FEDERAL COURT OF AUSTRALIA

Palmer v State of Western Australia (No 4) [2020] FCA 1221

ORDERS

THE COURT ORDERS THAT:

1.    The costs of the remitted issue be reserved to the High Court of Australia.

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

REASONS FOR JUDGMENT

TABLE OF CONTENTS

RANGIAH J:

1    On 11 March 2020, the World Health Organisation declared COVID-19 to be a pandemic. On 15 March 2020, the Minister for Emergency Services for Western Australia declared a state of emergency.

2    On 5 April 2020, the second respondent, the State Emergency Coordinator for Western Australia, issued the Quarantine (Closing the Border) Directions (WA). They have since been amended several times, most recently on 19 July 2020. I will refer to the current version as the “Directions”. The Directions were made pursuant to the Emergency Management Act 2005 (WA).

3    The Directions prohibit entry into Western Australia of persons other than those defined as “exempt travellers”.

4    On 25 May 2020, the applicants, Clive Frederick Palmer and Mineralogy Pty Ltd, commenced proceedings in the High Court of Australia seeking a declaration that the Emergency Management Act and/or the Directions are invalid, in whole or in part, on the basis that they contravene s 92 of the Constitution.

5    On 16 June 2020, the Chief Justice of the High Court made the following order:

Pursuant to section 44 of the Judiciary Act 1903 (Cth) so much of this matter as concerns the claim by the defendants of the reasonable need for and efficacy of the community isolation measures contained in the Quarantine (Closing the Border) Directions…made on 5 April 2020 be remitted to the Federal Court of Australia for hearing and determination.

6    From 27 to 31 July 2020, I conducted a hearing of the remitted issue. The applicants, the respondents and two interveners, the Attorney-General for the Commonwealth (Commonwealth) and the Attorney-General for Queensland (Queensland), actively participated in the hearing.

7    The respondents called two expert witnesses to give evidence, while the applicants called one expert witness and the Commonwealth called two. The Commonwealth supported the applicants’ case, and Queensland supported the respondents’ case. Each party cross-examined each of the witnesses and made opening and closing submissions.

8    On 2 August 2020, after the conclusion of the hearing, the Commonwealth notified the High Court that it was withdrawing from the proceeding. On 5 August 2020, the Commonwealth belatedly notified this Court of its withdrawal.

9    The respondents claimed that the Commonwealth was not entitled to withdraw without the leave of this Court, and that if the Commonwealth was given leave to withdraw, there should be a new hearing on the basis that they were prejudiced by the Commonwealth’s conduct. I held a hearing in respect of those claims on 7 August 2020. At that hearing, the applicants indicated that they relied upon the evidence that had been called and the submissions made by the Commonwealth. I have delivered separate reasons determining that the Commonwealth did not require leave from this Court to withdraw and that there should be no rehearing of the evidence: Palmer v State of Western Australia (No 3) [2020] FCA 1220.

10    Accordingly, I intend to decide the remitted issue on the basis of the whole of the evidence and the submissions. In view of the way the hearing was conducted, it remains convenient to distinguish between the evidence called by the applicants and the evidence called by the Commonwealth.

11    The remitted issue focuses on the determination of the factual matters involved in the respondents’ defence of the proceeding in the High Court. That defence involves a contention that the Community Isolation Measures, or border restrictions, contained in the Directions, are justified because:

(a)    they are reasonably necessary for the protection of the Western Australian community against the health risks of COVID-19;

(b)    they are reasonably appropriate and adapted to advance that object or purpose;

(c)    there are no other equally effective means, which would impose a lesser burden on interstate trade, commerce and intercourse, available to achieve that object or purpose.

12    The remitted issue requires assessment of the risk of COVID-19 spreading into the Western Australian population were the border restrictions to be removed. That risk depends substantially upon the ability of public health authorities to control the outbreaks presently occurring in several States. However, the extent of the outbreaks is in a state of flux. While the parties agree that the facts are to be determined on the basis of the evidence presented at the hearing, there has necessarily been a time-lag between the hearing and the delivery of these reasons. These reasons, therefore, cannot take into account any factual developments since the hearing. That lends a degree of artificiality to these findings.

13    In order to give context to the remitted issue, it is necessary to consider the constitutional context in which that issue arises and the facts pleaded in the proceeding before the High Court.

The constitutional context and the pleadings

14    Section 92 of the Constitution provides that, “trade, commerce, and intercourse among the States…shall be absolutely free”.

15    The applicants allege in their Second Further Amended Statement of Claim that, in contravention of s 92, the Directions impose an effective burden on the freedom of intercourse, or impose an effective discriminatory burden with protectionist practical effect on the freedom of trade and commerce, among the Australian people in the several States.

16    In response, the respondents allege in para 47(c) of their Second Amended Defence that the relevant provisions of the Emergency Management Act and the Directions do not have the purpose of economically protecting Western Australia, but are for the legitimate purpose of protecting the population of Western Australia against the risks of an emergency situation.

17    The respondents plead in para 47(d) of their Second Amended Defence that the continuance in force of the Directions:

(iii)    … is reasonably necessary to achieve, and is compatible with, the legitimate purpose of protecting the population of Western Australia against the health risks of COVID-19 where there are no other equally effective means available to achieve that purpose or object, but which impose a lesser burden on interstate trade or commerce;

(iv)     does not prevent intercourse with the State of Western Australia among the States, except for the purpose of, and is reasonably necessary for, regulating or preserving the population of Western Australia against the health risks of COVID-19; and

(v)    is reasonably appropriate and adapted to advance that purpose or object where there are no other equally effective means available to achieve that purpose or object, but which impose a lesser burden on interstate intercourse.

18    The respondents contend that whether a burden imposed upon the freedoms provided by either the trade or commerce limb or the intercourse limb of s 92 is constitutionally valid should be assessed in a similar way as is a burden imposed upon political communication under the test in McCloy v New South Wales (2015) 257 CLR 178 at [2]–[3] (French CJ, Kiefel, Bell and Keane JJ). A test of that kind appears to have been contemplated by the High Court in respect of the trade or commerce limb of s 92: see Betfair Pty Ltd v Western Australia (2008) 234 CLR 418 at [102]–[103] (Gleeson CJ, Gummow, Kirby, Hayne, Crennan and Kiefel JJ); Unions NSW v New South Wales (2019) 264 CLR 595 at [42] (Kiefel CJ, Bell and Keane JJ). A cognate approach has also been supported for the intercourse limb of s 92: see APLA Ltd v Legal Services Commissioner (NSW) (2005) 224 CLR 322 at [173]–[177] (Gummow J), [402], [408] and [421] (Hayne J).

19    Whether the respondents’ asserted analysis applies to either limb of s 92 is not a matter to be determined by this Court. It is sufficient to accept that this analysis is asserted by the respondents, and facts must be found upon that basis.

20    In para 15B(b) of their Amended Reply, the applicants plead that:

(a)    the probability of a person infected with COVID-19 travelling to Western Australia is so small as to provide no reasonable justification for the continuance of the border restrictions;

(b)    the relevant risk is the probability of uncontrolled and uncontrollable community transmission if the border restrictions were not in place; and the probability of such transmission is not sufficiently high to provide reasonable justification for the continuance of the border restrictions.

21    In para 20 of their Amended Reply, the applicants deny or do not admit the allegations in paras 47(d)(iii), (iv) and (v) of the Second Amended Defence.

The issues

22    Paragraphs 47(d)(iii), (iv) and (v) of the Second Amended Defence allege that the border restrictions imposed under the Directions are:

(a)    reasonably necessary for the protection of the Western Australian population against the health risks of COVID-19; and

(b)    reasonably appropriate and adapted to advance that purpose or object, in circumstances where there are no other equally effective means which impose a lesser burden on interstate trade, commerce and intercourse, available to achieve that purpose or object.

23    In these paragraphs, only “health risks” are raised in justification of the border restrictions. The parties submit that this Court ought not to consider other risks, such as economic or social risks. I will refrain from doing so.

24    The parties submit that this Court should not decide whether the Directions are in fact reasonably necessary for the protection of the Western Australian population, nor whether they are reasonably appropriate and adapted to advance that purpose or object, nor whether there are no other equally effective means which impose a lesser burden on interstate trade, commerce and intercourse, available to achieve that purpose or object. They submit that the purpose of the remitter is for the Federal Court to find the facts that will allow the High Court to determine these issues. I will refrain from deciding these matters. However, as there is no clear line of demarcation, some incursion may be unavoidable.

25    The parties agree that the respondents’ defence must be considered by reference to whether the border restrictions are currently justified, not whether they were justified when they were introduced.

26    There are several overlapping factual premises involved in paras 47(d)(iii), (iv) and (v) of the Second Amended Defence.

27    The first premise is that COVID-19 poses a substantial danger to the health of the Western Australian population.

28    The second is that the populations of other States and Territories are or may be infected with COVID-19, whereas the Western Australian population is not.

29    The third is that the border restrictions effectively protect the Western Australian population from COVID-19 by reducing the probability that infected people from other States and Territories will enter Western Australia.

30    The fourth is that, while there may be alternative methods available of reducing the probability that infected people will enter Western Australia, and which may impose a lesser burden on interstate trade, commerce and intercourse, they would be less effective, and inadequate, to protect the health of the Western Australian population.

31    The fifth is that if infected people did enter Western Australia, while there are measures that are or could be put in place to reduce the probability of the infection spreading into and amongst the population, which may impose a lesser burden on interstate trade, commerce and intercourse, they would be less effective, and inadequate.

32    It will be necessary to consider the applicants’ allegation that the probability of a person infected with COVID-19 travelling to Western Australia is so small as to provide no reasonable justification for the continuance of the border restrictions.

33    Further, the respondents plead that the risk is the risk of any community transmission of COVID-19, whereas the applicants plead that the relevant risk is the risk of uncontrolled and uncontrollable community transmission. It will be necessary to make findings relevant to both kinds of risk.

34    These issues make it necessary to determine:

(a)    whether there is ongoing community transmission of COVID-19 in Australia;

(b)    the public health consequences of persons infected with COVID-19 entering Western Australia and transmitting the virus;

(c)    the extent of the contribution made by the border restrictions to reducing the probability of community transmission in Western Australia;

(d)    the probability of COVID-19 being imported into Western Australia and community transmission occurring, including uncontrolled and uncontrollable community transmission, if the border restrictions were removed;

(e)    the efficacy of measures other than the border restrictions in reducing the risk of introduction of COVID-19 into, and transmission within, Western Australia.

35    The parties and the interveners have agreed a Statement of Issues. The Statement of Issues largely, but not entirely, captures the facts which, in my opinion, need to be determined. In particular, the disputed issues do not deal expressly with the principles relevant to public health decision-making in the context of the pandemic, or the potential consequences of outbreaks of COVID-19.

36    I will proceed by, first, considering the onus and standard of proof; second, briefly describing the agreed facts and evidence and assessing the evidence of the expert witnesses; third, addressing the premises and factual issues I have identified; and, fourth, addressing the factual matters specifically pleaded in the Statement of Issues.

Onus and standard of proof

37    The respondents accept that they bear the ultimate onus of proof on the remitted issue.

38    However, the respondents submit that the facts required to be found by this Court are “constitutional facts”, and that concepts of legal onus and legal standards of proof are inapposite in respect of such facts. They rely upon Thomas v Mowbray (2007) 233 CLR 307 at [620]–[639] (Heydon J), Maloney v The Queen (2013) 252 CLR 168 at [355] (Gageler J) and Clubb v Edwards (2019) 93 ALJR 448; (2019) HCA 11 at [152] (Gageler J). They submit that, rather, the Court is required to be satisfied of the facts’ existence.

39    The applicants submit that the issues to be determined by this Court involve not only “constitutional facts”, but also “adjudicative facts”, in respect of which findings on the balance of probabilities should be made. They also submit that s 140 of the Evidence Act 1995 (Cth) applies to this proceeding and requires that findings of fact in civil proceedings be made on the balance of probabilities. They submit that in Re Day (2017) 91 ALJR 262; [2017] HCA 2, Gordon J at [14] indicated that facts would be found on the balance of probabilities, whether the facts were classifiable as “constitutional” or “adjudicative”.

40    The issue of the standard of proof in respect of fact finding in a case remitted for the making of findings of fact relevant to a constitutional issue is unsettled. It is unnecessary for me to enter upon the issue. That is because the respondents accept that the extent of the proof necessary to obtain the required satisfaction will be informed by the nature of the factual inquiry in question. They accept that, in particular, this is so where the High Court has determined that questions of fact are to be ascertained by judicial process: cf Re Day at [22]–[26]. The respondents accept that it is open to this Court to decide the “intermediate facts” applying a standard of the balance of probabilities. I propose to adopt that course.

41    The parties are in agreement that where an intermediate fact is essential to drawing the relevant inference for the existence of the ultimate fact, the Court requires the same standard of persuasion for the intermediate fact as it does for the ultimate fact: Chamberlain v The Queen (No 2) (1984) 153 CLR 521 at 538–539. However, they accept that where an intermediate fact is merely one circumstance, which is not of itself essential but which, together with other circumstantial evidence, would sustain the drawing of an inference as to an ultimate fact on the balance of probabilities, then it is not necessary that the intermediate fact itself be proven to that standard, provided the intermediate fact is not one the existence of which the Court doubts: Chamberlain at 537; Australian Broadcasting Corporation v Wing (2019) 271 FCR 632 at [134].

The agreed facts and the evidence

42    The parties have agreed upon a large number of facts. The agreed facts are set out in a Draft Consolidated Special Case. These reasons adopt the terminology used in, and assume familiarity with, the Draft Consolidated Special Case.

43    There were five expert witnesses who gave evidence. The experts, the date or dates of their reports and the parties by whom they were called, are:

    Dr Andrew Robertson — 24 June, 3 July 2020 (the respondents);

    Associate Professor Kamalini Lokuge — 26 June, 21 July 2020 (the respondents);

    Associate Professor Sanjaya Senanayake — 7 July 2020 (the applicants);

    Professor Peter Collignon — 7 July 2020 (the Commonwealth);

    Professor Tony Blakely — 8 July 2020 (the Commonwealth).

44    The dates of the reports assume some significance because some aspects of the reports have been overtaken by recent developments concerning outbreaks of COVID-19 in Australia. The reports were admitted into evidence without objection.

45    In addition, the experts prepared a joint report on 23 July 2020 following a conference between them, setting out their areas of agreement and disagreement.

46    The experts gave their evidence concurrently, with each witness being called in turn, but cross-examining counsel being able to ask questions of, not only the witness, but also the other experts.

47    I will summarise various aspects of the reports of the expert witnesses later in these reasons. I will give my general assessment of the witnesses at this stage.

48    Associate Professor Lokuge was a highly impressive witness. She is a public health physician and medical epidemiologist, with a doctorate in epidemiology. She leads the Humanitarian Health Research Initiative at the National Centre for Epidemiology and Population Health at the Australian National University Research School of Population Health. Associate Professor Lokuge has over two decades of experience as an epidemiologist investigating transmission and implementing control of infectious disease outbreaks and pandemics. She is a specialist in the control of infectious diseases.

49    Associate Professor Lokuge’s expertise covers front-line epidemiological, clinical and public health responses for controlling high-risk pathogen outbreaks. Her experience includes leading and participating in field-level responses to Ebola, Avian Influenza, Pandemic H1N1 Influenza and Lassa, in Africa, Asia and Australasia. She is a member of the National COVID-19 Health and Research Advisory Committee, an independent committee providing advice to the Chief Medical Officer of Australia. As part of her role on that committee, she chaired a working group which produced a report advising the Australian Government on priorities for preventing the resurgence of COVID-19 transmission. She is presently advising the Victorian health authorities in their attempts to control the current outbreak.

50    Associate Professor Lokuge demonstrated an impressive command of the relevant research and literature, both in her oral evidence and her reports. She was generally able to support the propositions she contended for by reference to data. Where the data did not allow conclusions to be reached, she was frank in so stating. Of the experts who gave evidence, Associate Professor Lokuge not only had the most comprehensive grasp of the academic research, but, by far, the greatest practical, front-line experience in the control of pandemics.

51    Associate Professor Lokuge also demonstrated an understanding of the principles involved in the containment of large-scale outbreaks of infectious diseases, in a way that some of the other experts did not. Her analysis took into account principles of risk management which balance the potential consequences or impact of an outbreak with the probability of the outbreak occurring. Associate Professor Lokuge, Dr Robertson and Professor Blakely were the only experts who expressly took into account the necessity for a precautionary approach in the management of the pandemic.

52    The applicants and the Commonwealth criticise Associate Professor Lokuge’s evidence on a number of grounds. A number of those grounds are unfounded. For example, the applicants’ criticism of examples given by Associate Professor Lokuge to illustrate the potential for large outbreaks despite an overall situation of good control, fails to recognise that the experts, citing some of her examples, unanimously agreed that rapid, uncontrolled transmission resulting from a single infected individual has occurred in multiple settings even where there is otherwise good surveillance/testing control. Some of the criticisms have some substance, as I will discuss in the course of these reasons, but that does not materially affect my overall view of her evidence.

53    I accept the opinions expressed by Associate Professor Lokuge, except to the extent I will indicate otherwise. I will more specifically consider the merits of her opinions on the disputed issues later in these reasons.

54    Dr Robertson has been the Chief Health Officer for Western Australia since June 2018. Since 2003, he has held several high-level positions involving disaster preparedness and management in the Western Australian Department of Health. He has specialist medical training and a masters’ degree in Public Health and Health Service Management. Dr Robertson’s recommendations were influential in the making of the Directions.

55    Dr Robertson’s expertise is upon general public health, rather than the specific area of transmission of infectious diseases. In his management of the pandemic, he has relied, at least in part, upon the advice of other experts. However, he has demonstrated considerable understanding of the practical implications of outbreaks of COVID-19 in Western Australia and the ability of the health authorities to deal with and control such outbreaks. Much of Dr Robertson’s evidence was consistent with that of Associate Professor Lokuge.

56    Associate Professor Senanayake is a Senior Staff Specialist in Infectious Diseases at the Canberra Hospital and an Associate Professor in the Australian National University Medical School. He is a specialist physician in the area of infectious diseases and has a masters’ degree in Applied Epidemiology. Associate Professor Senanayake has considerable experience in the treatment of infectious diseases, but has had little practical experience in the control of large-scale outbreaks. His record of research and publications in the area of pandemic control is also very limited.

57    Associate Professor Senanayake’s report was focused substantially upon the statistical probability of a person infected with COVID-19 entering Western Australia. He conceded under cross-examination that his calculations were estimations based upon average figures, and that average figures are not a proper basis for estimating inherently stochastic events such as COVID-19¹. Associate Professor Senanayake also accepted that there were problems with a number of assumptions that he had made and that there was no evident explanation of the source of some figures he used². He said that he suspected he had used the wrong version of an appendix to his report, and that had led to inaccuracies in the report³. In view of the concessions made by Associate Professor Senanayake, and the limits of his expertise and experience in the management of pandemics, I approach his evidence with caution. Where there is conflict between his opinions and those expressed by Associate Professor Lokuge, I prefer the opinions of Associate Professor Lokuge.

58    Professor Collignon is a specialist physician in infectious diseases and a specialist medical pathologist in the field of microbiology. He is a Senior Staff Specialist at ACT Pathology. He is also a Professor in the Australian National University Medical School. He has a doctorate in anti-microbial resistance and believes he is regarded as one of the world’s experts in the spread of resistant bacteria. He has over 30 years’ experience. Professor Collignon is a member of the Australian Government’s Infection Control Expert Group on COVID-19, which provides advice about the control of infections in hospitals and the community.

59    The respondents have not submitted that Professor Collignon was not qualified to give his opinion evidence. However, he was subjected to a vigorous cross-examination about the level of his expertise in respect of viral pandemics (when his primary area is antibiotic resistance), his absence of formal qualifications in epidemiology and the number of papers he has published concerning pandemics. It was also suggested that Professor Collignon was advocating for a position against the closure of State borders. While it was not improper for the respondents to cross-examine in this manner, in fairness to Professor Collignon, I will expressly state that I reject the attack upon his expertise and impartiality. The evidence he gave was within the bounds of his expertise. Further, while he had a firm position upon the issue of State-wide border restrictions, there is nothing to suggest that it was other than a genuinely held opinion.

60    Having said that, I do have reservations about some aspects of Professor Collignon’s evidence. Upon some important issues in his report and his oral evidence, he tended to fall back upon language such as “I consider”, “I believe” and “my view”. These expressions tended to mask the basis for his opinions, in contrast to Associate Professor Lokuge, whose reasoning tended to be more transparent. For example, Professor Collignon stated in his report that, “[I]n my view these outbreaks or clusters are better handled by targeted community and regional interventions rather than State-wide interventions”⁴. It may be inferred, from an earlier comment he had made, that this conclusion was based on economic, social and individual impacts⁵. Accordingly, it is not apparent that there is any epidemiological or public health basis for Professor Collignon’s opinion that outbreaks or clusters are better handled by targeted interventions than State-wide border closures. I do not think that Professor Collignon’s evidence exhibited the same level of transparency and clarity of Associate Professor Lokuge’s evidence, nor was it supported by the same level of detailed research. Nor does Professor Collignon have anything like the same level of experience in the practical containment of fast-spreading infectious diseases. Further, Professor Collignon did not display the same understanding of the principles of the management of a pandemic as Associate Professor Lokuge. Where there is conflict between the opinions of Professor Collignon and those expressed by Associate Professor Lokuge, I prefer the opinions of Associate Professor Lokuge.

61    Professor Blakely is an epidemiologist and public health medicine specialist at Melbourne University’s School of Population and Global Health. From 2010 to 2019, he directed the Burden of Disease Epidemiology, Equity and Cost Effectiveness Programme at the University of Otago. Prior to that, he conducted studies concerning smoking and cancer. He has medical qualifications, as well as a doctorate. His particular focus is on epidemiological and quantitative research methodologies.

62    Professor Blakely’s report states that it aims to quantify the probability of transmission of COVID-19 into Western Australia from other States and Territories if the border restrictions were removed. Professor Blakely expressed confidence in his methodology, while frankly and transparently acknowledging the uncertainties involved in a number of the assumptions underlying it. He acknowledged other limitations of his modelling, particularly that it does not account for intentional non-compliance with protective measures. Professor Blakely also acknowledged that while his modelling provides a tool that can be used as part of a suite of considerations when making decisions about border restrictions, it should not be used by itself to make such decisions. I accept that Professor Blakely’s model is a useful tool of analysis, subject to its inherent limitations which I will discuss. Apart from their differing opinions as to the usefulness of modelling, Professor Blakely’s opinions tended to align with those of Associate Professor Lokuge.

Areas of agreement and disagreement between the expert witnesses

63    The expert witnesses conferred on 22 and 23 July 2020. They produced a joint report in which they identified areas of their agreement and disagreement.

64    The experts summarised their areas of agreement as follows:

65    The experts set out their areas of disagreement by reference to a series of questions they posed. The questions are as follows:

1.1    What is the probability of an infected case of COVID-19 arriving at the border of Western Australia?

1.2    Having regard to question 1.1 above, is a quantitative and/or qualitative assessment of probability more useful in these circumstances?

2.1    What are the possible impacts of the importation from within Australia of COVID-19 into Western Australia and what is the probability of those outcomes occurring?

3.1    What is the value of border restrictions in preventing the transmission of COVID-19?

3.2    Are state borders a useful delineation between populations for the purposes of managing the transmission risk of COVID-19?

4.1    What, if any, additional and/or alternative value is derived from border control measures over and above all non-border measures?

66    I will address the experts’ opinions in respect of these areas of disagreement in the course of these reasons.

The border restrictions under the Directions

67    The Directions provide that a person must not enter Western Australia unless the person is an exempt traveller (cl 4). That restriction applies to all persons in the other States and Territories, regardless of whether they ordinarily reside in Western Australia.

68    There are 18 categories of “exempt traveller” (cl 27), which may be summarised as follows:

(a)    certain Commonwealth and State government officials and military personnel (paras (a)-(f));

(b)    persons responsible for provision of transport or freight and logistics services into or out of Western Australia, provided they remain for only so long as reasonably required (para (g));

(c)    persons whose specialist skills are required for time-critical businesses or infrastructure (paras (h)-(i));

(d)    persons whose presence is required for agriculture, food production or a primary industry (para (j));

(e)    “FIFO” workers and their family members, who have undertaken a period of mandatory 14 day isolation (paras (k)-(l));

(f)    emergency service workers (para (m));

(g)    judicial officers and staff members of a court, tribunal or commission (para (n));

(h)    persons who enter on medical, or specified compassionate grounds (paras (o)-(q));

(i)    persons specifically authorised to enter on other grounds (para (r)).

69    The Directions also provide, broadly:

(a)    an exempt traveller must comply with any specified terms or conditions — which may be imposed on a particular exempt traveller or persons in a category of exempt traveller (cll 27, 41);

(b)    an exempt traveller must not enter Western Australia if the person has symptoms; has received notice that the person has a close contact who has tested positive; is awaiting a test result; or has received a positive test and has not been certified as having recovered (cll 5(a)-(d));

(c)    an exempt traveller must not enter if the person has been in New South Wales or Victoria in the previous 14 days, unless:

(i)    if the person falls within one or more of the categories in paras 27(a)-(g), the person complies with certain social-distancing, hygiene and mask-wearing precautions for 14 days;

(ii)    if the person falls within paras 27(h)-(r), the person complies with terms and conditions (including any quarantine direction) authorised by the State Emergency Coordinator (cl 5(e));

(d)    an authorised officer who is satisfied that a person has entered Western Australia contrary to cll 4 and 5, and that the person can leave Western Australia within a reasonable time and in an appropriate manner, must give the person a direction to leave (cl 9);

(e)    a person who has entered Western Australia contrary to cll 4 and 5, and believes that they are unable to leave at all, or within a reasonable time, or in an appropriate manner, must request a quarantine direction (cl 7); and an authorised officer must give such a person a quarantine direction, whether or not requested by the person (cl 8);

(f)    an authorised officer who gives a quarantine direction must give a direction to quarantine in a centre, unless a self-quarantine direction is given (cl 10);

(g)    a self-quarantine direction may be given where an authorised officer is satisfied, inter alia, that a Western Australian resident has satisfactorily completed 14 days of supervised quarantine elsewhere in Australia (cl 11).

Approach to assessment of risk

70    The Directions state that their purpose is to limit the spread of COVID-19. They evidently aim to do so by limiting the numbers of people who enter Western Australia in order to reduce the probability that people infected with SARS-CoV-2 will enter.

71    As Associate Professor Lokuge points out, the most effective way for Western Australia to limit the risk to public health would be to close the borders to all travellers. However, for legal, economic and social reasons, exempt travellers are permitted to enter Western Australia. The formulation of the border restrictions plainly involves a balancing by the second respondent of public health risks against the other considerations. However, these reasons are only concerned with the public health risks posed by COVID-19.

72    The issues to be determined are concerned with the risk that a disease which is not presently amongst the Western Australian population, may enter the population in the future. There are many uncertainties about whether the disease might enter, the ways it might spread and the effectiveness of measures for the control of its entry and spread. The issues, accordingly, involve making predictions about what may happen in the future in hypothetical scenarios. What is known is that in the worst-case scenario, there may be catastrophic consequences for the population. These circumstances call for identification of principles that ought to be applied when making decisions about measures to protect against such risks to public health.

73    In his evidence, Dr Robertson referred to the “precautionary principle” and the need to take a precautionary approach. The application of these concepts in managing public health risks was explained by Associate Professor Lokuge in her report of 26 June 2020. She considers the probability of importation of COVID-19 into Western Australia from a State where community transmission cannot be known or quantified from reported data. It is her opinion that such uncertainty mandates the application of risk management principles including the “precautionary principle”⁶. Associate Professor Lokuge quotes from a paper entitled, “The Precautionary Principle: Protecting Public Health, the Environment and the Future of our Children”, published by the World Health Organisation⁷:

Public health is inherently about identifying and avoiding risks to the health of populations as well as in identifying and implementing positive interventions to improve population health. However, traditional public health interventions have generally focused on removing hazards that have already been identified. In contrast, the precautionary principle states that action should be taken to prevent harm ‘even if some cause and effect relationships are not fully established scientifically’. The precautionary principle therefore seeks to shift health and environmental policy from a strategy of ‘reaction’ to a strategy of ‘precaution’.

(Citations omitted.)

74    In her oral evidence, Associate Professor Lokuge explained that the precautionary principle indicates that action should be taken based upon an understanding of what might occur, because if authorities wait for clear evidence of its occurrence, it may be too late⁸. She said that in the context of a pandemic disease of high mortality, which is highly infectious and of rapid spread, the principle requires, from a public health perspective, implementation of all available and effective mitigation measures⁹. In the joint experts’ report, Associate Professor Lokuge states that this is exactly the situation that risk management approaches and the precautionary principle were developed to address — substantive uncertainty where important harms are plausible¹⁰.

75    In her oral evidence, Associate Professor Lokuge indicated that the precautionary principle is an accepted principle of management of infectious diseases, saying that in all the outbreaks she has worked on, particularly those involving high risk pathogens, it is the guidance that she and her colleagues have used¹¹. She considers that Western Australia has been applying that principle in its management of the pandemic¹².

76    I accept the evidence of Associate Professor Lokuge that the precautionary principle is an accepted principle of management of a pandemic which involves the potential for grave public health risks.

77    Dr Robertson’s evidence is consistent with Associate Professor Lokuge’s view. In addition, Professor Blakely expressly states that he recognises and supports the application of the precautionary principle in the management of the current pandemic¹³. Associate Professor Senanayake, commenting upon Associate Professor Lokuge’s report of 26 June 2020, agreed with the paragraph in which she indicated that the circumstances mandated the application of the precautionary principle¹⁴.

78    A further principle emphasised by Dr Robertson and Associate Professor Lokuge is that public health risks are a function of probability and impact. In other words, the risks are measured, not merely by the probability that COVID-19 will be imported into the Western Australian population, but by the seriousness of health impacts for that population if that probability manifests¹⁵.

79    The approaches described by Dr Robertson and Associate Professor Lokuge have the advantage of being logical. Although the probability that a particular health risk will manifest may be small, if its consequences are potentially catastrophic, a precautionary approach is required. This means, from a purely public health perspective, all reasonable and effective measures to mitigate that risk should ideally be put in place. This analysis, however, does not take into account the legal, economic and social considerations that must, in practice, be considered.

80    The adoption of these approaches to public health risks reflects, in the case of Dr Robertson, the perspective of a public official who is required to make recommendations and decisions for the practical protection of the health of a population. Their adoption reflects, in the case of Associate Professor Lokuge, the considerable practical experience of an expert used to providing advice on the management of pandemics where the consequences of wrong decisions may be severe. Professor Blakely’s frank acknowledgement of the considerable uncertainties involved in providing estimates of probabilities and that his estimates should not form the sole basis for decision-making, also reflects a precautionary approach.

81    In contrast, the evidence of other experts tended to underplay or overlook these approaches. For example, Associate Professor Senanayake said he was “optimistic” that an uncontrollable outbreak would not arise from the introduction of a single case¹⁶. Professor Collignon “preferred” a targeted quarantine approach over State-wide border restrictions without comparing their advantages and disadvantages from a public health perspective¹⁷. These opinions do not reflect a precautionary approach to public health risks.

The potential public health consequences if COVID-19 were introduced into the Western Australian population

82    I will commence by considering the second of the integers of risk to public health identified by Dr Robertson and Associate Professor Lokuge, namely the potential health impacts, or consequences, if COVID-19 were introduced into the Western Australian population.

83    If the disease were introduced, transmission within the community may be able to be controlled, as it was in some States and Territories, or it may be uncontrolled for at least some period of time, as appears to be the case in Victoria. By the time an outbreak is brought under control, there may be substantial health consequences.

84    First, COVID-19 has the capacity to kill. In Australia, the crude case fatality rate has been up to 1.4% (although the figure does not take into account cases not tested for)¹⁸. As at 27 July 2020, there were 161 deaths in Australia out of 14,935 confirmed cases¹⁹. The disease affects vulnerable groups with co-morbidities, such as people in aged care facilities and Indigenous communities, the most substantially. The Australian crude case fatality rate for those aged 65–79 has been 3.1%, and for those aged 80 and over has been 22.7%²⁰.

85    Second, COVID-19 can cause illness, the symptoms of which can include fever, coughing, sore throat, fatigue, shortness of breath, nasal congestion, headache, conjunctivitis, diarrhoea, loss of taste or smell, skin rash or discoloration of fingers or toes²¹.

86    Third, about 13% of cases notified in Australia have required admission to hospital²². Dr Robertson estimates that for 100 new cases per day (or 1,000 active cases), 130 hospital beds and 25 ICU beds would be required, and 14 deaths expected; while for 500 new cases per day (or 5,000 active cases), 650 hospital beds and 124 ICU beds would be required, and 70 deaths expected. These numbers would remain within the capacity of the Western Australian health system to manage, but would substantially increase the burden upon the health system²³.

87    Fourth, there is presently no known vaccine or cure for COVID-19²⁴.

88    Fifth, SARS-CoV-2 is infectious during the incubation period (ie before symptoms first develop). Further, approximately 18–42% of cases remain asymptomatic, but are capable of transmitting the virus. There are others who remain only mildly symptomatic²⁵. The median incubation time is 5–6 days (with a range between 1 to 14 days)²⁶. A case is capable of transmitting the virus 1–3 days before exhibiting symptoms. Therefore, a person may spread the virus without the person knowing or suspecting that they have it²⁷.

89    Sixth, SARS-CoV-2 is highly infectious and is transmitted exponentially. If no measures are implemented to prevent spread of the virus, the growth rate (also known as the reproduction number) is approximately 2.3–2.5 (ie every infected person will infect on average 2.3–2.5 contacts)²⁸ . The expert witnesses agree that rapid uncontrolled transmission resulting from the introduction of a single infected individual to a community has been demonstrated to have occurred in multiple settings where there is otherwise good surveillance and testing control. So, for example, Associate Professor Senanayake comments that so-called “super-spreading events” have been associated with coronavirus infections such as SARS, MERS and COVID-19, and that this is a possible impact, but not a certain one, if a case is imported into Western Australia²⁹.

90    Seventh, testing for SARS-CoV-2 is imperfect. An infectious person can return a negative test³⁰.

91    Eighth, there are disincentives for testing, and the likelihood of each person undertaking testing is not equal. As Associate Professor Lokuge observes, whether a person is likely to submit to testing depends upon factors such as whether the person is a casual employee, has sick leave available and has access to Medicare, and upon the person’s personal background and experiences³¹. It can be inferred that some people who display symptoms or come into contact with a known carrier will not undergo testing and will continue to work and go about their routine activities unless and until they become seriously ill, and may, in the meantime, create chains of infection and clusters of COVID-19. As a result, the spread of the disease is stochastic³². As transmission relies on human interaction, the way such diseases move through a community reflects patterns of human interaction and disease control within that community, rather than an average across all of the population³³.

92    Associate Professor Lokuge estimates that the uptake of testing for people with symptoms consistent with COVID-19 is at most 50%³⁴. Associate Professor Senanayake agrees³⁵. I accept that this is a reasonable estimate.

93    Ninth, there is necessarily a lag time in identifying chains and clusters. Once an identified chain commences, or a cluster builds up, the hotspot can be locked down. As the median incubation period is 5–6 days, it will take at least a week for a hotspot to be identified, by which time a number of generations of transmission may have already occurred³⁶. In that time, many people can be expected to have entered or left the hotspot, and potentially picked up and spread the infection outside of the identified hotspot.

94    Tenth, the effectiveness of Containment Measures depends upon the willingness and ability of people to comply with the measures. The disease is mainly spread by contact with infectious respiratory droplets or surfaces on which these droplets have gathered³⁷. The prevention of transmission through such measures depends upon, for example, people keeping 1.5 metres away from each other, washing their hands regularly, increasing the frequency with which they clean surfaces, not gathering in large groups at family, religious and social events, and wearing a face mask. The experts agree that in many cases, people will not comply with these measures³⁸.

95    Eleventh, as COVID-19 has only recently emerged in the human population, the clinical, epidemiological and scientific knowledge base in respect of the disease is limited. There are a number of uncertainties about the disease, reflected in the qualifications placed upon many of the facts agreed in the Draft Consolidated Special Case. So, for example, the longer term health impacts of COVID-19 on those infected are not known, but the literature indicates it may include cardiac problems, coagulopathy, stroke, and, in children, multisystem inflammatory disorder³⁹. There is also the possibility of airborne transmission of the virus, but more research is needed⁴⁰.

96    Associate Professor Lokuge is the only expert who has addressed the potential health impacts or consequences of the introduction of COVID-19 into the Western Australian population to a substantial extent.

97    In the joint experts’ report, Associate Professor Lokuge notes that all the experts agree that the most concerning form of transmission is community transmission which is undetected, and therefore uncontrolled⁴¹. That is a theme which recurs throughout the experts’ evidence. However, it is not the sole determinant of the risk of transmission, as there have been transmissions even when it is known that people are at substantial risk of being infectious: for example, there have been transmissions from people in quarantine.

98    In her report of 26 June 2020, Associate Professor Lokuge states that the impact depends upon the magnitude of transmission which results from a case, the severity of morbidity and mortality in those infected, and the response measures that are required to control further transmissions⁴². She observes that the importation of even a single case can result in a rapid increase in community transmission, with major consequences. She observes that systems for detection in Australia are comparatively strong, but not infallible⁴³.

99    Associate Professor Lokuge notes that if a case were imported into Western Australia and initiated community transmission, that would occur in a setting where COVID-19 has been eliminated. She states that large outbreaks are particularly likely if the disease is introduced to settings where containment measures have been relaxed and large gatherings are occurring. Of 1,100 individual “super-spreading” events which have been reported globally and catalogued, 480 of them involved over 100 cases⁴⁴.

100    Associate Professor Lokuge states that the impact depends upon the severity of the disease among those infected. It will have the highest morbidity and mortality in those who are older, obese, have chronic cardiac disease, chronic respiratory disease, neurological disease, liver disease or cancer. The introduction of disease into settings, such as residential care facilities, where such risk factors are more prevalent may have devastating consequences⁴⁵. She states that if interstate importations result in community transmission in Western Australia, this will necessitate implementing stricter Community Isolation and Containment Measures, as are currently in place in Victoria⁴⁶.

101    To assess the severity of the impact, Associate Professor Lokuge applies a decision-tree algorithm from a 2009 paper entitled, “Assessing the Risk from Emerging Infections”, published in Epidemiology and Infection. Algorithms were developed to assess risks from emerging infectious diseases, and give a qualitative estimation of the probability and impact of introduction of the disease. Under an algorithm for, “Impact on Human Health: the Scale of Harm Caused by the Infectious Threat in Terms of Morbidity and Mortality”, the impact is characterised as “minimal”, “low”, “moderate”, “high” or “very high” based upon the answers to a series of questions⁴⁷. In applying the decision-tree model to SARS-CoV-2, Associate Professor Lokuge answers the following questions, “Yes”:

    Is there human-to-human spread?

    Is the population susceptible?

    Does it cause severe disease in humans?

    Would a significant number of people be affected?

    Is it highly infectious to humans?

102    Associate Professor Lokuge gives a negative answer to the final question, “Are effective interventions [treatment or prophylaxis] available?”. On this basis, her opinion is that the impact of SARS-CoV-2 is categorised as “very high”, the highest possible categorisation⁴⁸.

103    Associate Professor Lokuge provided a supplementary report dated 21 July 2020. She notes that cases in the Victorian community from late May onwards have been linked to a few breaches, or perhaps a single breach, in quarantine measures. She states that the epidemiological links between the initial seeding events and cases of community transmission, were not recognised until genomic data was available several weeks later. This indicates that there must have been undetected community transmission occurring throughout that period between multiple community clusters. By the time new cases of locally-acquired disease had been identified in June 2020, levels of undetected community transmission were such that they resulted in a rapid increase in case numbers over the subsequent month, with 1,280 cases reported in the four days prior to her report. This was despite Victoria having the highest testing rates per capita and the highest levels of compliance with social distancing measures in Australia. This underlines that such measures are unlikely to control transmission within the community in Western Australia and, instead, if an importation initiates transmission, a return to much more stringent social distancing measures will be required⁴⁹. I understand this to mean that a fuller range of Personal Isolation Measures, Community Isolation Measures and Containment Measures would be required.

104    Associate Professor Lokuge states that COVID-19 is not distributed evenly throughout the population. It is also highly clustered. In infectious disease modelling, the randomness in transmission is termed “stochasticity”, while the clustered and uneven distribution of cases is termed “heterogeneity”. As transmission relies on human interaction, the way such diseases move through a community reflects patterns of interaction and disease control within that community, rather than an average across all of the population. She states that even if testing and contact tracing levels are high across the population in general, as they are in Victoria, there are likely to be subgroups of the community with very different characteristics, who are both highly interconnected with each other and the wider community through family, work and social networks, and at the same time have lower engagement in response measures due to social, economic, cultural and demographic barriers to uptake. It is in such subgroups that disease would most widely circulate undetected when introduced. In the case of COVID-19, the period from initial introduction to multiple generations of unrecognised transmission is a few weeks at most in such a context. The characteristics of transmission in the early phases of the resurgence in Victoria reflects such a pattern⁵⁰.

105    Associate Professor Lokuge states that Western Australia has now returned to a level of social interaction similar to pre-COVID-19 levels and that very few, if any, in the Western Australian population would be immune to SARS-CoV-2. This means that the most likely reproduction number if a case of COVID-19 were introduced into the Western Australian community now would be similar to numbers seen early in the transmission of the virus globally, that is 2–2.5 people infected per case. She considers that the risk of large outbreaks from even a single importation is high in Western Australia, and higher than it was in Victoria when the current resurgence commenced⁵¹. It must be recognised that Western Australia is already at risk of importation even under the current border restrictions, but here Associate Professor Lokuge is talking about impact rather than probability.

106    Associate Professor Lokuge’s opinion is that the current resurgence of transmission in Victoria, and the seeding of transmissions into New South Wales from that resurgence, demonstrate the potential risks related to COVID-19 transmission. Even when testing, contact tracing and compliance with social distancing measures is high, rapid amplification can occur from one or a few cases, to the point where much more stringent measures preventing movement and interaction are required to slow the transmissions⁵².

107    Dr Robertson considers that if cases of COVID-19 were imported into Western Australia, the exponential nature of the spread of the disease, and the possible failure to identify the disease immediately (although possibly mitigated by testing), would mean that the impact of the disease spread is serious, particularly if vulnerable communities, such as health care workers or Indigenous communities, were impacted. He states that as physical distancing decreases, the potential for outbreaks increases proportionately and the importation of one case could have more significant consequences⁵³. He considers that “super-spreader” events would be of particular concern, as they can lead to rapid increase in both the initial number and subsequent growth of disease cases⁵⁴.

108    These aspects of the evidence of Associate Professor Lokuge and Dr Robertson have not been contradicted by the other experts. Professor Blakely acknowledges that the modelling he carried out does not take into account impact or consequences — it does not differentiate between an outbreak consisting of 5 cases or 500. I accept the opinions of Associate Professor Lokuge and Dr Robertson.

109    If COVID-19 is introduced into the Western Australian population, community transmission may be controlled or uncontrolled. Not all cases of community transmission will become uncontrolled, particularly as Australia has good systems of community controls and tracing, as the experts agreed. The initial outbreaks from February to April 2020 were brought under control in all States and Territories, except perhaps Victoria. However, if left uncontrolled for any substantial length of time, outbreaks will cause very severe consequences for the health of the Western Australian population. In the worst-case, the consequences could be catastrophic.

The utility and effectiveness of the border restrictions

110    The border restrictions aim to guard the Western Australian population from the importation of COVID-19 through the arrival of infected persons from other States and Territories. The utility of the border restrictions depends upon, first, whether there is such a risk of importation to guard against, and, second, whether they are effective to guard against such risk.

111    The last identified case of community transmission of SARS-CoV-2 in Western Australia was on 12 April 2020⁵⁵.

112    The parties have agreed in the Draft Consolidated Special Case that community transmission of any infectious disease can only be regarded as having ceased when two incubation periods have expired from the last confirmed case of community transmission within a community, although there remains a risk that there may be undetected disease within the community⁵⁶.

113    The expert witnesses agree that where there have been no reported cases of community transmission of COVID-19 for two incubation periods (28 days), the disease can be described as “eliminated”. The experts also agree that where there have been no reported cases of community transmission with an unknown source of infection for 28 days, that is “as low risk a situation as can reasonably be hoped for”⁵⁷ .

114    The experts’ opinions do not, however, suggest that it will be certain that there are no ongoing community transmissions. First, the use of the expression “eliminated” may be misleading because, as the experts agreed, there may be underlying transmissions of asymptomatic or mildly symptomatic cases that have not been reported. Second, both the application of the 28 day period and the existence of risk more generally must be subject to the particular circumstances of, and changing circumstances within, a relevant State or Territory. For example, if the borders of a particular State have remained open, or have recently been opened to another State where there is ongoing community transmission, there must be a risk of as yet unidentified community transmission within the first State.

115    Since the Western Australian borders are not completely sealed to overseas and interstate travellers, and as the incubation period for COVID-19 is up to 14 days before symptoms are produced, it is possible that there may be unreported and undetected cases of community transmission in Western Australia. However, in the absence of any identified cases since 12 April 2020, it can be inferred that the probability of any present community transmission in Western Australia is negligible.

116    On 27 July 2020, data released by the Commonwealth Department of Health showed that there were 4,542 active cases in Victoria and 160 active cases in New South Wales. In both States, there was community transmission. I infer that in both States, there are cases where the source of the infection had not been identified because the total number of cases identified as “locally acquired — contact not identified” increased by 127 in Victoria and by 2 in New South Wales in the period from 24 to 27 July 2020⁵⁸.

117    I conclude that the probability of there being any community transmission of COVID-19 in Western Australia at present is negligible. However, there is ongoing community transmission, both from known and unknown sources, elsewhere in Australia. The experts specifically agree in their joint report that border restrictions are important to ensure higher transmission risk populations do not spread COVID-19 to lower transmission risk populations. Therefore, the border restrictions aim to guard the Western Australian population against an ongoing risk.

118    It is necessary to next consider the effectiveness of the border restrictions in guarding against the importation of COVID-19 cases from interstate.

The numbers of people entering Western Australia from interstate before and after the border restrictions

119    The respondents plead in para 39C of their Second Amended Defence that, on average, approximately 5,000 people per day entered Western Australia in 2019, and approximately 3,500–4,000 per day entered in March 2020. Those figures are the subject of non-admissions in the Amended Reply. However, the parties have proceeded on the basis that they are correct. For example, Associate Professor Senanayake and Associate Professor Lokuge were asked to assume that those were the numbers of arrivals from interstate.

120    These figures are not consistent with the figures agreed in the Draft Consolidated Special Case, which states that 2,995,133 passengers arrived in Western Australia from interstate between 1 January 2019 and 31 December 2019. That is an average of 8,205 arrivals per day. The parties have also agreed that 172,634 passengers arrived between 1 March 2020 and 31 March 2020. That is an average of 5,568 arrivals per day⁵⁹. It has not been explained how the agreed figures were calculated, but it may be that they reflect the inclusion of arrivals from overseas. I will proceed on the basis that the figures of 5,000 per day in 2019 and 3,500–4,000 per day in March 2020 are correct because the hearing was conducted on that basis.

121    The parties also proceeded upon the basis that since the border restrictions were implemented, the number of people arriving in Western Australia from interstate has been, on average, about 470 per day (approximately 3,290 per week). Dr Robertson and Associate Professor Lokuge were asked to assume the correctness of the daily figure. Dr Robertson also said in oral evidence that he believed the weekly figure to be correct⁶⁰.

122    The experts disagree about the effectiveness of the border restrictions. I will consider their opinions upon this issue.

Dr Robertson

123    When Dr Robertson provided advice to the second respondent on 29 March 2020 about border restrictions, he stated that closing the Western Australian borders would have an impact on slowing the spread of COVID-19, but it would not reduce the risk significantly further than that achieved by measures already in place, such as isolation and restricting opening of retail outlets and mass gatherings⁶¹.

124    In his first report of 24 June 2020, Dr Robertson expresses the opinion that the likelihood of importation of COVID-19 into Western Australia would be significantly higher if the Directions are revoked. He states that the likelihood is a function of the number of people crossing the border and the probability that those people are infected with the virus — the more people who come in from other States or Territories with community transmission, the greater the likelihood that the virus will be imported. Dr Robertson assesses the likelihood of importation of COVID-19 into Western Australia if the Directions are revoked as low (less than 10%) — that assessment being based upon the present low level of travellers coming into Western Australia. He considers that if the State were to move back towards its travel figures from six months ago, the risk would be higher. Dr Robertson considers that if the Directions are maintained, the likelihood of importation of the disease would be negligible. Although there remains a risk of both re-introduction of the disease and subsequent community transmission from exempt travellers, particularly those who are not subject to quarantine, Dr Robertson considers that with the Directions in place, the risk of disease re-introduction and community transmission in Western Australia is less than 1%⁶².

125    As Dr Robertson acknowledged in re-examination, the percentage figures he gives are estimates rather than calculations. They are based upon his perceptions of risk, rather than reflective of any attempt at modelling or mathematical calculation⁶³. I consider that the figures themselves should not be given any weight. Nevertheless, his impressionistic assessments have some value.

126    In his second report of 3 July 2020, Dr Robertson states that when he recommended the border restrictions, he noted weaknesses around people not self-isolating properly and essential workers not isolating at all. He came to the view that “closing” the borders would have the effect of slowing the spread of COVID-19. Dr Robertson took the view that the border restrictions would only be effective at that time because to be effective there had to be a differential risk across the country of developing the disease — if Western Australia had a similar rate of spread to the other States at that time, the closures would have far less impact⁶⁴.

127    In the joint experts’ report, Dr Robertson states that Containment and Personal Isolation Measures are likely to be adequate in most circumstances to prevent further community transmission, provided there is no re-introduction of the disease back into the community. He considers that border measures, through home or hotel quarantine of people crossing the border, will generally be required to prevent such re-introduction. He also considers that Containment and Personal Isolation Measures become less adequate over time, as people become less compliant with them. They also become less effective as restrictions are eased, with less physical distancing, increased mixing and less compliance with hygiene measures, leading to an increased potential for outbreaks; but also leading to a more normally functioning society with positive impacts on general health, through more available health services, and mental health stressors⁶⁵.

128    Dr Robertson considers that the United Kingdom provides a classic example of the consequences of not using border measures. The United Kingdom used Containment and Personal Isolation Measures, but did not lock their borders or impose quarantine until after the first wave, resulting in significantly increased case numbers and deaths⁶⁶.

Associate Professor Lokuge

129    In her report of 26 June 2020, Associate Professor Lokuge states that border measures are an accepted and essential component of the public health response to the control of infectious disease outbreaks. She states that the most effective way to control an infectious disease is through early and stringent border controls. She considers that this is particularly important for preventing cases from entering a region that is free of disease. She says that the most effective border measure would be to prevent people moving from regions with active community transmission into regions that do not have active community transmission⁶⁷.

130    In her first report, Associate Professor Lokuge, at one point, states that by “stringent” border measures, she means closing borders to all those outside the region⁶⁸, but later describes the Western Australian border restrictions (which allow exempt travellers) as “stringent”⁶⁹. The applicants and the Commonwealth criticise that contradiction, but I consider that it is merely a slip or inconsistency in language, since Associate Professor Lokuge also expressly considers the current level of arrivals in her report. She states that the number of interstate visitors to Western Australia influences the probability of importation of COVID-19. She considers that the reduction in numbers of visitors to the present level decreases the probability of a case entering, on average, to 9–13% of what it would have been without the border restrictions (comparing the current figure of 470 arrivals per day to 3,500 to 5,000 per day before the pandemic)⁷⁰. The average figures applied by Associate Professor Lokuge do not account for the stochastic nature of outbreaks of COVID-19, but are of some utility in assessing the impact of the border restrictions. The inference she draws that the probability has been substantially decreased is logical.

131    Associate Professor Lokuge considers that effective and early border closures would have prevented the resurgence in Victoria and its subsequent seeding into New South Wales. She notes that all other States and Territories have now closed their borders to Victoria, including New South Wales and the Australian Capital Territory, which had previously maintained open borders. She states that this is consistent with border controls being an essential component of public health measures for controlling COVID-19⁷¹.

132    Associate Professor Lokuge states that in many settings where a single importation resulted in uncontrolled community transmission (such as Singapore and Melbourne), initial amplification occurred within subgroups that had lower engagement in control measures such as testing, community messaging, quarantine and isolation. The drivers for this included poverty, uncertain and casual work, language barriers, distrust of authorities, temporary visa status, overcrowded and unsanitary living conditions, and specific cultural and religious beliefs and behaviours. She notes that in modelling terms, this variation within subgroups is referred to as heterogeneity, and is why heterogeneity must be included in models of highly clustered infectious diseases such as COVID-19 if they are to be valid. She agrees that if all of these drivers were addressed and every person in the community had equal, equitable and high access to and uptake of all control measures, the risks of rapid, uncontrolled amplification of transmission would be greatly reduced. Until this is achieved, it remains her opinion that the border restrictions under the Directions provide the most immediate, effective and feasible strategy for preventing the impact of an importation⁷².

133    In the joint experts’ report, Associate Professor Lokuge states that border measures are aimed at preventing infected individuals entering the community, whereas all other measures can only reduce the risk that infected individuals pass on disease to the community. She notes that, as all experts have agreed, even in settings where non-border measures have been implemented to a high degree, uncontrolled outbreaks can occur. She considers that border measures will always have additional value above and beyond other measures, as they are the only measures that can prevent entry of disease. She states that all places which have achieved good control of COVID-19 have had border controls as an integral part of their response. She therefore disagrees with Professor Collignon’s view that other measures can replace border measures⁷³.

134    Associate Professor Lokuge considers that as Western Australia has now greatly reduced implementation of key measures such as prevention of mass gatherings, and compliance with other internal control measures (such as allowing contact with non-household members), border measures assume even greater importance. Associate Professor Lokuge notes that a critical factor in the effectiveness of these measures is compliance, and that many people do not comply with such measures⁷⁴.

135    Associate Professor Lokuge disagrees with the view that the Australian Capital Territory is a good example of open borders not increasing risk. She states that the Australian Capital Territory in fact had a higher rate of interstate importations than she would expect for its size. In her opinion, the reason it did not experience a large outbreak is due both to chance and factors that decrease risk of spread in the Australian Capital Territory, such as an educated and wealthy population, less high-density social housing and the absence of risky industries such as abattoirs⁷⁵.

Professor Blakely

136    In the joint experts’ report, Professor Blakely states that border restrictions are an essential tool in preventing the incursion of COVID-19 into a destination State. Professor Blakely also states that a key benefit to a State like Western Australia in having eliminated community transmission is that the population can live in a near-normal manner (but with measures like physical distancing still in place). For this reason, he considers that border controls are important to a State like Western Australia which has no community transmission⁷⁶.

Professor Collignon

137    Professor Collignon provided a report dated 7 July 2020. He was asked to assume that there are “Common Measures” adopted by all States and Territories in response to COVID-19. These are:

    measures taken to isolate all cases, and quarantine all potential cases, of COVID-19;

    measures taken (or that have been previously taken and could be reimposed) to prevent or limit the movement of persons geographically within a State (including stay-at-home directions, maximum distance travel-from-home directions and regulations preventing movement into and out of particular locations, including remote communities);

    measures taken (or that have been previously taken and could be reimposed) to limit the number of people that may meet in a group;

    measures encouraging individuals to practice social distancing;

    measures encouraging individuals to undertake regular handwashing;

    measures facilitating contact tracing; and

    measures taken to increase the testing of potential COVID-19 cases.

138    It may be noted that the Common Measures are a combination of the measures described in the Draft Consolidated Special Case as “Personal Isolation Measures”, “Containment Measures” and “Community Isolation Measures”, but excluding border restrictions.

139    Professor Collignon’s report deals with efficacy of the border restrictions in containing transmissions of COVID-19 in Western Australia in comparison to the Common Measures. His opinion is that the additional contributions from the Western Australian border restrictions in containing the spread of COVID-19 within Western Australia “are fairly small” compared to the Common Measures⁷⁷. Professor Collignon states that, looking at the data, the number of locally acquired cases from interstate visitors appears to have been very small. He states that there are also minimal (if any) obvious extra effects on the number of new COVID-19 cases diagnosed in Australia after 5 April 2020 (the approximate date of the border closures of most States). Some cases associated with interstate travel did occur after 5 April 2020, both in Australia and Western Australia, but there were already rapidly falling numbers of new cases. The fall in numbers associated with interstate travel paralleled the rapid fall in overall numbers of new cases in Australia. Professor Collignon infers, based on his expertise, that it was other factors, rather than border closures, which were responsible for the fall in cases related to interstate travel. He also states that in general for the effects from any intervention to be seen, it would usually take about ten days (two average incubation periods), so that 15 April 2020 is a more appropriate date to use as a cut-off to examine any likely benefits from State border closures on the number of new cases of COVID-19. He states that he cannot see any obvious additional benefits from State border closures, judged by looking at the number of new cases in Australia, using 15 April 2020 as the relevant date⁷⁸.

140    Professor Collignon states that the peak of new cases in Australia was 20 March 2020. After that date there was a rapid decrease in the daily numbers of new cases. The decrease was in community transmission in Australia and also those cases linked to interstate travel. These decreases started to occur around the same time. Professor Collignon states, based on his experience, that the decrease occurred because the Australian population as a whole appeared to be mostly following recommendations from the health authorities, including physical and social distancing, together with actions such as reducing crowd numbers and closure of places where large groups could mingle together. He states that the closure of the international borders, case finding and isolation of people with infection and quarantining of return travellers and close contacts of those with COVID-19 cases all made major contributions to lowering the numbers of new cases⁷⁹.

141    Professor Collignon states that there was a total of 5,914 cases of COVID-19 until 5 July 2020 (not including cases occurring and diagnosed within Victoria). Of these, only 102 cases were linked to interstate travel, representing a very small proportion (0.17%) of all cases diagnosed. He also states that the numbers were falling well before any interstate border restrictions were applied. He considers that, accordingly, border restrictions for Western Australia, the Northern Territory, Queensland, Tasmania and South Australia appear to have had minimal additional impacts and took effect when the effective reproduction rate for COVID-19 in all States and Territories in Australia was already less than 1⁸⁰.

142    In his report, Professor Collignon further states that New South Wales and the Australian Capital Territory did not have border closures, but had no marked increases in cases and, especially, no increase from interstate visitors compared to the States that closed their borders. He points out that there have been no additional cases from interstate visitors in the Australian Capital Territory since 25 March 2020 despite extensive testing⁸¹.

143    Professor Collignon concludes that, compared to all the other measures that can be taken to minimise the risk of transmission of COVID-19 in a community that has a low prevalence of COVID-19 currently, the closure of State borders has added minimal additional protection⁸².

144    In the joint experts’ report, Professor Collignon states that uncontrolled transmission is a very major concern, but the factors that favour uncontrolled transmission have much more to do with lack of physical distancing, crowded indoor venues, people going to work or visiting when they have respiratory symptoms, delays in testing, isolation and quarantining than closure of State borders. Professor Collignon also considers that closure of State or internal borders would have had minimal impact in the South Korean, Singaporean, Victorian and Tasmanian examples⁸³. Even if that is so, I consider that Professor Collignon’s opinion misses the point that, in other circumstances, border restrictions may stop the spread of the virus by preventing it from entering a region.

145    Professor Collignon states in the joint experts’ report that he also believes that the Western Australian border restrictions would make a very much smaller contribution compared to the Common Measures in the future⁸⁴. He does not explain the basis for this view. In my opinion, it calls for an explanation, particularly given that the Common Measures have not protected Victoria from its current wave of community transmission.

146    I infer that Professor Collignon has taken his conclusion that the border restrictions played a fairly small role compared to the Common Measures in eliminating COVID-19 from Western Australia by 13 April 2020 and extrapolated that into a conclusion that border measures would play a similarly small role in the future. There are flaws in this reasoning. First, it fails to recognise that border restrictions and the Common Measures play quite different roles — the border restrictions aim to prevent COVID-19 from being imported, whereas the Common Measures aim to restrict or contain the community spread of COVID-19 once imported. Second, if COVID-19 were imported, border restrictions would continue to have the important role of minimising further importations.

147    Under cross-examination, Professor Collignon conceded that border restrictions do operate to reduce the risk of importation of COVID-19. He said, “If you drop your number of people travelling by 50 per cent, broadly, I would think you have got a 50 per cent reduction in your risk”⁸⁵.

Associate Professor Senanayake

148    In his report of 7 July 2020, Associate Professor Senanayake states that a number of measures — targeted restriction for entry from regions with surges of community transmission, physical distancing, avoidance of visits to aged care facilities and hospitals for 14 days, hand hygiene, border screening, mandatory use of the COVIDSafe application, with or without testing — are as effective as a full border closure⁸⁶. This appears to be inconsistent with his later agreement in the joint experts’ report that border controls are important to ensure that COVID-19 is not spread from higher-risk populations to lower-risk populations⁸⁷. It is also partly inconsistent with his agreement that many people do not comply with some of these measures⁸⁸.

149    In the joint experts’ report, Associate Professor Senanayake states that he believes that a region which has eliminated or “aggressively suppressed” COVID-19 will reduce that risk further with a border closure, but to be truly effective, it would ideally have to be a rigid border closure with zero travel from inside that State or into that State (including freight). He says that even though this is not the case currently with Western Australia, it has not had any recent cases of community transmission with an unknown source. He considers that this is because Western Australia is exercising discretion with regard to who is let in and that such discretion should continue when considering allowing low-risk populations in⁸⁹.

150    Associate Professor Senanayake notes that the Australian Capital Territory is surrounded by New South Wales, has commuters travelling to work daily from New South Wales, and has never had border restrictions. Despite outbreaks in New South Wales and Victoria, the Australian Capital Territory has not had community transmission with an unknown source for more than 28 days⁹⁰. It may be noted that the Australian Capital Territory has in fact now introduced border restrictions in respect of Victoria⁹¹.

Conclusion upon the effectiveness of the border restrictions

151    I accept the opinions of Associate Professor Lokuge and Professor Blakely that border measures are an accepted and effective component of the public health response to the control of infectious disease outbreaks. That is reflected in the fact of controls upon Australia’s international borders, and the effectiveness of border closures in New Zealand. The experts specifically agree in the joint experts’ report that border restrictions are important to ensure higher risk transmission populations do not spread COVID-19 to lower risk transmission populations. I accept that they are a particularly important measure for the protection of a region with no community transmission from importation of the virus from regions with ongoing community transmission.

152    Professor Collignon’s opinion that the contribution of the Western Australian border restrictions commencing on 5 April 2020 to containing the spread of COVID-19 in Western Australia at that time was “fairly small” compared to the Common Measures is consistent with the view that Dr Robertson expressed in his advice on 29 March 2020. That is because the growth rate was already below 1 when the border restrictions were introduced⁹². I accept this aspect of Professor Collignon’s evidence.

153    However, the border restrictions were left in place to guard against future importations of the disease to Western Australia. Professor Collignon’s opinion about the relative contributions the border restrictions and the Common Measures made to containing the spread of COVID-19 in April 2020 has very limited relevance to the question of the protection that border restrictions presently offer, particularly given the large outbreak in Victoria which commenced in mid-June 2020. In the joint experts’ report Professor Collignon accepts that residents of Melbourne should be restricted in their movements⁹³. Since the border restrictions at least substantially achieve this, it must follow that they have value in preventing the spread of COVID-19 to Western Australia. Although Professor Collignon’s report points to New South Wales as an example of a State with open borders which had not had substantial interstate transmissions, that position changed with the Crossroads Hotel outbreak in early July 2020. There have been at least 56 cases associated with that outbreak to date. The cluster commenced with a single traveller from Victoria⁹⁴. Dr Robertson’s opinion, which I accept, is that the traveller would not have been permitted to enter Western Australia under the border restrictions⁹⁵.

154    As Associate Professor Lokuge and Professor Blakely acknowledge, chance plays a substantial part in the incidence of community transmissions of COVID-19. The Australian Capital Territory did not have substantial border restrictions until it restricted travel from Victoria, yet had not recorded community transmissions of COVID-19 in the 28 days prior to the hearing⁹⁶. However, Associate Professor Lokuge points out that the Australian Capital Territory had more interstate transmissions than she would expect for its size. I accept the opinion of Associate Professor Lokuge that the absence of community transmissions was a combination of good luck, demographics and the absence of high risk industries. In contrast, New South Wales was unfortunate with the Crossroads Hotel outbreak having commenced with a single Victorian traveller. It does not follow from the fact that the Australian Capital Territory has not had an outbreak, that the outcome would have been the same in Western Australia if the border restrictions had not been in place.

155    It is possible that even if the border restrictions were not in place in Western Australia, there would have been no imported cases of COVID-19 resulting in community transmissions. However, the border restrictions substantially reduced the layer of chance, given that there would otherwise have then been substantially greater numbers of potential carriers entering Western Australia.

156    I consider that the opinions of Associate Professor Senanayake and Professor Collignon as to the relative contributions of the border restrictions and the Common Measures to the control of COVID-19 say little about the efficacy of the border restrictions. There is no doubt that the Common Measures would be vital to controlling the spread of COVID-19 if the disease entered Western Australia. There is no suggestion that the Common Measures should be removed. The importance of the border restrictions is that they substantially limit the chance that COVID-19 will be imported from States where there is ongoing community transmission into Western Australia, where there are no cases of COVID-19 present.

157    I accept the opinions of Dr Robertson and Associate Professor Lokuge that the border restrictions are presently making a substantial contribution to keeping Western Australia free of COVID-19. An indication of the extent of that contribution is provided by comparing the numbers of people entering Western Australia before and after the restrictions. There is a direct correlation between numbers of people travelling to Western Australia and the probability that an infected person will enter the State. This was the evidence of Dr Robertson and Associate Professor Lokuge. Ultimately, Professor Collignon also accepted that the risk was reduced in proportion to the reduction in the numbers travelling because of the border restrictions. The numbers are now about 470 per day, compared to between 3,500 and 5,000 people per day before the restrictions were imposed. Associate Professor Lokuge’s estimation that the border restrictions have reduced the probability of importing the virus to 9%–13%⁹⁷ of what it would otherwise have been is somewhat crude, but informative in this context. I find that the border restrictions have very substantially reduced the probability that the virus will be imported into Western Australia from interstate. I find that they are effective.

The effectiveness of the border restrictions over the Common Measures

158    There is an issue of whether the border restrictions have additional protective value over the Common Measures, and to what extent. I have addressed this in part, but will now do so directly.

159    Professor Collignon considers that the additional contributions from the Directions in containing the spread of COVID-19 within Western Australia was “fairly small” compared to the contribution of the Common Measures, and that these relative contributions will continue in the future⁹⁸.

160    Western Australia has in place all of the Common Measures, with the exception that it does not have an absolute limit on the numbers of people permitted to meet in groups (although it has distancing requirements which apply to groups)⁹⁹.

161    I accept Associate Professor Lokuge’s evidence that border measures reduce the risk of infected individuals entering the community, whereas other measures reduce the risk that infected individuals who enter the community will transmit disease into the community. Associate Professor Lokuge notes that the experts have agreed that even in settings where non-border measures have been implemented to a high degree, uncontrolled outbreaks can occur¹⁰⁰.

162    The Common Measures and the border restrictions contained in the Directions target different risks of transmission of COVID-19. The border restrictions aim to restrict the chance that COVID-19 will be imported. The Common Measures aim to limit or control the spread of the virus once it has been imported. The Common Measures do not affect the risk of the importation of the virus.

163    I accept Associate Professor Lokuge’s opinion that border measures have additional value above and beyond other measures, as they are the only measures that prevent entry of disease.

164    The experts have agreed that border restrictions are important to ensuring that COVID-19 does not spread from populations with a higher risk of transmission to a population with a lower risk of transmission¹⁰¹. That is particularly important in respect of the transfer of the virus to communities with no community transmission from those with ongoing community transmission. I have already indicated that I accept the opinions of Associate Professor Lokuge and Professor Blakely about the importance and effectiveness of border controls.

165    The experts have agreed that rapid, uncontrolled transmission resulting from the introduction of a single infected individual to a community has been demonstrated to have occurred in multiple settings where there is otherwise good surveillance and testing¹⁰². It follows that any importation of COVID-19 has the potential to cause an uncontrolled outbreak.

166    Western Australia has been free of community transmission of COVID-19 since 12 April 2020. The extent of the contribution to the “elimination” of the virus by that date does not matter. What is relevant is that since then, the border restrictions have contributed to the absence of imported cases. I have concluded that the border measures have very substantially reduced the chance of importation of the virus by dramatically reducing the number of travellers.

167    I also accept Professor Blakely’s opinion that border restrictions are important to Western Australia, which has eliminated community transmission, because the restrictions allow the population to live in a near-normal manner. Dr Robertson’s evidence is that the absence of imported cases has allowed the relaxation of some of the Common Measures. These include relaxing the limits on the gathering of large groups, and easing restrictions upon travel, weddings, funerals, sporting events, restaurants, gyms and other businesses¹⁰³. That is important in light of the evidence of Dr Robertson and Associate Professor Senanayake that measures like the Common Measures become less adequate over time because people become more complacent and less compliant with them¹⁰⁴. The experts also agree in their joint report that many people do not follow such measures.

168    Professor Collignon argues that closing the State borders may have the perverse effect of increasing the potential for spread within communities. He considers that once there is very little spread, and particularly if people think the virus has been eliminated where they live, there is a desire and expectation to go back to “normal”. He considers that if the virus is then reintroduced into the community, it may then be rapidly spread. He states that this is why the most important aspect of control is not closing State borders, but for the population as a whole to follow practices that decrease the risk of the virus spreading¹⁰⁵.

169    There are some difficulties with accepting Professor Collignon’s argument. The logic of his argument may suggest that it is better to have some outbreaks from time to time so that people do not become complacent in respect of their compliance with the Common Measures. However, this does not sit well with Professor Collignon’s acceptance that even the introduction of a single case can lead to an uncontrolled outbreak. Professor Collignon’s reasoning also implicitly acknowledges that in any circumstance where there had been no community transmission there would be the risk of complacency setting in, even in the absence of border restrictions. The “perverse effect” described by Professor Collignon is substantially outweighed by the value of border restrictions in reducing the chance of importation of COVID-19 into the Western Australian community. I do not accept that this is an adequate reason for removal of the border restrictions.

170    The border restrictions and the Common Measures together provide substantial ongoing protection to the Western Australian population. It is not possible to disaggregate the relative contributions they have made since COVID-19 was “eliminated” from Western Australia. It may be that there were some cases imported despite the border restrictions, but which died out without being detected because of the Common Measures. It may be that there were cases that were not imported because of the border restrictions, but which would not have been controlled by the Common Measures. It is impossible to know. What can be said is that the border restrictions have very substantially reduced the chance of importation of COVID-19. I find that the combined operation of the border restrictions and the Common Measures has prevented the re-introduction of community transmission of COVID-19 into Western Australia since 13 April 2020.

171    I conclude that the border restrictions offer a tangible and substantial layer of protection to the Western Australian community over the protection offered by the Common Measures.

The probability of an infectious person entering Western Australia from interstate if the border restrictions were removed

172    I have accepted that the risk to public health that the border restrictions guard against is a function of probability and impact. In para 15B(a) of their Amended Reply, the applicants plead that the probability of a person infected with COVID-19 from an unknown source travelling to Western Australia while infectious is so small as to provide no reasonable justification for the border restrictions.

173    Accordingly, the probability of persons carrying the virus into Western Australia from interstate in the absence of the border restrictions must be determined. An appropriate starting point is to assume the complete removal of the border restrictions without replacement by alternative measures.

174    Several issues arise. First, there are differences between the experts as to whether a quantitative assessment of the probability can usefully be made, or whether only a qualitative assessment is useful. Associate Professor Senanayake and Professor Blakey have carried out quantitative assessments, but accept that they are subject to a number of uncertainties and limitations. Professor Collignon, and perhaps Dr Robertson, consider that quantitative assessments should be made, but that a qualitative assessment is more useful. Associate Professor Lokuge considers that a quantitative assessment is too uncertain to be useful and has carried out a qualitative assessment using a decision-tree algorithm.

175    Second, there are differences between the experts about the level of probability, described in qualitative terms, of persons entering Western Australia while infected.

176    Third, it is relevant to compare the probability of an infected person entering Western Australia if the border restrictions were removed with the existing probability of that occurring while the border restrictions are in place in order to understand the marginal, or incremental, difference made by the border restrictions.

177    Fourth, there is an issue of the probability of persons who enter Western Australia while infectious actually transmitting the virus to the Western Australian population and, in addition, the probability of such transmissions causing uncontrolled outbreaks.

178    Fifth, there is an issue as to the extent to which imposing alternative measures to border restrictions, such as mandatory quarantining of interstate travellers and wearing of face masks, would reduce the probability of COVID-19 entering the Western Australian population.

179    I will commence by examining the opinions of the experts upon these issues.

Associate Professor Senanayake

180    The applicants asked Associate Professor Senanayake for his opinion upon the statistical probability of a person who is carrying SARS-CoV-2 travelling to Western Australia while infectious.

181    In his report of 7 July 2020, Associate Professor Senanayake estimates that the daily probability of an infectious person arriving in Western Australia from elsewhere in Australia in the period from 25 June to 1 July 2020 with the border restrictions in place is one case per 68,188 people; for persons coming from anywhere other than Victoria, his estimate was one case per 8,100,000 people. If the border restrictions were removed and approximately 5,000 people per day arrived in Western Australia, his estimate is one case per 12 people, or 8.3%; and if people from Victoria were excluded, the probability would be one case per 1,780 people, or 0.06%¹⁰⁶.

182    Associate Professor Senanayake states that he is unable to estimate the probability of a person arriving in Western Australia carrying the virus and then causing an uncontrolled outbreak of COVID-19¹⁰⁷. He said he would be “optimistic” that an uncontrollable outbreak would not arise within Western Australia from the introduction of a single case¹⁰⁸. That optimism must be considered to be qualified by the unanimous view of the experts in their subsequent joint report that rapid uncontrolled transmission resulting from the introduction of a single infected individual to a community has been demonstrated to have occurred in multiple settings where there is otherwise good surveillance and testing control.

183    There are some difficulties with Associate Professor Senanayake’s quantitative estimates. Several of his assumptions appear to be overly optimistic. For example, he assumed that 95-100% of contacts would be identified through contact tracing, even though, as he accepted under cross-examination, that assumption was not based on any firm data¹⁰⁹. He accepted that there was no factual basis for his estimate that 95% of people exhibiting flu-like symptoms would obtain testing¹¹⁰. He accepted that it was no longer reasonable, having regard to the circumstances of the recent outbreaks in Victoria, to exclude hotel quarantine cases from his calculations¹¹¹. Further, Associate Professor Senanayake’s calculations represent averages, which make no allowance for the stochastic nature of the transmission of COVID-19¹¹². The calculations fail to include any range for uncertainty. He also accepted that he had attached the wrong version of an appendix to his report and that had led to inaccuracies in the report. Associate Professor Senanayake accepted that where he had relied upon the assumptions and calculations in the appendix, it would be reasonable to say that the Court should place no weight on those parts of the report¹¹³.

184    I would have considered that calculations as to average probabilities would have some, although limited, value in decision-making if satisfied that they were done accurately and included an assessment of the margin for error. However, I do not think that any weight can be placed upon Associate Professor Senanayake’s calculations. That is not to say that his opinions upon other issues should not be accorded due consideration.

Professor Blakely

185    Professor Blakely’s report of 8 July 2020 aims to quantify the risk of transmission of SARS-CoV-2 into Western Australia from other States and Territories. There are essentially two parts to Professor Blakely’s report: first, estimating the probability of an undetected infectious person travelling to Western Australia; and, second, estimating the extent to which the risks of such a case entering Western Australia and the risk of an outbreak could be mitigated or reduced through alternative measures. Importantly, he was asked to have regard only to people who are asymptomatic or pre-symptomatic, and to disregard those who are contacts of a known case and those who are symptomatic.

186    Professor Blakely commences the first part of the task by estimating the number of infected people who had not been detected as a ratio to each daily notified case of a locally acquired infection without a known source. He assumed that undetected infected people are the major risk for cross-border transmission. Such undetected infected people were assumed to consist of asymptomatic and pre-symptomatic cases¹¹⁴.

187    Professor Blakely acknowledges that the proportion of asymptomatic people is not known with precision, but said that it probably lies in the range of 20% to 50%, and he used a figure of 35% as a “best estimate”. He assumes that asymptomatic cases would pose a risk of infection for 16 days. For the purposes of calculating the ratio, Professor Blakely assumes 20 new infections per day in a State or Territory other than Western Australia, of whom 7 per day (35%) would be asymptomatic and pose a risk of carrying the virus across the border. On this basis, there would be 112 “risky asymptomatic people” at any one point in time (ie seven asymptomatic infections per day multiplied by 16 days until, on average, they are not infectious). This produces 5.6 “risky asymptomatic people” per each newly infected person (ie 112 divided by 20 new infections per day). For the 65% of people who become symptomatic (ie 13 out of 20 new infections per day), Professor Blakely assumes it would take 7 days from infection to diagnosis. There would be 91 “risky pre-symptomatic but destined to be symptomatic people” at any one point in time (ie 13 per day multiplied by 7 days). This produces 4.55 such people per each newly infected person (ie 91 divided by 20). Combining these figures gives 5.6 + 4.55 = 10.15 risky people per daily new infection¹¹⁵.

188    Professor Blakely then observes that the approximately ten risky people per daily new infection is the “true infection”. However, the new infections are based on actual reported cases. Asymptomatic people will not be detected in the absence of widespread community testing and, in addition, Polymerase Chain Reaction (PCR) testing for SARS-CoV-2 (the current best testing option) misses some 15% of infections. These factors raise the ratio. However, offsetting this is the fact that community-wide testing and contact tracing will “unearth” some of these risky people. Professor Blakely acknowledges that the net effect of these considerations is difficult to quantify. He assumes a net zero impact on average, but allowed for wide uncertainty by applying a range of 5 to 20 “risky people per detected and reported case”¹¹⁶.

189    Professor Blakely concludes, accordingly, that for every daily reported case of locally acquired infection that cannot be tracked back to a known contact, there are about ten other undetected infected people at that point in time. On a given day, if 20 cases are reported in a locality and approximately the same number had been reported in the last week (ie assuming an effective reproduction rate of 1), then there might be another 200 undetected infected people in that locality at the same time. Professor Blakely notes that implicit in this calculation is that the greatest risk of cross-border transmission is from undetected people infected through community transmission. He says that while there are other risks of transmission (ie people who are infected despite completing quarantining, staff associated with quarantining and contacts of known cases who do not isolate and instead travel), they are assumed to be minimal¹¹⁷.

190    Professor Blakely then considers the number of locally acquired cases with an unidentified source in each State and Territory in the 28 days up to 4 July 2020. He assumes that if the Western Australian borders were reopened, there would be half as many arrivals as before the border restrictions were implemented.

191    Professor Blakely noted that there had been no locally acquired cases with no identified source in the last 28 days in the Australian Capital Territory, the Northern Territory, Queensland, South Australia and Tasmania and concluded that there is a small or negligible risk from those States and Territories¹¹⁸.

192    Professor Blakely then calculates the probability of a single person in New South Wales having a locally acquired case with no identified source by reference to the population of that State. For 35,000 arrivals per month (half the pre-COVID figures), the expected number of people carrying the virus would be 0.004. The probability of one such person boarding a plane to Western Australia from New South Wales was 0.4% per month. Professor Blakely also calculates, using the same method, that the number of such people arriving from Victoria (excluding the locked-down postcodes) would range from 0.56 – 4.35 per month¹¹⁹.

193    Professor Blakely was also asked to calculate the probability of an infected person starting an outbreak in Western Australia, assuming the border restrictions were removed. He considers that while the reproduction rate of SARS-CoV-2 is 2.5 in its “natural state”, the rate would be reduced to 2 by Western Australia’s Containment Measures. This means that, on average, one infected person would spread the virus to two others. Professor Blakely says whether any single case will infect others is inherently stochastic, so that some may not spread the virus at all, but that may be compensated for by so-called “super-spreaders”¹²⁰.

194    Professor Blakely states that it is not known what percentage of people entering Western Australia from another State or Territory would generate an outbreak. He considers that it seems safe to assume that at least half of incoming infected people would generate an outbreak, with the upper limit being that all would do so. He says that it is unknown whether any outbreak would be small and quickly stamped out, or large and could not be stamped out. It would depend largely upon how good the Western Australian surveillance, testing, contact tracing and outbreak management systems are. It would also depend, inevitably, on chance. The growth and size of an outbreak could depend upon whether some of the infected persons are super-spreaders, whether they work in an aged care home, and other matters of chance¹²¹.

195    The second part of Professor Blakely’s report is concerned with whether the risk of SARS-CoV-2 entering Western Australia and the risk of an outbreak could be mitigated or reduced by imposing a number of alternative measures. He attended to that task by adapting a recent simulation study by Professor Wilson et al published at medRxiv, a repository for breaking research that has not yet been peer-reviewed, modelling the chance of infection and outbreaks occurring if New Zealand borders were opened to Australian arrivals¹²². I will consider the results later.

196    Professor Blakely reaches the following overall conclusions¹²³:

The risk of a cross-border incursion of SARS-CoV-2 due to arrivals from NT, QLD, SA, TAS and ACT is – as of early July – negligible, assuming that known contacts of cases in the origin S&T do not travel and that people coming out of quarantine in the origin S&T before travelling are not infected (i.e. quarantine was done well).

The risk of a cross-border incursion from NSW, where it appears community transmission is very low – possibly even eliminated – is very low, especially if border control measures like mask wearing for 14 days on arrival are instituted.

The risk of cross-border incursion from regions of Victoria outside of the currently locked-down hotspots is high (e.g. an average of one week to two months until an outbreak occurred with air arrivals at half the volume they were pre-COVID-19), very uncertain, and unlikely to be mitigated sufficiently by border controls (e.g. border screening, inflight masks, mask wearing for first 14 days of arrival, testing on days 3 and 12 after arrival, might reduce this risk to an outbreak occurring an average or 0.3 to 2.3 years after loosening border restrictions). Moreover, since these calculations were done (4 July), the risk has increased in Melbourne (as of 7 July) due to the escalating number of cases.

However, any decision of whether to open the borders and with what controls in place would depend on a risk assessment by WA authorities, balancing all of: the risk of incursion and outbreaks occurring (the focus of this Report); the effectiveness and cost of WA authorities to stamp out any outbreak; the risk of a large outbreak leaving residual community transmission with the associated morbidity and mortality and economic losses.

197    Professor Blakely’s calculations were done on 4 July 2020. During the hearing, he undertook further calculations based on figures available up to 27 July 2020 in respect of Victoria. He used the actual numbers of reported cases in the previous week, an average of 403 per day. Professor Blakely calculates, assuming 44,000 Victorians travelling to Western Australia per month (half the pre-COVID-19 numbers), that 21.84 undetected infected Victorians would arrive each month into Western Australia. On that basis, and assuming that travel was unrestricted, he calculates a 100% chance of at least one outbreak of COVID-19 in Western Australia each month¹²⁴.

198    Adapting and applying the Wilson simulation model to the alternative scenario for travellers from Victoria to Western Australia of the implementation of exit and entry screening plus the use of masks on planes, Professor Blakely calculates the probability of an outbreak at 99.99%. With those measures and PCR testing on the second and twelfth days after entry, it is 99.8%. With all of those measures plus mandatory mask wearing in the first 14 days in Western Australia, it reduces to 75.7%. With quarantining for the first 14 days, together with exit and entry screening and masks on planes, the probability reduces to 62.3%¹²⁵.

199    Professor Blakely also conducted further modelling using the figures as at 23 July 2020 in respect of travellers from New South Wales to Western Australia. He estimates that there is a 4% chance that there will be a single undetected infected traveller in a month. In other words, assuming 35,000 travellers per month from New South Wales, 0.04 travellers would be in that category. He estimates a 3.62% probability of at least one outbreak per month in Western Australia¹²⁶.

200    When Professor Blakely modelled the scenario for travellers from New South Wales of exit and entry screening plus the use of masks on planes, the probability of an outbreak is 1.74%. With those measures and PCR testing on the second and twelfth days after entry, it is 1.13%. With all of those measures plus mandatory mask wearing in the first 14 days in Western Australia, it reduces to 0.26%. With quarantining for the first 14 days and exit and entry screening and masks on planes, the probability reduces to 0.18%¹²⁷.

201    In their joint expert report of 23 July 2020, the experts were asked to answer the question, “What is the probability of an infected case of COVID-19 arriving at the border of Western Australia?”. Professor Blakely’s answer is that there is a low — but not zero — probability of an importation from a jurisdiction with no detected cases of community transmission (ie locally acquired cases of an unknown source). He notes that in his report he had estimated the probability of an infected traveller arriving from New South Wales based on New South Wales cases up to 4 July 2020 at 0.004 per month, but that there is considerable uncertainty about this estimate. He states that, for example, it might range from half or even a quarter of this value, to twice or even four times this value. It would also vary depending upon other characteristics, such as the socio-economic status of travellers. It would also change over time. Using case data for New South Wales up to and including 16 July 2020 (eight days after his report) had increased the probability approximately eight-fold. Professor Blakely considers that a sensible ‘rule of thumb’ is that borders should be closed to States and Territories with evidence of community transmission in the last 28 days. However, he also considers that the risk to a destination State varies with the number of travellers from the origin State, and the State population size¹²⁸.

202    The uncertainties conceded by Professor Blakely are reflected in a number of assumptions he makes. In his report, he refers to the necessity to make assumptions because of a lack of solid data¹²⁹. He states that the exact proportion of people who are infected with SARS-CoV-2 but remain asymptomatic is not known with precision, but probably lies in the range of 20–50%, and used a figure of 35%. Professor Blakely agreed in his oral evidence that this area provides a source of uncertainty¹³⁰.

203    In the simulation model used by Professor Blakely, an assumption is made that the majority of people would get tested if they showed symptoms. Dr Robertson thinks this an underestimate, particularly if the symptoms were mild. In his oral evidence, Professor Blakely considers this to be a reasonable objection¹³¹. Associate Professor Lokuge points to flu tracking data which suggested that the uptake of testing in the community of people with symptoms consistent with COVID-19 is at most 50%, and is likely to be even lower in some subgroups of the population¹³².

204    Dr Robertson also considers that the assumption in the model that entry screening would pick up 50% of symptomatic cases is overly optimistic¹³³.