability at the levels of both duty and breach, the question of breach generally and causation.

665    Now Gleeson CJ said (at [38]):

In considering the powers and responsibilities of the Council, for the purpose of determining whether it owed a duty of care to oyster consumers, an aspect of the facts should be noted. Wallis Lake is large, and there were many different ways in which, and places at which, human activity on or around the lake could result in pollution of its waters. There was no particular place of pollution that was shown to be responsible, or mainly responsible, for contamination of the oysters. As Lindgren J pointed out, assertions of a duty to reduce or minimise pollution are difficult to give practical content of relevance to the harm suffered by Mr Ryan. As with the State, the complaint is that the Council did not do enough to reduce pollution, but it is not possible to point to any specific act or omission that would have prevented harm to Mr Ryan.

666    But in my view, the fact that there are many actors goes nowhere. We are dealing with the Minister and any role she might play in creating or exacerbating the relevant risk.

667    It is convenient to address one other but different point here. The Minister made what was little more than a rhetorical if not in terrorem submission that if she owed a duty of care, then other authorities, entities or individuals might also owe a duty of care for GHG emissions. This argument went nowhere. Perhaps she may be correct, perhaps not. So what? If it is otherwise appropriate to find that she owed a duty of care, she should not be permitted to shirk her, as she would have it, personal responsibilities by such a distraction.

668    In summary, I do not accept that the Minister did not, in a relevant sense, control the risk flowing from her approval.

Vulnerability

669    The primary judge held that the vulnerability of the claimant class to the harm that may result from the impugned conduct of the Minister was one of the affirmative indicators of a duty of care. But the Minister complains that there was a lack of clarity as to what was meant when referring to the claimant class’ vulnerability. It is said that his Honour used the term to refer to the increased likelihood that the claimant class will suffer harm because they will live through a period in which the effects of climate change are more likely to be experienced. But the Minister says that those observations do not establish relevant vulnerability as a salient feature.

670    I would agree with the Minister that the primary judge has blended several concepts that should be kept discrete.

671    It is important to be clear what one is focusing on when addressing vulnerability. It is not any generalised sense of vulnerability to climate change. And it is not vulnerability to high temperatures brought about by GHG emissions. Rather, it is vulnerability to the relevant harm, here personal injury, potentially causally connected to any breach of duty, where members of the claimant class could not reasonably be expected to adequately safeguard themselves from such harm (Crimmins at [93(3)] per McHugh J), that is, personal injury rather than climate change.

672    It is vulnerability in this narrower sense that is being considered, and is particularly relevant where a defendant has brought about or exacerbated the relevant risk or at the least been in a position to control the risk (Graham Barclay Oysters at [149] per Gummow and Hayne JJ and Perre v Apand Pty Ltd (1999) 198 CLR 180 at [127] per McHugh J).

673    Now reference has been made to the concept of a plaintiff being especially vulnerable. And such an intensifier has been used from time to time, particularly and rightly by McHugh J in Crimmins. Of course, vulnerability is not focusing on members of the public generally or everyone who may be alive in the relevant period who may be exposed to climate change. And so, the target of especially re-inforces the notion that one is focused and must be focused on a narrower set of individuals. But what narrower set?

674    In my view, many members of the claimant class will likely be vulnerable to a real risk of personal injury from events causally related, on this hypothesis, to the Minister’s breach of the posited duty of care, in the sense that they are unlikely to be able to protect themselves from such personal injury. But to be clear, it would not be all members of the claimant class. All members of the claimant class will be subjected to climate change and its potential effects. But to pitch the concept of vulnerability at that level is much too broad. It is only some members of the claimant class who will be vulnerable in the relevant sense, that is, to personal injury that they cannot protect themselves against.

675    Of course, all members of the claimant class will generally be subject to the physical and economic effects of climate change. None may be able to escape that. But most will not be vulnerable to personal injury that they cannot reasonably protect themselves against, for example, those living in major cities on the Eastern seaboard. That is the focus, rather than a more diffuse and generalised concept of vulnerability.

676    In summary, the primary judge was partially correct in his conclusion on vulnerability. Many members of the claimant class are or will be relevantly vulnerable to personal injury. But in my view one cannot reasonably ascertain their parameters or likely number today. Who knows how or where members of the claimant class will be living in 80 years and the then technological and adaptive ingenuity of humanity to deal with the consequences flowing from any increases in temperature? It is not sustainable to say that all members of the claimant class today may be at a real risk of personal injury in 80 years in a manner that they cannot reasonably protect themselves against.

677    It is convenient to dispose of one other matter at this point. The Minister says that his Honour’s reasoning was not improved by the conclusion that the Minister is in a special protective relationship with the claimant class by virtue of her role within the Commonwealth Executive. I agree with the Minister that no constitutional or other legal principle or authority supports the notion that there is such a special protective relationship between all members of the Executive and all Australian children. That point goes nowhere on the question of vulnerability or the next topic that I should address.

Special relationship/sufficient closeness and directness

678    Let me at this point deal with three inter-related questions. Was there a relevant relationship between the Minister and the claimant class? Relatedly, is the statutory power in question for the benefit of a particular or indeed any class of persons including the claimant class? And more generally, was there a sufficient closeness and directness between the Minister and her exercise of statutory power and the potential effects on the claimant class?

679    Of course, the answers to the first and second questions are clear. There was no such relationship, except in an indirect causal sense. But that is not sufficient. Moreover, and related to the question of any relationship, the statutory power was not for the benefit of or focused on any class of persons including the claimant class.

680    But given that I am dealing with a third scenario case as I discussed at the outset, I would prefer to decide the case by answering the third question, although the answers to the first and second questions inform the answer to the third question. Let me explain.

681    The Minister says that nothing about the applicable statutory scheme created any relationship between the Minister and the claimant class. But it may be queried whether a special relationship need be shown between the Minister and the respondents. After all, the present context does not involve an asserted failure to act or a failure to go far enough. Here it is said that if there is a positive exercise of power then a risk of harm may be created.

682    Now in the context of a failure to exercise a statutory power, a relationship may be required to be gleaned from the statutory regime. So, in Stuart v Kirkland-Veenstra, Gummow, Hayne and Heydon JJ said (at [112] and [113]):

…Does that regime erect or facilitate “a relationship between the authority [here the holder of statutory power] and a class of persons that, in all the circumstances, displays sufficient characteristics answering the criteria for intervention by the tort of negligence”?

Evaluation of the relationship between the holder of the power and the person or persons to whom it is said that a duty of care is owed will require examination of the degree and nature of control exercised over the risk of harm that has eventuated, the degree of vulnerability of those who depend on the proper exercise of the relevant power, and the consistency or otherwise of the asserted duty of care with the terms, scope and purpose of the relevant statute. Other considerations may be relevant.

683    But of course that was in the context where it was said that there was a duty to act. That is not our case. Further, the observations of Crennan and Kiefel JJ must also be seen in context. They accepted (at [131]) that “the statutory powers in question must be directed towards some identifiable class or individual, or their property, as distinct from the public at large.” And they recognised (at [135]) that a “relationship might be seen to arise when an authority has commenced exercising its powers towards a class of individuals”. The question of a relationship is tied in with the question of whether the powers are for the protection of a particular class. But this is a question more relevant to a first or second scenario type case. What relationship existed or what was it about the relationship that required the repository of the statutory power to act or go further? But you are not asking such questions where a positive act or exercise of power has created the risk or given rise to the harm. Here, of course, it may be accepted that the relevant power was not for the benefit of a particular or indeed any class of persons.

684    In Graham Barclay Oysters, Gleeson CJ said (at [39]):

The powers conferred upon the Council, insofar as they are presently relevant, were conferred for the benefit of the public generally; not for the protection of a specific class of persons... But there is nothing in the relevant statutory provisions, or in the circumstances concerning the relationship between the Council and oyster consumers, to justify a conclusion that the Council’s powers were given for the protection of oyster consumers, or any other particular class.

685    He also said (at [32]):

…The power given by s 189 is a power to protect the public, not a specific class of persons. Similar powers, covering a wide range of activities, are given to Ministers and government authorities in the interests of public health and safety. A legislative grant of power to protect the general public does not ordinarily give rise to a duty owed to an individual or to the members of a particular class.

686    McHugh J said (at [82]):

The likelihood of the common law imposing an affirmative duty of care whose content may require the exercise of a statutory power increases where the power is invested to protect the community from a particular risk and the authority is aware of a specific risk to a specific individual. If the legislature has invested the power for the purpose of protecting the community, it obviously intends that the power should be exercised in appropriate circumstances…

687    Of course that is not our case. But at [91], he said:

…Unless a particular exercise of power has increased the risk of harm to an individual, the Executive government of a polity does not ordinarily owe any common law duty to take reasonable care as to when and how it exercises its powers…But such cases are less likely to arise than in the case of other public authorities. In particular, knowledge of specific risks of harm or the exercise of powers in particular situations is less likely to be a factor in creating a duty than in the case of an ordinary public authority. This is because the powers and functions of the Executive government are conferred for the benefit of the public generally and not for the benefit of individuals.

688    At [95], he said:

In my opinion, the State had no relationship with the consumers of oysters that imposed on the State an affirmative duty to protect those consumers from harm created by the growers and distributors of oysters…

689    And at [99], he said:

…There was simply no relationship between the Council and oyster consumers sufficient to create a duty of care…To create a duty, the relationship between the public authority and persons affected by the conduct of the authority must be “so closely and directly affected by [its] act [or omission] that [it] ought reasonably to have them in contemplation as being so affected” when it directs its mind to the relevant conduct in question. In considering whether it should exercise its powers over pollution, the Council was no more concerned with oyster consumers than any other section of the public or individual. There was no close and direct relationship between oyster consumers and the Council such that it had a duty to take care for the safety of each and every one of them…

690    Now McHugh J in Crimmins at [93] posed his second question of whether by reason of the defendant’s statutory or assumed obligations or control, the defendant had the power to protect a specific class, rather than the public at large, from a risk of harm; for the moment, put to one side that his six questions approach has not been generally accepted. He then said (at [99]):

These statements bring out the point that some powers are conferred because the legislature expects that they will be exercised to protect the person or property of vulnerable individuals or specific classes of individuals. Where powers are given for the removal of risks to person or property, it will usually be difficult to exclude a duty on the ground that there is no specific class. The nature of the power will define the class — eg, an air traffic control authority is there to protect air travellers…

691    One question to consider is whether such an analysis only applies where the case involves a failure to exercise a power, rather than the negligent exercise of power which causes a danger. Assume that the power was not there to protect safety or avoid risks to persons or property, such that it is not there to protect specific classes of individuals. Yet if its positive exercise causes harm to humans, does this matter?

692    In Pyrenees, Brennan CJ in the context of a failure to act scenario said (at [26]):

No duty breach of which sounds in damages can be imposed when the power is intended to be exercised for the benefit of the public generally and not for the protection of the person or property of members of a particular class…

693    But this reference to “a particular class” was all in the context of a failure to exercise a statutory power and whether there was a common law duty to exercise it. McHugh J (at [101]) made the point that “the common law has drawn a distinction between causing damage by a positive act and “causing” damage by a failure to act.” He then said (at [102]):

In the absence of a contract, fiduciary relationship or statutory obligation, the common law makes a person liable in damages for the failure to act only when some special relationship exists between the person harmed and the person who fails to act. By a person’s failure to act, I mean that person’s failure to act divorced from positive conduct by that person that causes damage such as the failure to brake while driving a car. A special relationship may arise from the ownership, occupation or control of land or chattels, from the receipt of a benefit or from an undertaking, assumption of responsibility or invitation which might induce the person harmed to act or to refrain from acting. None of those matters is present in this case…

694    Again, his Honour’s reference to the necessity to show a special relationship is in the context of a failure to act.

695    Now in the present case there is no physical, temporal or relational closeness between the Minister and her exercise of power and its consequences on the one hand, and the claimant class on the other hand. Further, the relevant statutory regime and the specific exercise of the statutory power in question has no direct function or purpose concerning human safety or indeed GHG emissions. And the statutory regime and powers in the present case are no analogue with Crimmins, Pyrenees, or Graham Barclay Oysters. For example, in Crimmins, the Authority’s powers and functions were directed to securing the safety of stevedoring operations, albeit that this was “an overarching supervisory and regulatory role” (McHugh J at [60]). No doubt they were compatible with the common law duty of care so found.

696    I doubt in the present case that a special relationship is required to be shown. But if that is a requirement, it has not been shown here. But the central question is whether there is a sufficient closeness and directness. And in my view, the claimant class have not demonstrated the requisite sufficient closeness and directness (Donoghue v Stevenson [1932] AC 562 at 580 and 581 per Lord Atkin).

697    First, there is no temporal closeness between the Minister and her exercise of statutory power to approve on the one hand, and the effects on the claimant class on the other hand by reason of such exercise. Indeed, the gap is many decades.

698    Second, there is no geographic closeness. The claimant class consists of members all over Australia. Contrastingly, the extension project and the decision to approve are in one sense more locally positioned.

699    Third, there is no causal closeness and directness between the approval on the one hand, and the effects on the claimant class on the other hand, in the sense that there are many links and other actors in the causal chain. I have dealt with this in a slightly different context when considering reasonable foreseeability.

700    Fourth, there is no otherwise relationship between the Minister and the claimant class, whether to be gleaned from the Act or otherwise. The Act has little to do with or say concerning GHG emissions or their consequences. Moreover, the Act does not make or recognise the claimant class as a protected species or a potential beneficiary of any exercise of power. Members of the claimant class are all strangers.

701    Generally, I cannot see how any member of the claimant class could reasonably be considered to be “so closely and directly affected” by any act of the Minster with respect to her exercise of statutory power in the present context.

Indeterminacy

702    I would say upfront that in my opinion the posited duty of care cannot be sustained by reason of indeterminacy. But why is it so? I need to first discuss some questions of principle directed to two issues.

703    First, the present context involves a personal injury case where indeterminacy is not usually analysed as a separate negative control mechanism or salient feature. But given the exceptional scope of the posited duty of care, I will seek to justify why it separately needs to be considered here, as distinct from merely observing indeterminacy by reason of the failure of the affirmative elements to establish the duty.

704    Second, and as I will explain, indeterminacy arises because of the lack of ascertainability of the relevant class. There are several points that I should make at the outset. The class is not just the claimant class, for that is the pleader’s construct that cannot artificially define the boundaries of the relevant class. Further, howsoever the class is otherwise defined, it must deal with those that are at risk of harm in the sense that they are vulnerable, that is, they cannot take adequate measures to protect themselves. So, we are not dealing with anyone at risk from GHG emissions or warmer temperatures. Rather, we are dealing with those at risk of suffering personal injury from such temperatures or their consequences who are not able or likely to be able to protect themselves. Now as soon as that is the focus, it will be appreciated that such a class is not readily ascertainable today. Sure, you can reasonably hypothesise that many of the claimant class (as presently defined 80 years or so in advance of the likely time of personal injury) are likely to be so vulnerable. But that does not define the inquiry. There is no reasonable ascertainability today of the boundaries of the true class of likely vulnerable victims.

705    Let me build up these two points upon which insufficient attention was given in the parties’ submissions.

Some questions of principle

706    The law has a control mechanism to avoid the spectre described by Chief Judge Cardozo of “liability in an indeterminate amount for an indeterminate time to an indeterminate class” (Ultramares Corp v Touche 255 NY 170 at 179 (1931)). This phrase encapsulated the policy behind the former exclusionary rule in pure economic loss cases, until Caltex Oil (Australia) Pty Ltd v The Dredge ‘Willemstad’ (1976) 136 CLR 529 removed the rule against relational pure economic loss claims and any artificial distinction previously made between words and other conduct.

707    Various approaches to indeterminacy need to be distinguished. The forensic approach used in pure economic loss cases treats indeterminacy as a separate question of fact and law which if found in a particular case may negate a duty of care. In the application of this approach, examination is made of the defendant’s ex ante knowledge of, or capacity to ascertain, its likely ambit of liability. If this is unascertainable ex ante, then usually no duty arises. In contrast, the design approach used in standard personal injury and property damage cases negates indeterminacy by utilising the affirmative elements to establish a duty. These elements are designed so that their satisfaction is treated as producing determinate liability. No separate question relating to indeterminacy is usually analysed as a negative control mechanism.

708    The focus of Cardozo CJ’s expression is on indeterminate liability derived from three elements: indeterminate amount, indeterminate time and indeterminate class. But there are three qualifications to any application of Cardozo CJ’s phrase.

709    The first qualification is that to produce indeterminate liability does not require satisfaction of all three elements. In theory, any of these elements can produce indeterminate liability. It also follows that to find determinate liability, all three elements must be negated.

710    The second qualification is that the reference to indeterminate amount is not to be read literally. A defendant could not work out ex ante the precise amount of liability. Liability literally is almost always in an indeterminate amount if the relevant perspective is ex ante. It would be a rare case where you could realistically calculate the size of any likely claim at or prior to the time of the negligent act, although there are cases where you may be able to work out the realistic maximum amount of damages. At most, the practical application of the mechanism focuses on the indeterminacy of the nature of the likely claims. No quantification of amount is necessary. Ascertainability of the likely general heads of loss is sufficient. Further, indeterminacy of amount is not concerned with the size per se of the potential claim(s), only the ascertainability of their likely nature. It is also not about capping liability but ensuring that liability can be realistically calculated. In the present case I need say nothing more concerning indeterminate amount.

711    The third qualification is that indeterminate time is not usually an independent element. Indeterminate time would usually be subsumed in, or reflected by, the element of indeterminate class or the element of indeterminate amount. Liability following a tortious act could be time dependent in two ways. The time at which a person may become a claimant (ie, suffers loss at all) may be uncertain but this may reflect itself in the indeterminacy of the class. The time at which loss may manifest itself may be uncertain but this may reflect itself in the indeterminacy of the nature of the claim. Further, negating both indeterminate class and indeterminate amount may entail negating indeterminate time, but there are cases where indeterminate time may have independent significance. For a defectively constructed building, you might have a determinate class (present and future owners) and a determinate amount (rectification work expense) for work to be carried out at some indeterminate time. But even the time span may be limited or controlled by the element of reasonableness both in the requirement that damage be foreseeable and in the content or scope of the duty. On this basis, indeterminate time may practically be given little additional work to do. In the negligent mis-statement context, there may be scope for an offending document to be relied upon within some indeterminate period, but even then the element of an indeterminate class or an indeterminate amount would usually also be established.

712    In summary, in the present context the principal focus for indeterminacy rests on class ascertainability. And if the class is indeterminate, indeterminacy follows.

713    Now indeterminacy of the class is the issue rather than its absolute size. Size and indeterminacy are different concepts. A class may be large yet have ascertainable boundaries or members. Conversely, an indeterminate class may have only one member. Indeterminacy is the quality of something which is not fixed in its parameters or is uncertain in extent or character, but a large size is more likely to reflect indeterminacy than not. Conversely, a class which is determinable is likely to be relatively smaller. However, size is not a proxy for indeterminacy. Accordingly, the issue of size per se is not generally relevant to the forensic enquiry.

714    The policy rationale underlying the control mechanism of indeterminacy has not always maintained a distinction between size per se and indeterminacy. The rationale has been to produce a pragmatic line between the recoverable and non-recoverable case, described by the United States Court of Appeal, Seventh Circuit, as “a sensible stopping point in order to preclude open-ended crushing liability on a tortfeasor” (Leadfree Enterprises Inc v United States Steel Corporation 711 F 2d 805 at 808 (1983)). Indeterminacy and size are two factors that can conjunctively produce the line. The line is also derived from a concern about causal indirectness. And if both size and indeterminacy are surrogates for this concern as Professor Peter Cane has explained (P Cane, ‘The Blight of Economic Loss: Is there Life after Perre v Apand?’ (2000) 8 Torts Law Journal 246 at 256 and 257), it is hardly surprising that differentiation between size and indeterminacy has sometimes been blurred in the authorities. Further, different policy considerations have frequently not been segregated.

715    For example, many cases have considered a utility’s liability in negligence to the public where there has been an interruption to a utility service but where the claimants have had no contract with the utility. In H R Moch Co Inc v Rensselaer Water Co 247 NY 160 (1928), Cardozo CJ denied recovery, saying (at 168) “[t]he assumption of one relation will mean the involuntary assumption of a series of new relations, inescapably hooked together”. This passage, which Ultramares then referred to (at 189), has been applied in later utility cases in preference to the more famous phrase from Ultramares. I have referred to these in a paper that I wrote 17 years ago (Indeterminacy: The uncertainty principle of negligence (2005) 13 Torts Law Journal 129 at 133) and I do not need to set them out here. However, in addition to indeterminacy, these cases have also been influenced by size, with the occasional reference to liability which is “enormous”. Further, other policy considerations have sometimes been blended to produce the pragmatic line without making clear distinctions.

716    Size is not part of indeterminacy, but part of proportionality. Neither the size of the class per se nor the magnitude of the claims per se is part of indeterminacy.

717    Proportionality considers whether imposing a duty would impose a burden out of proportion to the wrong; see Caltex at 551 and 552 per Gibbs J; San Sebastian Pty Ltd v Minister Administering the Environmental Planning and Assessment Act 1979 (1986) 162 CLR 340 at 354. Proportionality in this sense is in the context of the individual case and not the concepts of cardinal and ordinal proportionality between tort remedies and fault based tort liability in the general context. The size of the class and each claim is relevant to the burden (Perre at [108] per McHugh J). However, size is not considered in absolute terms but as a variable weighed against another variable, viz, the nature and context of the tortious conduct.

718    What determines whether a burden is proportionate to the wrong? Is it enough to ask whether the burden is reasonable? After all, negligence is an area “which has reasonableness as its central concept” (Tame at [35] per Gleeson CJ). Reasonableness requires that “account be taken both of interests of plaintiffs and of burdens on defendants” (Gifford v Strang Patrick Stevedoring Pty Ltd (2003) 214 CLR 269 at [9] per Gleeson CJ). However, the content of reasonableness is elusive. If the focus for proportionality is on the burden, the correct question may be whether liability for the breach “would be disproportionate to the risk which a person might reasonably be expected to bear as an incident of engaging in the particular activity” (Hawkins v Clayton (1988) 164 CLR 539 at 556 per Brennan J). This formulation is general and embraces an aspect of reasonableness, but at least the correct angle is used. You would consider both the nature of the conduct (assessed factually and normatively) and the risk a person might reasonably be expected to bear as an incident of engaging in that conduct. Let me return to indeterminacy.

719    What is necessary to assess whether the class is indeterminate? The first step is to stipulate the characteristics of the members of the class that you are analysing; ascertainability of such members is the second step. However, care is necessary in formulating the characteristics.

720    Defining the class too narrowly is more likely to negate indeterminacy. Equally, defining the class too broadly is more likely to produce indeterminacy. Say an electricity outage occurred for a generator supplying a city causing economic loss to various electricity users. If you define the class as those members who might be at risk of suffering loss rather than those persons who would be at risk, you would produce a broader class that was less likely to be ascertainable. But counter-intuitively a broader class may sometimes make it easier to determine its members. If the class is all persons who were connected to the city’s electricity grid (class A), you may have a very large but determinate number from available utility or public records. However, if the class is only those members of Class A who were likely to suffer loss by reason of the outage (Class B), you may have a sub-class which may be indeterminate. It may not be ascertainable ex ante which grid users were likely to suffer loss by reason of the outage. The theory of indeterminacy is to be applied to Class B even though Class B may be part of a broader determinate class. If Class A is used, indeed perhaps by analogy the claimant class in the present case, you would define away the problem. Further, it matters not that if Class B is a sub-class of Class A, you could ascertain the theoretical maximum liability referable to Class B by ascertaining the maximum liability referable to Class A, for liability to Class B must have that maximum boundary limit. Indeterminacy is not about capping liability. For Class B you would not be identifying the likely liability, only a theoretical maximum divorced from any relevant reality or policy consideration. In summary, a sub-class can be indeterminate even if the relevant universe, of which the sub-class forms a part, is determinate.

721    Now the guiding principle in selecting the characteristics of the class must be their relevance to the context. Given that the context is indeterminate liability, the class characteristics must be defined by or have a nexus with elements relevant to potential liability. And Perre discusses the necessary class characteristics for a forensic application of indeterminacy, but the descriptions given are not uniform. I have set out the references in Indeterminacy at 137.

722    The first description given is those persons who would be or were likely to be directly affected by the alleged tortfeasor’s conduct or likely to suffer loss. Those persons who might be at risk are not part of the class. The distinction between those who were likely to suffer harm and those who might be at risk is important. The latter characteristic would produce a broader class that is more likely to produce indeterminacy.

723    The second description is confined to the likely victims who were vulnerable because they were not able to take reasonable protective measures. It is a sub-set of the first description. Now it might be thought that to require identification of the likely persons who were likely to suffer loss implicitly carries with it identification of those who were vulnerable. How could you identify a person who was likely to suffer loss without identifying necessarily that they were vulnerable? However, identifying a person who was likely to suffer loss may only entail identifying that such a person had no protection in fact; it says nothing as to whether the person had the ready capacity to protect against loss but chose not to because say the risk/benefit analysis did not justify the protection as being economic.

724    The second description may be perceived to be closer to what is required. If indeterminacy concerns the ascertainability of potential liability, then to use the first description and to define the characteristic as persons likely to suffer loss, and whether or not vulnerable, considers a boundary beyond those to whom the duty is owed. However, there are arguments for the first description. The first description requires a more realistic task to be performed from the perspective of the alleged tortfeasor. It focuses on potential victims per se, not any underlying reason for their loss other than the potential causal connection between the defendant’s conduct and potential damage. In contrast, vulnerability looks at whether there were no readily available means to avoid the loss and so might explain why a person is likely to suffer loss.

725    Whatever the first description’s advantages, most judges in Perre accepted the second description. Accordingly, the applicable class description is that each member is likely to suffer loss or damage by the alleged tortious conduct and is vulnerable thereto.

726    The second necessary class characteristic, although not relevant to the context that I am considering, is implicit in Perre. It is that each member must be a first-line victim. This is required to take the ripple effect out of the equation. The law of negligence has no affection for the ripple effect. This is because indeterminacy may reflect the ripple effect or be its necessary outcome.

727    The ripple effect describes or produces secondary claims. The loss of the first-line victim can produce loss to the second-line victim. The growth pattern can be linear or exponential. A’s loss could produce B’s loss which in turn produces C’s loss. Alternatively, A’s loss could produce loss to the class of members B1 … Bn with the loss of each member of that class producing losses to further sub-classes of members CA–Cz each separately associated with a member of class B1 … . Bn. A terminal point may also be preceded by many more links. Either the linear or exponential path could produce indeterminacy, but it is the latter that produces the paradigm example. The ripple effect entails causal indirectness. Causal indirectness is likely to entail indeterminacy, although indeterminacy can manifest itself even in the absence of causal indirectness. Second-line victims form part of an indeterminate class and therefore are persons to whom no duty of care to avoid loss can be owed. This is another pragmatic rule. In a field of law where the concern is “not only with the compensation of injured plaintiffs, but also with the imposition of liability upon defendants, and the effect of such liability upon the freedom and security with which people may conduct their ordinary affairs” (Tame at [5] per Gleeson CJ), the outer bounds of a defendant’s liability must be confined by pragmatism. In the present context, I do not need to discuss the ripple effect further.

728    Let me now deal with the question of ascertainability. If the class is not ascertainable, indeterminacy follows. Now the perspective is ex ante. Ex ante knowledge is the theme of ascertainability, though one should not travel too far back in time. The earlier the frame of reference within which ascertainability is posed, the more likely that you are to produce indeterminacy, which is a problem that has manifested itself in the case before us. Ascertainability at or shortly before the time of the alleged tortious conduct is the relevant perspective and any rule needs to reflect this time frame. However ascertainability of what? The answer is the likely number of members of the class satisfying the characteristics previously discussed, namely, those likely to suffer loss. Further, you do not need to be able to ascertain the actual identity of each member but only the likely number of members. Further, actual knowledge is not essential. Liability can be determinate when the members of the class could have been ascertained by the defendant. The class need only be ascertainable rather than ascertained, although finding the latter would satisfy the former. Means of knowledge are sufficient. However, what is the ambit of means of knowledge? Where there is no clear spatial, temporal or relational boundary, there can be a spectrum of means of knowledge. Should you ask whether it was merely possible for the alleged tortfeasor to ascertain the class? However, in using this threshold, indeterminacy would be more easily negated. It is divorced from a consideration of reasonableness. Should the defendant, if reasonable enquiry had been made, have been able to ascertain the class? Such a question would not impose an unreasonable test for means of knowledge. Further, the forensic and normative exercise in assessing what the alleged tortfeasor could reasonably have ascertained has no greater difficulty than analogous exercises in this area, particularly if the focus is on whether the means of knowledge were readily available. In Perre, development of the concept of constructive knowledge was unnecessary, but there was some discussion in the ripple effect context.

729    Now so far I have discussed the concept of an ascertainable class. But reference has occasionally been made to what can be described as the “known plaintiff” rule. Is the known plaintiff rule part of Australian law? Now although Caltex seems to have left the matter open, Perre would reject it. The Perre approach is to analyse indeterminacy in terms of ex ante ascertainability of the class. Knowledge of the plaintiff’s position and vulnerability are specific elements required for the duty with a narrower focus, rather than something which negates indeterminacy, which has a broader focus. In my view, a derivative of Perre is the ascertainable class rule.

730    In summary, authority supports the use of an ascertainable class rule with the members of the class defined by the characteristics previously suggested. But does such a rule apply to personal injury and property damage cases? In my view, it can and should in an unusual case or context such as the present.

731    The accelerating complexity of society has enabled conduct to have a broader reach in inflicting personal injury or property damage. Recently there has been greater observation of indeterminacy in personal injury and property damage cases, although some of these observations have not involved the application of any negative control mechanism or salient feature; rather, they have been associated with the absence of the affirmative element of sufficient closeness and directness in the specific case. Let me start with personal injury cases.

732    Indeterminacy was observed by Gleeson CJ in Agar at [21] in denying a duty owed by members of a supervisory rugby body to players to take reasonable care in monitoring rules. Gaudron, McHugh, Gummow and Hayne JJ also denied a duty, saying that it might entail the holding of “a similar duty to the many thousands, perhaps hundreds of thousands, of persons who played rugby throughout the world” (at [67] to [70]). Their language was not that of indeterminacy or a lack of proportionality, but in terms that it could not be said that such players “were so closely and directly affected” by the relevant board members’ activities. So the observation was an application of Donoghue v Stevenson at 580, but neighbourhood concepts and indeterminacy are related themes. One may partially reflect the inverse of the other. Indeterminacy was also discussed by Gleeson CJ in Cattanach v Melchior (2003) 215 CLR 1 at [26], [32] and [38]. Gleeson CJ considered that indeterminate liability was a relevant consideration; but in that case there was some debate about the true characterisation of the claims, particularly that of the father. However, indeterminacy was not accepted by the majority.

733    Indeterminacy has also been observed in cases where a failure in electricity supply has produced a failure in building lighting, which has then caused people to suffer personal injury because of consequential accidents: see Crane v City of New York 65 NY 2d 859 (1985); Goldstein v Consolidated Edison Company of NY Inc 115 AD 2d 34 at 38–9 (1986); Espinal v Melville Snow Contractors Inc 98 NY 2d 136 at 139–42 (2002); Strauss v Belle Realty Co 65 NY 2d 399 at 403 and 405 (1985). Indeterminacy could also be observed if a computer virus caused widespread shut-downs in computerised systems for utility services in hospitals or buildings, leading to personal injury. Another example might be the diffusion of electro-magnetic radiation or a chemical which had a mutagenic effect on the germ cells of a first generation population resulting in an inherited deformity for the second generation. Claims by the second or later generation may entail indeterminacy. Claims of the first generation, whether for direct harm or economic loss for the additional expense in raising the second generation, can be put to one side for the purposes of the example. In contrast, the thalidomide episode did not involve indeterminacy. The second generation was embryonic at the time of exposure of the first generation. Thalidomide had a direct teratogenic effect on the embryos. There was sufficient closeness and directness, but in the mutagen example, arguably there is no such closeness and directness. The mutagenic effect on the germ cells would be before inheritance, that is, the second generation would not even exist in embryonic form. But there may be some causal directness even for the second (or even later) generation because of the predictable (although not time specific) biological pathway. This biological ripple effect may not be a reason to negate a duty owed to the second (or even later) generation.

734    The need to avoid indeterminacy has also explained the now rejected control mechanisms of “sudden shock” and “direct perception” (see Tame at [192] per Gummow and Kirby JJ) in psychiatric injury cases, but replacements may be unnecessary if the element of a sufficient closeness and directness embraces factual issues relating to “sudden shock” and “direct perception”. In Sullivan at [61], the court also explained that indeterminacy could arise in relation to potential liability flowing from misdiagnosis or mistreatment.

735    Generally, the authorities do not take a forensic approach to indeterminacy in the personal injury field. Moreover, such an application may be unclear in many cases. Take the example of a jumbo jet falling on a crowded stadium. Ascertainability of the likely number of victims would be the issue. The direct impact may have a limited geographic range, but ex ante where the jumbo jet was likely to crash may be unknown. It might as predictably have crashed in an open field. It may not be possible to say ex ante what the likely number of claims would be (excluding the passengers), as Gillard J said in Johnson Tiles Pty Ltd v Esso Australia Pty Ltd (2003) Aust Torts Reports 81-692; [2003] VSC 27 at [920]. However, what would be known is the likelihood of personal injury to those within the vicinity of the crash site, wherever that may be. Further, “[t]he class [would be] limited by the nature of the harm inflicted and the mechanics of its infliction” (Tame at [265] per Hayne J). The potential class would have finite temporal and geographic boundaries although ascertainability would be a more ambulatory concept. But the case before us is an exceptional case where indeterminacy should separately be addressed. Let me now say something about property damage cases.

736    In the property damage context, observations of indeterminacy have been made where the connection between the tortious act and the property damage has been indirect: Milliken & Co v Consolidated Edison Co 84 NY 2d 469 at 477–9 (1994) and H R Moch Co at 168. However, indeterminacy was not observed in the property damage context in Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973] 1 QB 27, SCM (United Kingdom) Ltd v WJ Whittall and Son Ltd [1971] 1 QB 337 and British Celanese Ltd v A H Hunt (Capacitors) Ltd [1969] 1 WLR 959. Each allowed recovery for property damage and consequential economic loss for factories whose electricity supply had been disturbed without any concern expressed as to indeterminacy for those specific losses, but each factory was in the immediate vicinity of the tortious conduct and known by the tortfeasor to be affected. Buckley LJ in SCM (at 357) distinguished this from “an electric generating station or main cable or a principal water-main serving a large number of consumers over a wide area [which was] put out of action”. In SCM, the factory that had its power cut off was in the same road as where the contractor was working. In Spartan Steel, the factory was a quarter of a mile from the digger and there was a direct cable from the Midlands Electricity Board to the factory. Further, in Dorset Yacht Co Ltd v Home Office [1970] AC 1004, particularly at 1034, 1070 and 1071, there was sufficient closeness and directness, between the negligent conduct and the damaged yachts and no indeterminacy was observed.

737    Now in the vast majority of personal injury and property damage contexts, the requirement of a sufficient closeness and directness between the tortious act and the damage usually avoids the need for a negative control or to consider indeterminacy as a separate salient feature. If indeterminacy is an aspect of causal indirectness, then requiring such directness usually negates the application of its related inverse, indeterminacy. It is not a true inverse since a class could still be indeterminate even if there was closeness and directness between act and effect. Equally, a class may be determinate even though there was no such closeness and directness but the exceptions would be very unusual. Physical damage depends upon the operation of physical laws. The operation of physical laws usually provides directness and, by reason of that directness, determinacy. And in the case where physical laws produce indirect consequences, an indeterminate outcome may be the consequence. So where sufficient closeness and directness has been found, indeterminacy has not usually been observed. Correspondingly, where there has been no such closeness and directness, indeterminacy has sometimes been observed. However, then the duty fails for absence of the affirmative element. The authorities are generally consistent with these propositions.

738    In the supply of a defective product causing personal injury there could be no indeterminacy. The want of care and personal injury are directly and intimately connected. The ascertainable class, even with one or more links in the supply chain, would be the potential consumer although the precise identity may be unknown. There is also no ripple effect. Where there is the supply of a defective product, the fact that the product may pass through intermediate hands does not change the fact that the end-user’s loss is caused by the defect and not any loss to, or reaction by, any intermediate entity in the supply chain. Further, “[a] negligently made article will [usually] only cause one accident” (Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465 at 483 per Lord Reid). But in the case of the non-supply of a product, the ripple effect can manifest itself in the different economic loss context. Further, if a product such as a contaminated blood supply, a product causing dust disease, or a drug or radio-active waste injures many, there may still be a direct and intimate connection between the defect or dangerous quality produced by the want of care and the injury. However, indeterminacy may be observed where there was greater temporal and geographic distance between a non-manufacturer defendant’s conduct and the victim, with the defective manufacture being the act of a third party.

739    However, there may be non-classic cases. For example, the negligent shut-down of an electricity generator in South Australia that supplies the National Electricity Market may lead indirectly to the spoiling of food in a Victorian restaurant because the local electricity retailer has not been able to acquire sufficient electricity from the NEM for on-supply to the restaurant for refrigeration. Another example is the marine hydrographer who omits reference to a reef on a map, indirectly leading to loss by the captain of a ship who in reliance on the map steers the ship on to the reef. In such non-classic cases, indeterminacy may be observed, but then there would be no sufficient closeness and directness.

740    In summary, in the context of personal injury and property damage cases, if one finds sufficient directness and closeness, indeterminacy is usually taken not to arise either as to class or amount. Sufficient closeness and directness involves considering both the victim and type of damage. The expression “so closely and directly affected” (Donoghue v Stevenson at 580) entails a consideration of both aspects. If one does not find sufficient closeness and directness, indeterminacy may be observed. However, these observations are not usually the result of an application or recognition of any separate negative control mechanism or salient feature, but of the shadow associated with the absence of the said affirmative element in the specific case.

741    But in the present case we are dealing with an exceptional context. Given the potential geographic range, timing of and contingencies associated with potential effects flowing from any breach and the ambit of the claimant class, it is necessary to address indeterminacy as an important control mechanism or salient feature. True it is that I have found no sufficient closeness and directness, and that finding is sufficient to find against the posited duty. Nevertheless, if I am wrong on that issue, either as it presently is dealt with in the authorities or because it has been or may be reconceptualised, the issue of indeterminacy would still not be overcome given the potential causal effects and breadth that I have just described. In some respects, although we are dealing with a personal injury case, we are not dealing with a direct causal pathway such as involved with a defective product, a contact mechanical force or a deleterious agent such as dust or a chemical. In that sense, the case is further along the spectrum towards “open-ended” liability and unconstrained causal pathways that one usually finds in the pure economic loss context.

The present case

742    I have already found against the claimant class in the sense that there is no relevant sufficient closeness and directness as I have discussed. Unsurprisingly perhaps, and consistently with that conclusion, the respondents also fail on the question of indeterminacy.

743    Now I should say at the outset and as should be apparent from my earlier discussion, the fact that the claimant class is large does not concern me and is not the issue in terms of indeterminacy. Size per se is not the question. Moreover, the potentially large financial exposure of the Minister or the Crown if a duty were to exist and be breached is also not of concern to me. That is more a question concerning any lack of proportionality rather than indeterminacy. And on that score, the fact that the burden might be larger than usual is just a function of the serious consequences that might flow from any breach of duty in the present case. I would not find against the claimant class based upon any perceived lack of proportionality. Rather, they fail on indeterminacy.

744    Shortly put, in my view the likely number of members of the claimant class with the characteristics that I have identified are not readily ascertainable today. One can identify all those alive today in terms of the relevant universe and in a diffuse sense say that they are at risk from GHG emissions and higher temperatures. But that is not addressing the subset with the relevant characteristics who are simply unascertainable today, even assuming that the claimant class has been properly defined.

745    As is obvious, the risk of harm from future bushfires and heatwaves is highly variable across the variety of climatic conditions in Australia, impacting on the practicability of assessing the nature and extent of prospective liability. Contrastingly, conventional duties of care are usually concerned with a finite set of physical consequences flowing from an act or omission. But here we are not dealing with the possibility or consequences of a single catastrophic event. I agree with the Minister that the concern is with rolling events potentially causing damage where there is no meaningful limit on how many of the claimant class will suffer harm and how many times they will be so harmed, when that damage will occur over the next century or so, and the extent of that damage.

746    Further, why confine the duty to the claimant class, as opposed to all living people at the relevant time, including those presently unborn, who will be exposed to the same risks as the claimant class, albeit for different periods of time? And why exclude from the class those that only suffer property damage. Or why exclude consequential economic loss claims? There is no reason in principle to do so. The true potential claimant class, if a duty is to be owed, cannot be foreclosed by the pleader’s construct. Moreover, even on its present construct, it is not properly focused on the correct class characteristics as I have previously discussed.

747    In summary, whether the claimant class or the true potential claimant class, in my view indeterminacy forecloses the posited duty of care. The class of likely vulnerable victims is simply unascertainable today.

Conclusion

748    In summary, his Honour ought not to have declared that the posited duty of care was owed. My principal basis for so concluding is that the respondents did not establish the requisite sufficient closeness and directness. Moreover, in the present case indeterminacy is a powerful salient feature against the duty of care being owed.

749    For the foregoing reasons I would allow the appeal on grounds 1, 2(a) & (b) and 3(d) & (e) only and set aside the declaration. But I should make clear one matter, which may be of importance to the claimant class for the future. I would entertain an application to remove the representative nature of the proceedings. It may be that one day, one or more members of the claimant class may suffer damage and so have an apparently complete cause of action. If so, it should then be open to them to assert the existence of a relevant duty of care, breach of duty, causation and damage. No issue estoppel should then arise on any question concerning the existence of a duty by allowing the present appeal. But to achieve that outcome, the representative nature of the proceedings would need to be addressed. I would be prepared to receive further submissions from the parties on such questions before making formal orders to set aside the primary judge’s declaration. I agree with the orders proposed by the Chief Justice.

750    Let me make three final observations by reference to the following well-known anecdote recounted by President John F Kennedy in an address on 23 March 1962 given at the University of California, Berkeley:

The great French Marshal Lyautey once asked his gardener to plant a tree. The gardener objected that the tree was slow-growing and would not reach maturity for a hundred years. The Marshal replied, ‘In that case, there is no time to lose, plant it this afternoon.’

751    There are three themes worth drawing from this carbon capturing metaphor.

752    First, no one doubts that immediate sensible action is necessary to deal with GHG emissions even if the benefit of such action may not come to fruition for many decades, indeed even into the next century. But the present case is not about gainsaying that broader theme.

753    Second and resonating with the metaphor, the primary judge planted the seed of a cause of action in finding the posited duty, but envisaging that the seed may not fulfil its Aristotelian potential of a fully formed tort for many decades, if at all. This was a bold step to take given that trial judges normally only assess, admire or indeed chop down completed forms.

754    Third, it is for the High Court not us to engineer new seed varieties for sustainable duties of care, modifying concepts such as “sufficient closeness and directness” and indeterminacy to address the accelerating complexity, multiple links and cross-links of causal relations. Such concepts in their present form may have reached their shelf life, particularly where one is dealing with acts or omissions that have wide-scale consequences that transcend confined temporal boundaries and geographic ranges, and where more than direct mechanistic causal pathways are involved.

I certify that the preceding four hundred and seven (407) numbered paragraphs are a true copy of the Reasons for Judgment of the Honourable Justice Beach.

Associate:

Dated: 15 March 2022

REASONS FOR JUDGMENT

WHEELAHAN J:

Introduction

755    The respondents to this appeal are children who brought the proceeding below by a litigation representative. Invoking r 9.21 of the Federal Court Rules 2011 (Cth), they maintained the proceeding in a representative capacity on behalf all children born before the commencement of the proceeding and who ordinarily resided in Australia at the time of its commencement.

756    The question on this appeal is whether the primary judge was correct to declare that the Commonwealth Minister for the Environment, in exercising powers under s 130 and s 133 of the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act) in respect of an extension to the operations of the Vickery Coal Project in New South Wales (the Extension Project), owed to the respondents and those whom they represent a common law duty to take reasonable care to avoid causing personal injury or death arising from emissions of carbon dioxide (CO2) into the Earth’s atmosphere. The case is unusual, not the least because at the time of the making of the declaration, no common law cause of action in negligence against the Minister had accrued, and the only substantive relief given was the declaration. A claim by the respondents for a final injunction was dismissed, from which there is no cross-appeal.

757    I would allow the appeal. The declaration that a duty of care was owed by the Minister to the respondents and those whom they represent was made in error, and in consequence it should be set aside. In summary, my reasons for allowing the appeal are that: (1) the Minister’s function under s 130 and s 133 of the EPBC Act in this case does not erect or facilitate a relationship between the Minister, and the respondents and those whom they represent, that supports the recognition of a duty of care; (2) it would not be feasible to establish an appropriate standard of care, with the consequence that there would be incoherence between the suggested duty and the discharge of the Minister’s statutory functions; and (3) I am not persuaded that it is reasonably foreseeable that the approval of the Extension Project would be a cause of personal injury to the respondents, as the concept of causation is understood for the purposes of the common law tort of negligence.

Preliminary remarks – concrete facts

758    Before turning to the issues raised on the appeal in more detail, I will make some preliminary remarks concerning the subject matter of the appeal, being a declaration as to the existence of a duty of care in a novel case. At the time of making the declaration, no common law cause of action in negligence had accrued. There was no allegation by the respondents of any actual breach, or causation, or d