New Zealand Pork Industry Board v Director-General of the Ministry for Primary Industries
[2013] NZSC 154
•20 December 2013
| IN THE SUPREME COURT OF NEW ZEALAND |
| SC 36/2013 [2013] NZSC 154 |
| BETWEEN | THE NEW ZEALAND PORK INDUSTRY BOARD |
| AND | THE DIRECTOR-GENERAL OF THE MINISTRY FOR PRIMARY INDUSTRIES |
| AND | THE CHIEF TECHNICAL OFFICER AND BIOSECURITY NEW ZEALAND |
| AND | NATIONAL BEEKEEPERS ASSOCIATION OF NEW ZEALAND INC |
| Hearing: | 26 June 2013 |
Court: | Elias CJ, McGrath, William Young, Glazebrook and Arnold JJ |
Counsel: | F M R Cooke QC and J B Kaye for Appellant |
Judgment: | 20 December 2013 |
JUDGMENT OF THE COURT
AThe appeal is dismissed.
BThe appellant is to pay costs of $25,000 to the first and second respondents collectively, plus reasonable disbursements as fixed by the Registrar.
____________________________________________________________________
REASONS
Para No
Elias CJ [1]
McGrath, William Young, Glazebrook and Arnold JJ [92]
ELIAS CJ
The Director-General of the Ministry for Primary Industries[1] is empowered by s 22(1) of the Biosecurity Act 1993 to set import health standards for the effective management of risks associated with the importation of risk goods.[2] “Risk goods” are defined to include “any organism … that may … cause unwanted harm to natural and physical resources or human health in New Zealand”.[3] Before an import health standard can be made, the Director-General is obliged to obtain a recommendation from a chief technical officer[4] after the chief technical officer has consulted with those the chief technical officer considers to be “representative of the classes of persons having an interest in the standard”.[5] Under s 22A of the Act, enacted in 2008,[6] if a person consulted by the chief technical officer raises “a significant concern” about the science that has been employed, the Director-General may obtain a report from an independent review panel “to review whether, in developing an import health standard, there has been sufficient regard to the scientific evidence” about which the “significant concern” has been raised.[7] On receiving the report, the Director-General is required, “as soon as is reasonably practicable” to “determine the issue in dispute after taking into account the findings and recommendations of the independent review panel, giving reasons for that determination”.[8]
[1]The Ministry of Agriculture and Forestry merged with the Ministry of Fisheries and the New Zealand Food Safety Authority and was renamed the Ministry for Primary Industries as of 30 April 2012.
[2]Biosecurity Act 1993, s 22(1). In these reasons, references to the Biosecurity Act are to the Act prior to the 2012 amendments, unless otherwise indicated. See below at [11]–[14].
[3]Section 2.
[4]Section 22(1).
[5]Section 22(6).
[6]Biosecurity Amendment Act (No 2) 2008, s 6.
[7]Section 22A(1).
[8]Section 22A(3).
In 2001, after an outbreak of foot-and-mouth disease in England and of Porcine Reproductive and Respiratory Syndrome (PRRS) in South Africa, a new scientific study suggested that the destructive viruses, including PRRS, can be transmitted through the ingestion by pigs of infected raw meat (and not simply by contact between live pigs or through infected semen, as had previously been thought). The Director-General for the Ministry then put in place a provisional import health standard which prevented the importation of raw pig meat from countries where the virus PRRS has been detected in pigs. Because of the need for urgency in responding to the new scientific information, the standard was adopted as a precautionary and interim response on an incomplete assessment of risk. It was based on a draft import risk analysis by the Ministry as to the risk of the virus entering New Zealand which did not extend to the likely effects if pigs in New Zealand were exposed to the virus, as is required by s 22(5)(b).
PRRS is highly contagious and destructive. New Zealand and Australia are two of only five countries in the developed world which are known to be free of it. It has the capacity to cause substantial harm to the pork industry.
The appeal concerns the import health standards issued by the Director‑General on 13 April 2011. They are substantially the same as the provisional import health standards issued in 2009, which were the ultimate replacements for the 2001 provisional import health standards. The new standards were adopted after a process of risk assessment and consultation which is outlined in what follows.
The standards set in April 2011 permit uncooked pig meat to be imported from the European Union, Canada, the United States and the Sonora State of Mexico (in all of which the virus is endemic) provided specified conditions are met. The two principal conditions are that the meat either consists of consumer-ready cuts packaged for direct retail sale not exceeding 3 kg per package (and excludes minced meat, the head and neck, and lymph tissues) or is processed in an approved transitional facility in New Zealand to meet the standards set for direct importation.[9] The purpose of confining the raw pig meat imported to high value retail cuts and of providing that larger raw pig products be first processed into such cuts in an approved facility in New Zealand is to reduce the risk that infected raw meat will be fed to pigs in New Zealand. Key to the decision of the Director-General in setting the 2011 standards is his assessment that the risk to New Zealand is effectively managed by the conditions set in the standards.
[9]Transitional facilities must be approved under s 39 of the Biosecurity Act.
The issue on the appeal is not whether the standards set in April 2011 are appropriate to manage the risk effectively. The only issue for the Court is whether the statutory processes, designed to assess and manage biosecurity risk effectively, have been properly followed.
Insistence on strict adherence to the procedures required by the statute is not to elevate form over substance.[10] The requirement that specified processes be followed reflects the legislative judgment that those processes enable better substantive decision-making in an area where assessment of risk is necessarily a matter of judgment incapable of exact proof and where the adverse consequences of faulty judgment may be very damaging to New Zealand because of its dependency on primary production and the advantages it enjoys because of its remoteness. The system employed depends on private contribution to the decision-making by interested parties, in a form of regulation not uncommon in recent years, which harnesses additional expertise outside government and transfers some costs of government regulation.
[10]As was suggested in the decision appealed from by Harrison and Stevens JJ: New Zealand Pork Industry Board v Director-General of the Ministry of Agriculture and Forestry [2013] NZCA 65 (Harrison, Stevens and White JJ) [New Zealand Pork Industry Board (CA)] at [96]–[97].
Well-known and costly errors in management of biosecurity in recent years have occurred despite the statutory controls available to the Director-General. As a direct result, amendments to the legislation in 2008 provided for more extensive participation in risk management assessment under s 22A, to ensure that contentious biosecurity decisions are made on the best available scientific information and that the science is subject to some independent scrutiny. A purpose of the independent review process introduced in 2008 with the new s 22A was described by the Chair of the Select Committee which recommended its adoption as being to address the problem that the Ministry was perceived as being “judge and jury on the issue”.[11]
[11](20 March 2008) 646 NZPD 15145.
It is clear from ss 22 and 22A of the Act that industry participants have an important contribution to make in setting import health standards. While no doubt they can be expected to represent their own interests, the debates at the time of the 2008 amendments do not suggest the views of industry participants are to be discounted for that reason and clearly there is common cause with the Ministry in seeking to manage risk of harm.[12] Conversely, the debates in Parliament treat the Ministry itself as a participant with its own views and constraints which have sometimes impeded its assessment of risk and whose assessments therefore benefit from independent and public scrutiny by others with the capacity and resources to make such contribution.[13] With this background, I do not think that it is helpful or appropriate to depreciate the interest of producers in the assessment of risk and its effective management because of their self-interest. The legislation, and the circumstances which led to its adoption, indicate the importance placed on the process prescribed and the participation of those affected.
[12]See (1 April 2008) 646 NZPD 15201–15202.
[13]See (1 April 2008) 646 NZPD 15204.
The appeal turns on the application of ss 22 and 22A of the Biosecurity Act. For the reasons to be given, and in disagreement with the other members of the Court, I am of the view that s 22 was not followed and that the import health standards issued on 13 April 2011 are invalid.
The Biosecurity Act 1993
The import health standards in issue were adopted under s 22 of the Biosecurity Act and after the process provided for in s 22A. Those provisions have since been repealed by the Biosecurity Law Reform Act 2012 which replaced them with equivalent provisions now contained in ss 22 to 24K of the Biosecurity Act. It has been accepted that the present appeal turns on the provisions in force at the time the import health standards challenged in the proceedings were adopted by the Director‑General.
The changes in the legislation do not affect the substance of the issues on the appeal, but they are organised in accordance with the sequence to be followed in developing an import health standard or in deciding not to issue one. The sequence may have been less easy to see in the provisions as enacted before 2012, although they were to the same effect. The former s 22 (“Import health standards”) has now been reorganised and split into s 22 (“Meaning of import health standard”), s 23 (“From draft to recommendation”), and s 24A (“Issue”). The former s 22A (“Process for independent review panel to be established”) is in substance reproduced, although reorganised, in the present s 24 (“Review”) and is now located before s 24A (“Issue”), making clear the sequence of decision-making (a little obscure in the former provisions which began with the issue of an import health standard and only then described the processes to be followed).
In the new s 24A, it is made explicit (what may not have been as easy to follow in the layout of the former provisions) that the Director-General’s decision to issue an import health standard following the recommendation of the chief technical officer is a decision taken “[a]fter … complying with section 24(4) [the determination of the dispute referred to a review panel under the former s 22A], if it applies” (emphasis added). The two determinations, where the review panel process is invoked, are sequential and distinct.
In what follows the questions raised in the appeal are dealt with by reference to the pre-2012 provisions, the course followed by the Court of Appeal and by counsel in their submissions. I have referred to the 2012 provisions only to point out that they are consistent with the meaning I attribute to the provisions in issue here.
The importation of “risk goods” is dealt with under Part 3 of the Act. Its purpose is described as being “to provide for the effective management of risks associated with the importation of risk goods”.[14] Section 22(1) permits the Director‑General to set import health standards “specifying the requirements to be met for the effective management of risks associated with the importation of risk goods before those goods may be imported”. The Director-General is not obliged to set an import health standard (without which such goods cannot be imported) if he considers that the purpose of the Act would not be met by any requirements that could be imposed.[15] The Director-General may set import health standards only “following the recommendation of a chief technical officer”.[16] To the extent relevant to the appeal, s 22 provides:
[14]Section 16.
[15]Section 22(3).
[16]Section 22(1).
22 Import health standards
(1)The Director-General may, following the recommendation of a chief technical officer, issue an import health standard specifying the requirements to be met for the effective management of risks associated with the importation of risk goods before those goods may be imported, moved from a biosecurity control area or a transitional facility, or given a biosecurity clearance; and may, in a like manner, amend or revoke any import health standard so issued.
(1A)An import health standard issued under this section applies to goods the importation of which involves, or might involve, an incidentally imported new organism.
…
(3)Nothing in this Act obliges the Director-General to have an import health standard in force for goods of any kind or description if, in the Director-General’s opinion, the requirements that could be imposed on the importation of those goods would not be sufficient to enable the purpose of this Part to be met if the importation of those goods were permitted.
(4)An import health standard issued under this section may apply to goods of a certain kind or description imported from—
(a)a country or countries specified in the import health standard; or
(b)countries of a kind or description specified in the import health standard; or
(c)all countries; or
(d)a location or locations specified in the import health standard.
(5)When making a recommendation to the Director-General in accordance with this section, the chief technical officer must have regard to the following matters:
(a)the likelihood that goods of the kind or description to be specified in the import health standard may bring organisms into New Zealand:
(b)the nature and possible effect on people, the New Zealand environment, and the New Zealand economy of any organisms that goods of the kind or description specified in the import health standard may bring into New Zealand:
(c)New Zealand’s international obligations:
(d)such other matters as the chief technical officer considers relevant to the purpose of this Part.
(6)Before making a recommendation to the Director-General on the issue or amendment of an import health standard, the chief technical officer must, unless the standard needs to be issued or amended urgently, or unless the chief technical considers that the amendment is minor, consult with those persons considered by the chief technical officer to be representative of the classes of persons having an interest in the standard.
(7)The consultation may be on the import health standard or on a document that analyses or assesses the risks associated with the goods or class of goods to which the goods belong.
…
In making a recommendation to the Director-General, the chief technical officer must have regard to “the likelihood that [the goods] may bring organisms into New Zealand”[17] and the likely effect of such organisms on “people, the New Zealand environment, and the New Zealand economy”.[18] The chief technical officer must also have regard to “New Zealand’s international obligations”[19] and to “such other matters as the chief technical officer considers relevant to the purpose of this Part”.[20]
[17]Section 22(5)(a).
[18]Section 22(5)(b).
[19]Section 22(5)(c).
[20]Section 22(5)(d).
Unless there is some urgency or the recommendation concerns amendment to a standard which is “minor”, the chief technical officer must, under s 22(6), first “consult with those persons considered by the chief technical officer to be representative of the classes of persons having an interest in the standard”. This consultation “may be on the import health standard or on a document that analyses or assesses the risks associated with the goods or class of goods to which the goods belong”.[21]
[21]Section 22(7).
Section 22A of the Act provides a process for reviewing whether, “in developing an import health standard, there has been sufficient regard to the scientific evidence about which a person consulted under section 22(6) has raised a significant concern”:
22A Process for independent review panel to be established
(1)The Director-General must, by notice in the Gazette, set out the process by which an independent review panel is to be established to review whether, in developing an import health standard, there has been sufficient regard to the scientific evidence about which a person consulted under section 22(6) has raised a significant concern.
(2) The notice required by subsection (1) must cover the following matters:
(a)the criteria for setting up an independent review panel; and
(b) how the Director-General will appoint an independent review panel, including the knowledge and experience required for appointees; and
(c)the procedures to be followed by—
(i)a person eligible to seek a review under subsection (1); and
(ii)an independent review panel, in undertaking its review; and
(d) the reporting requirements for an independent review panel.
(3) The Director-General must receive any report from an independent review panel and, as soon as is reasonably practicable, determine the issue in dispute after taking into account the findings and recommendations of the independent review panel, giving reasons for that determination.
…
As is provided in s 22A(3), on receipt of a report obtained under this process of independent review, it is the responsibility of the Director-General to “determine the issue in dispute”. He must give reasons for the determination.
The process the Director-General was obliged by s 22A to adopt and publish was published in the Gazette on 26 June 2008.[22] Clause 9 of the Gazette notice deals with the decision to grant a request for a review. In considering whether to accept a request for review, the Director-General “must take into account”:
(a)the extent to which the scientific evidence is or may be material to the measures in the proposed import health standard;
(b)the extent to which the request for review appears to be based on credible scientific evidence;
(c)whether the evidence has been the subject of an earlier review; and
(d)any other relevant matter.
Background
[22]“Biosecurity (Process for Establishing Independent Review Panel) Notice 2008” (26 June 2008) 104 New Zealand Gazette 2765.
As has been mentioned, the provisional standards were adopted in 2001 as a precautionary measure based simply on new appreciation of the risk of incursion of the PRRS virus through raw imported pig meat fed to pigs in New Zealand. Further work was necessary before a non-urgent standard could be adopted under s 22. Further work was also necessary to justify the restriction adopted for the interim, or any modification of it, in order to comply with New Zealand’s international obligations, particularly under the Agreement on the Application of Sanitary and Phytosanitary Measures (the SPS Agreement) adopted by the World Trade Organisation.[23] Before 2012 the chief technical officer had a general duty under s 22(5)(c) to “have regard to New Zealand’s international obligations”. That is a duty now made explicit in relation to the SPS Agreement under the new s 23(4)(c), enacted in 2012, which requires the chief technical officer to be “satisfied that the requirements proposed for inclusion in the standard are consistent with New Zealand’s obligations under the SPS Agreement”.
[23]The WTO Agreement on the Application of Sanitary and Phytosanitary Measures (signed 15 April 1994, entered into force 1 January 1995).
Under the SPS Agreement, restriction of trade for biosecurity reasons is permitted only where standards are justified by scientific principles and evidence,[24] including as to risk assessment,[25] and then only to the extent necessary to protect human, animal or plant life or health.[26] Such standards must not amount to a disguised restriction on trade.[27] Article 5 of the SPS Agreement permits provisional measures to be put in place “[i]n cases where the relevant scientific evidence is insufficient” on the basis of the available information.[28] The precautionary provisional standard set in 2001 was justified on this basis. Where a provisional standard is put in place, however, states are obliged to “seek to obtain the additional information necessary for a more objective assessment of risk” and to “review the sanitary or phytosanitary measure accordingly within a reasonable time”.[29] It was therefore incumbent on the Ministry once it had put in place the provisional standard to obtain the information it needed to make “a more objective assessment of risk” and to review the provisional standard.
[24]Article 2.2.
[25]Article 5.1.
[26]Article 2.2.
[27]Article 5.6.
[28]Article 5.7.
[29]Article 5.7.
The scientific assessment of risk and consideration of its effective management in review of the provisional 2001 measure took some time. Eventually, in July 2006, the Ministry released for consultation a paper assessing the risk of importing pig meat from countries with the virus, “Import Risk Analysis: Porcine reproductive and respiratory syndrome (PRRS) virus in pig meat”.[30] The analysis concluded that the likelihood of the virus being introduced through chilled or frozen imported pig meat was “low”[31] but not negligible.[32] It was acknowledged that, if the virus were introduced to New Zealand, the consequences would be “significant on affected farms, particularly in breeding units”.[33] The risk analysis therefore accepted that some sanitary standards were required. It pointed to the fact that the feeding of raw meat to pigs was illegal under the garbage feeding regulations[34] (so that “an exposure pathway” would exist only on those farms not complying with the regulations) [35] and that, if infection did occur, “the likelihood of spread to other pig farms would be low as long as standard biosecurity practices were observed”.[36]
[30]Noel Murray and Howard Pharo Import Risk Analysis: Porcine reproductive and respiratory syndrome (PRRS) virus in pig meat (Ministry of Agriculture and Forestry and Biosecurity New Zealand, 25 July 2006) [2006 Import Risk Analysis].
[31]At [1].
[32]At [4].
[33]At [4].
[34]Biosecurity (Meat and Food Waste for Pigs) Regulations 2005.
[35]At [2].
[36]At [3].
On the basis of this analysis of risk, the Ministry proposed that the ban on importing raw pig meat from countries affected by the PRRS virus be relaxed to permit the import of “consumer-ready, high value cuts” from countries not free of PRRS.[37] In addition to these “consumer-ready, high value cuts”, imports of other raw pig meat from the same countries was to be permitted for the purpose of further processing on arrival, “in an officially approved facility, into consumer-ready high value cuts”.[38] The reason for the restriction to consumer-ready high value cuts was the view that such cuts minimised trimming or cutting during preparation and were therefore less likely to generate raw waste which might be fed to pigs.[39]
[37]At [4].
[38]At [4].
[39]At [5.2.2.2].
The New Zealand Pork Industry Board is a statutory body set up under s 4 of the Pork Industry Board Act 1997 with the object under s 5 “to help in the attainment, in the interests of pig farmers, of the best possible net ongoing returns for New Zealand pigs, pork products, and co-products”, while being obliged to “have regard to the desirability of the pork industry’s making the best possible net ongoing contribution to the New Zealand economy”. The Board’s functions include increasing demand for New Zealand pork products in existing and new markets,[40] helping to “maintain the confidence of consumers of pork products in the New Zealand pork and pig industries”[41] and improving access to overseas markets.[42] Four directors are elected by pig farmers and up to two are appointed by the Minister because of relevant expertise.[43] The adoption of a replacement import health standard to manage the risk to the New Zealand pig population of the introduction of PRRS was a matter on which the Board’s functions made it “representative of the classes of persons having an interest in the standard” under s 22(6), requiring the chief technical officer to consult with the Board before making his recommendation to the Director-General.
[40]Section 6(1)(a).
[41]Section 6(1)(b).
[42]Section 6(1)(c).
[43]Section 13(2).
The Board was concerned about the assessment of risk which underlay the proposals for a new import health standard published by the Ministry. It expressed the view that kitchen and commercial waste is commonly fed to pigs in New Zealand. It pointed to a review by the Ministry which said that regulations to control the feeding of scraps to pigs[44] were unlikely to be effective, whether or not there was awareness of the regulations (a circumstance contributed to by the nature of pig farming in New Zealand in small populations).[45] The Board considered that there was insufficient scientific justification for the proposed standard and that it was speculative for the Ministry to say that the risk of incursion of the virus through the feeding of imported scraps to pigs would be acceptably reduced by confining imports to high value cuts or limiting processing in the way proposed.
[44]Clause 5 of the Biosecurity (Meat and Food Waste for Pigs) Regulations provides that pigs must not be fed untreated (raw) meat or food waste that contains or (under (b) of the definition of “untreated food waste” in s 4) has come into contact with untreated meat.
[45]2006 Import Risk Analysis, above n 30, at [4.2.5]–[4.2.5.3].
As a result of the feedback it received on its July 2006 paper, the Ministry commissioned a peer review of the risk assessment from international experts, which took until June 2007. It released its review of the submissions it had received on the risk assessment in June 2007. In August 2007 the Board commissioned Dr Neumann and Professor Morris, both veterinary scientists with experience in quantitative modelling of disease in animals, to model a quantitative assessment of the risk of the introduction of the PRRS virus through the importation of raw pig meat from countries affected by the virus. Their model predicted an average of 4.3 outbreaks of PRRS a year from the feeding of infected raw imported pig meat to pigs in New Zealand. The report obtained was given to the Director-General but did not lead to a change in the proposal. On 12 November 2007, the Ministry released draft import health standards for pig meat. Under them, import of “ready-to-cook, high value cuts of pig meat” would be permitted from the European Union, Canada, the United States, and the Sonora State of Mexico. The Board responded with a submission seeking rejection of the draft standard, relying on the quantitative risk model developed by Dr Neumann and Professor Morris. Following further consultation, the Director-General on 7 April 2009 published further provisional import health standards adopting the standard as proposed in the drafts upon which it had consulted.
By that time, the Biosecurity Amendment Act (No 2) 2008 had inserted s 22A into the Biosecurity Act and the Director-General had published the independent review panel notice in the Gazette.[46] On 28 May 2009 the Board requested the Director-General establish an independent review panel under s 22A, identifying nine matters of concern to it relating to the adequacy of the scientific evidence. On 7 August 2009 the Director-General agreed to the Board’s request and set up a review panel. The review panel was chaired by a New Zealand Queen’s Counsel and comprised in addition Professor Katharina Stärk (from the Royal Veterinary College, United Kingdom), Professor John Wilesmith (a retired professor and specialist in veterinary epidemiology from the United Kingdom) and Professor James McKean (a veterinary epidemiologist of Iowa State University, United States).
[46]Gazette notice, above n 22.
In September, terms of reference for the review panel were provided. Under them, the review panel was asked to review whether sufficient regard had been paid by the Ministry to the scientific evidence in nine areas (the first eight of which had been suggested by the Board in its request):
(a)The identification and analysis of potential hazards associated with the importation of pig meat and pig meat products.
(b)The likelihood that meat from slaughter weight pigs will contain infectious PRRS virus.
(c)The impact of changes to volumes of trade in pig meat as a result of the proposed changes in the [import health standards].
(d)The impact of changes to the volume and distribution of the waste stream as a result of the proposed changes in the [import health standards].
(e)The likelihood that PRRS-infected imported pig meat will be fed to [New Zealand] pigs and cause infection.
(f)The structure and inter-relatedness of the New Zealand commercial and non-commercial pig industries, and consequent exposure and spread risks.
(g)Importance and likelihood of aerosol and ‘area’ spread of PRRS virus between herds.
(h)Quantitative modelling of the risk of PRRS virus exposure and consequence, using the model developed during the [import risk analysis/import health standards] process.
(i)Each of the above issues sits within the context of the overall assessment of risk. The Panel should consider whether [the Ministry’s] overall treatment of the issues was reasonably open on all the evidence.
The review panel met by teleconference and videoconference and exchanged emails between November 2009 and March 2010.Its report to the Director-General was presented on 31 March 2010.[47]
The review panel findings and recommendations
[47]Report of the Independent Review Panel on the Provisional Import Health Standards for Pig Meat and Pig Meat Products for New Zealand (31 March 2010).
In its report, the review panel indicated that it had confined itself to the threat of the PRRS virus alone, since a study of other “potential hazards” (which had been included in the terms of reference) was “not feasible within the available time” and the review panel had been provided with information only about PRRS. With respect to other hazards, it suggested greater transparency around hazard identification and better communication about risks in general.
The review panel recommended that the import risk assessment should be redone to ensure that it took into account the relevant and current science and either filled or acknowledged and modelled for gaps in the analysis. It took the view that the Ministry had insufficiently acknowledged the limitations in scientific knowledge particularly in relation to infectiousness and treatment of affected pig meat. It considered that there were gaps in important parts of the risk analysis, including in assessment of the volume of trade likely. The Ministry’s risk assessment “did not address the impact of trade volume directly”, despite its importance because of the sensitivity of the risk estimate towards trade volume and the likelihood that risk of introduction would increase with a rise in the volume of trade. Nor did the Ministry sufficiently explain why it had assessed the extent of discarded pig meat at 3.8 per cent when a study in the United Kingdom had considered that its assessment of 7 per cent (based on survey) was an underestimate. This factor was “of key significance in [the Ministry’s] risk assessment”. The Ministry had justified the difference on the basis of an “unsubstantiated assumption” that a proportion of the United Kingdom cuts would be of cuts greater than 3 kg (which the review panel thought hard to reconcile with the dominance of supermarket sales in the United Kingdom, in which smaller cuts predominate).
The review panel expressed concern that “there is no reliable estimate presented of the amount of uncooked pork/pigmeat which is likely to be discarded from the various sources in New Zealand”. It thought it surprising that effort had not been made “to fill gaps in the necessary information base”, because the gaps created uncertainty in the risk analysis “especially where there is no assessment of the sensitivity of the analysis to the assumptions made on estimates”.
Nor was there adequate information about the New Zealand pig population, such as was needed to assess the risk of spread of any infection and to justify the Ministry’s view that observance of biosecurity measures would adequately manage the risk of spread. The review panel identified the need to obtain the current population data for pigs and obtain better data on farm size, type and location “in order to reduce the uncertainty about the epidemiological and economical consequences of a potential PRRSv introduction”. It identified recent clinical experiences in the United States and elsewhere that suggested the Ministry was “overly optimistic” in its view that movement limitations and biosecurity measures would contain virus movement.
The review panel considered that the Ministry’s “qualitative” assessment should have been supplemented by a “quantitative” analysis of risk, as the Ministry’s external reviewer had also recommended (although it thought that some of the uncertainties used in the sensitivity analysis in the Neumann model gave some validity to the Ministry’s criticisms of it). It recommended that the Neumann quantitative model be “updated with the currently available information by a joint technical working group, including epidemiologists experienced in quantitative risk assessment named and agreed by MAF and stakeholders, including [the Board]” and that the Ministry consider the use of empirical information of the extent to which infected pig meat might be fed to pigs to update the risk assessment. Recent studies since the risk assessment was undertaken by the Ministry needed to be factored in (especially relating to the percentage of pig meat discarded as waste and the spread of infection by area and air) in order to ensure that the risk assessment “is based on more robust scientific and epidemiological data”. Of some importance is the review panel’s recommendation that the results of the further modelling “should be presented in a report and should then be integrated in a revised version of the [import risk assessment]”. (It acknowledged that the conclusions of the original import risk assessment might remain unchanged.) Significantly, it said that “[i]f significant uncertain and influential model inputs are identified, the collection of additional data should be considered”.
The review panel was critical of the lack of transparency in the selection of independent reviewers by the Ministry and the “overall process involved in obtaining and reviewing the relevant scientific evidence”. It considered that “in some areas [the Ministry] made assumptions based on relatively little data”. To the extent that this was unavoidable, the lack of knowledge needed to be stated clearly, and needed to be taken into account in “a level of sensitivity for the key elements”. It acknowledged the limited resources available to the Ministry but expressed some concern that the Ministry had been “too hasty” in dismissing comments adverse to its position, including comments it had solicited from reviewers. It had “made assumptions for which there was considerable uncertainty and for which no or limited effort was reported to substantiate or refute in the course of conducting the [import risk assessment]”. The review panel took the view that the reassessments it recommended “would not take more than 6 to 9 months to complete, depending on the resources available”. And it recommended cooperation in meeting the recommendations between the Ministry and the Board.
Responding to the review panel recommendations
In April 2010 the Ministry convened a meeting of stakeholders, including representatives of the Board, to discuss the findings of the review panel, the key recommendations of the report, and how the Ministry might respond to its recommendations. At the meeting the Board took the view that, while it was supportive of a stakeholder working group to progress the recommendations through providing advice to the Director-General, there should be a new risk analysis rather than an attempt to “backfill” the existing import health standard. There was discussion about the key issues raised by the Ministry in the agenda for the meeting and an indication by representatives of the Ministry that it was undertaking further work.
In an update of 2 July, the Ministry advised the Board that it declined its offer to have further input into the preparation of the advice for the Director-General. It advised that the Ministry was considering what the review panel had recommended as to quantitative risk modelling and expected to be in a position to provide its advice to the Director General “in the near future”. It advised that the Director‑General “has determined that no further stakeholder input will be required prior to the Director-General deciding how he wishes to respond to the review panel’s recommendations”, and that “[s]hould the Director-General decide that further work is required, it is likely that there will be an opportunity at that stage for further input from stakeholders”.
Part of the work undertaken in the meantime by the Ministry was refinement of the quantitative model originally put forward by Dr Neumann and Professor Morris, as the Ministry’s communications had indicated would be happening. The Ministry sought input from some of the expert review panel members on its adjustment for some inadequacies in the values entered into the original model – such as the failure to acknowledge in the assessment of pig meat imports that a proportion of imported pig meat would continue to come from other countries that are free of PRRS, namely Australia, Finland and Sweden. The Ministry indicated that its work showed that the four most influential input variables in the sensitivity analysis were number of meals consumed annually by each pork-consuming household, proportion of pig carcasses likely to be contaminated at the time of slaughter, the concentration of PRRS virus in contaminated carcasses at slaughter, and the proportion of pork in a meal that is discarded as fresh raw scrap. The first variable was the same as that used by Neumann. The second and third variables were said to reflect the current state of knowledge and it was thought that further studies were unlikely to provide a change in understanding. The fourth variable (the proportion of raw pork trimmed off and discarded as raw meat prior to cooking) was different in that there was no data to substantiate the 10 per cent assumption identified by Neumann. In the end, the Ministry decided not to proceed with the survey it originally planned to remedy the gap in knowledge in respect of this factor. One of the expert members of the review panel, Professor Stärk, was asked to peer review the Ministry’s adaptation of the Neumann/Morris model. The Ministry’s reworked model, with the variables it determined, returned a mean result of one introduction into a pig farm every 4,635 years.
On 1 September 2010, the Director-General released what was described as his “decision” on the findings and recommendations of the review panel.[48] The decision was in the form of an advisory report from the chief technical officer, setting out the background and ending with 10 recommendations on which the Director-General indicated that he either accepted or declined, or noted, each recommendation by striking out the inapplicable outcome and dating and signing his name at the end of the document.
The decision of the Director-General of 30 August 2010
[48]This decision was made on 30 August 2010 and released on 1 September 2010.
The advice given by the chief technical officer was that the review panel’s expectation that the further work it recommended could be carried out in six to nine months “indicates to me that the intent of the [review panel] is for additional work to be undertaken to modify or support the conclusions in the underlying risk analysis and the resulting risk management measures in the pork import health standards, rather than for [the Ministry] to begin the process anew”.
The chief technical officer’s recommendation on the approach to be followed, accepted by the Director-General, was that in deciding whether to accept and act “immediately” on any recommendation, the Director-General should apply a standard of “materiality” to the statutory objective of “effective management of the risks associated with the importation of risk goods”[49]:
In this instance, it means the degree to which a particular recommendation contributes to determining the effectiveness of available risk management measures for importation of pork. In some cases, a recommendation may be material to risk management in a general sense but decisions on whether to accept the recommendation and how best to implement it need not be taken prior to making a decision on issuing import health standards for pork.
[49]Biosecurity Act 1993, s 16(a).
As a result of this recommendation, the Director-General agreed on the further recommendation of the chief technical officer that all the review panel’s “process-related recommendations” could be passed on to a general review and did not need to be accepted in relation to the management measures for imported pig meat. Some of the recommendations (especially as to monitoring and management of risks after importation) could only be carried out “if import health standards for pork take effect and trade commences” but “could be part of a package of risk management measures”.
In response to the recommendations for further data collection and analysis, the chief technical officer advised that there were a “wide range of possibilities, from carrying out no further work to beginning the import risk analysis process anew”. The chief technical director did not recommend starting again:
[The Ministry] could choose to revisit the entire process of carrying out a new import risk analysis and subsequent development of draft import health standards, with attendant public consultations. Indeed, [the Board] has already called for this to occur. Given that [the Ministry] has spent the past five years closely studying the risks associated with pork imports, it is not clear to me that a further round of analysis and consultation would justify the resources required. Moreover, the six-to-nine month timetable suggested by the Panel for [the Ministry] to respond to their report indicates this option was not what they intended.
The chief technical officer commented that the carrying out of further research and modelling and the updating of the existing risk analysis as recommended by the review panel could be time-consuming and have significant resource implications. One option was to issue the import health standards immediately on the basis that further investigation was unnecessary. The chief technical officer suggested that it was necessary for the Director-General to:
decide how much, if any, additional activity you want [the Ministry] to undertake before I provide you with final advice about the effective management of the risks associated with pork importation.
It is not stated explicitly but it necessarily emerges from the paper and its recommendations that in effect the Director-General decided that further work before he could issuing the import health standards following further recommendation of the chief technical officer was necessary only in two respects. First, a quantitative model of the risk of a PRRS incursion was to be developed by the Ministry, using any recently published relevant material, and then reviewed by independent specialists, with key stakeholders being given four weeks to comment on the design of the model. Secondly, an expert working group was to be set up to review input variables for the quantitative modelling, with sensitivity analysis being used where necessary for inputs in the quantitative modelling. The Director-General noted that the Ministry did not propose to investigate further the extent of trimming of raw pork to test the hypothesis on which the 3 kg cut was based. Rather, “a range of values and test assumptions” would be modelled through the expert working group. Stakeholders were to be invited to nominate experts to have input into both the modelling design and the input variables.
Updated information readily obtainable would be used, but further research work suggested in relation to PRRS by the review panel would be undertaken by the Ministry for the purposes of assessing “whether risk management measures specified in the provisional import health standards are effective”. The chief technical officer recommended that the Director-General note that the “output of the model, in predicted frequency of outbreaks, will largely inform my future recommendations to you on the effectiveness of measures to manage the risk of PRRS introduction”. That is to say, the model output was acknowledged to be critical to the chief technical officer’s recommendation under s 22(1).
The effect of this decision seems to me to be that the Director-General took the view that the outstanding question before introduction of the import health standards was modelling the risk of introduction of the virus. Two pieces of work only were required: finalising the structure of the model, and determining the input variables (including by sensitivity analysis in respect of the pork trimmings information gap). Other risks in managing the virus, should it be introduced, and the information needed to assess the risks and develop management strategies in that event (such as in relation to the spread of the virus if introduced or the effect on the virus of chilling pigmeat), would be the subject of ongoing work after introduction of the import health standard and, if necessary, lead to adjustment of standards. The criticisms the review panel had made of the Ministry’s processes were left to be picked up in a more general review of border controls.
The expert working group
Dr Neumann was nominated by the Board to participate in the expert working group comprising 11 members. Also in the group were two experts from EpiX Analytics LLC, Dr Zagmutt and Dr Groenendaal, appointed by the Ministry. The Ministry’s model, which it had continued to work on in adaptation of the Neumann model, was considered by the expert working group. It resulted in no agreement. Some of the experts considered the model was fundamentally flawed and required reconstruction. Internal Ministry documents of the time indicated that the Ministry was by then bracing itself for legal challenge.
Dr Neumann himself produced a revised model which represented household and retail food sector waste fed to small commercial and non-commercial herds. The retail waste stream had not before been modelled. Of particular significance was that it included discard of whole cuts of meat (from freezer or cold store failures or sell by dates on packaging). Instead of the Ministry’s mean prediction of one incursion per 4,635 years, this model predicted a median of 9.8 PRRS introductions per year, with a mean result of 16 introductions per year. This model was produced by Dr Neumann too late for the expert working group to consider it.
The expert working group reported to the Ministry on 29 October 2010. Its report was released at the beginning of November.
The Zagmutt-Groenendaal model produced by EpiX Anaytics
With the expert working group unable to come to an agreement, the Ministry asked Dr Zagmutt and Dr Groenendaal of EpiX Analytics (who had been members of the expert working group) to review the latest Neumann model. They produced a report critical of the model, which they said threw up some impossible results and contained at least one insupportable assumption. The Zagmutt-Groenendaal model, received by the Ministry on 14 December 2010, produced a much lower risk assessment than Dr Neumann’s. It predicted one outbreak on average every 1,227 years.
The Zagmutt-Groenendaal model produced by EpiX Analytics was not released by the Ministry. The Board, which had learned of the report and risk model when it was mentioned in an email, unsuccessfully sought disclosure of it in February 2011. Its request was declined by the Ministry on 24 March.
Before then, the chief technical officer had submitted recommendations to the Director-General for decisions to be taken under s 22A(3) of the Biosecurity Act and which also recommended that the Director-General issue under s 22 of the Act the final import health standards issued as provisional in April 2009, in reliance on the Zagmutt-Groenendaal model.
The decision of the Director-General dated 10 April 2011 was released on 13 April 2011. On the same date, the Ministry released the EpiX Analytics report dated 14 December 2010 containing the Zagmutt-Groenendaal model for risk adopted by the Director-General in his s 22A determination. On the same date, also, the Director-General issued the new import health standard for pig meat, dated 18 March 2011 which adopted in final form under s 22(1) the standard provisionally made in April 2009, in the form of the draft released for consultation in November 2007. The Ministry’s earlier reliance on qualitative analysis in setting the standards was said to have been supported by the quantitative analysis carried out by EpiX Analytics in the Zagmutt-Groenendaal model.
The 10 April 2011 decision of the Director-General
On 13 April the Director-General issued what was described as a “decision in accordance with Section 22A(3) of the Biosecurity Act 1993”. The report recited the history of the April 2009 provisional import health standards and the request from the Board which led to the report of the review panel in March 2010. The decision recites that, following the report of the review panel, “the (then) Director General agreed to a programme of work in response to the report” (referencing the chief technical officer’s advice) and that the “work programme has now been completed”:[50]
The development of the quantitative risk model has been a long process involving several levels of peer review, including through the EWG process, and various structural changes to incorporate recommendations. The model described in the report of the expert consultants, EpiX Analytics LLC (an independent consultancy specialised in quantitative risk analysis), is accepted by [the Ministry] as “fit for purpose” to support a decision on whether import health standards issued under the Biosecurity Act section 22 provide for effective management of biosecurity risk. The model supports the conclusions of [the Ministry]’s earlier qualitative risk analysis in that respect. The model reports a mean of 0.0038 PRRS virus primary introductions per year if the importation of pig meat in the proposed consumer-ready form were permitted. This model output can be considered to be equivalent to an average of 1,227 years between outbreaks, given the current conditions and parameters assumed in the model. The expert consultants EpiX Analytics have offered their view that the model provides a conservative estimate of risk, in that the selection of parameters for severable variables are likely to over-state the risk.
[50]Footnotes omitted.
The decision identifies the “matter in dispute” as one that “can be described at two levels”:
Overall, the matter in dispute is whether [the Ministry] has taken appropriate account of the available science in determining that the provisional import health standards provide for effective management of biosecurity risk, considering the legal obligations of Section 22(5) of the Biosecurity Act 1993.
In detail, the matters in dispute are effectively summarised as each of the individual matters in the terms of reference [the Ministry] established for the [review panel]. These are referred to in the terms of reference under the following headings:
1. Scope of the standards
2. Virus levels in imported meat
3. Impact of changes in volume of trade
4. Likelihood of infection
5. Knowledge of New Zealand industry, and consequent spread risk
6. Relevance of quantitative models
7. Overall assessment of risk
In answer to the “overall” matter in dispute, the Director-General decided, for reasons which are contained in “two technical briefings” which are annexed to the decision:
·The final import health standards, and [the Ministry]’s process to develop them (including the additional work undertaken in response to the [review panel]) has taken appropriate account of the available science, and will provide for effective management of biosecurity risks, considering the legal obligations under Section 22(5) of the Biosecurity Act.
·A minor amendment to the specification of the requirements in relation to pH-cured pork in the final import health standards is appropriate to take account of the published science on viral stability, while effectively managing biosecurity risk and remaining aligned with manufacturing norms, thereby meeting the principle of being least-trade restrictive in accordance with the [SPS Agreement].
Each of the “detailed matters in dispute” was then addressed in the remainder of the document. The decisions accorded with the approach taken in the 30 August paper. The recommendations directed to the processes followed by the Ministry were treated as not raising “specific matters requiring a decision in relation to the pork import health standards” and were referred to a more general review, the Border Change Programme, being conducted within the Ministry.
The review panel had recommended that more recent scientific knowledge bearing on meat infectiousness should be used to review the risk assessment it had earlier carried out. The Director-General’s response referred to the 30 August response to the review panel in which it had been pointed out that the risk assessment had not relied on diagnostic tests “but instead is based on direct estimation through observational studies”. The Ministry had not considered that technological developments in diagnostic testing would “necessarily assist in determining effective biosecurity measures to manage the risk posed by PRRS at this time”. It considered that infectiousness and virus concentration could be addressed in quantitative modelling (including sensitivity analysis), which could draw on the findings of other risk assessment bodies. Further empirical studies were unlikely to greatly add to existing knowledge and the time and cost in carrying them out were not justified.
The quantitative modelling on the risk of a PRRS incursion was to be the principal tool for addressing the concerns expressed by the review panel. It would draw on any relevant recently-published material and would “incorporate dose-response data highlighted by the [review panel]” and “investigate the effect of meat treatments such as pH and age curing on PRRS virus survival”. While studies to measure trimming and disposal of fresh pork by New Zealand consumers would not be carried out, modelling would use “a range of values and test assumptions through the expert working group”. The decision given by the Director-General on 13 April 2011 confirmed the approach indicated in the August response and concluded that the updated quantitative risk assessment conducted by EpiX Analytics had taken appropriate account of the limitations of the diagnostic tests in determining the proportion of imported pork likely to contain an infectious amount of the PRRS virus. The decision concluded that the EpiX risk assessment had taken appropriate account of the available science relevant to the likelihood of meat being infectious, the level or virus likely to be present in imported pig meat and the infectious dose of the PRRS virus. It further concluded that development of the EpiX risk assessment had treated uncertainty appropriately, and had used scenario and sensitivity analysis to look at the effect of different parameterisation options and the influence of variables and their parameters on the model output.
The first decision turned on the key variable “Contamination_P”. It had been drawn by EpiX Analytics from a 2004 study. The expert working group had concluded that the variable drawn from the same study used in the model based on Dr Neumann’s original work was wrong. The Director-General accepted that EpiX Analytics had employed a “more methodologically correct” means of using the data from the study and had taken into account “the limitations of the current scientific evidence used to derive the Contamination_P variable”.
The innovation is to combine the two sets of data as two beta distributions multiplied together, using the Magar and Larochelle (2004) empirical evidence rather than expert estimates. This avoids the need to specifically set a lower bound estimate, as required when the Pert distribution was used in both the original Neumann et al (2007) model and the revised [Ministry] model.
The second decision turned on the key variable in step four of the quantitative model, which arrived at “an estimate for the variable Scrap_Viral_Concentration”. The Director-General expressed satisfaction that the final model incorporated “a comprehensive and robust process to parameterise the key variable Peak_Virus_Concentration_At_Slaughter, following the [expert working group] process to identify potentially relevant publications and the [EpiX Analytics] application of meta-analysis techniques to this data”. EpiX Analytics had “reviewed the model’s use of variables to represent the chain of events after slaughter that affect eventual viral concentration in meat”, and had rejected any suggestion of double counting. The parameters selected had been based on a risk assessment reported by the European Food Safety Authority (EFSA)[51] and had used recently published studies on viral survival.
[51]The use of the EFSA report was a matter of dispute on the evidence: see below at [73].
The third decision was based on an innovation in the Ministry’s revision of the Neumann model, incorporating a 2005 study on infectious dose. The expert working group had agreed this was acceptable use of the data. The Director-General noted that “[a]n issue arises, however, with the extrapolation to low doses from the Hermann et al (2005) data which is important as the number of trials increases (trials being feeding events i.e. a scrap of uncooked imported pig meat being fed to a pig in New Zealand)”:
This became particularly relevant when other model changes were introduced as a result of the [expert working group]. The expert consultants EpiX Analytics demonstrate the effect of different model fitting options at doses below that which the Hermann et al (2005) study demonstrated did not achieve transmission. These different statistical methods result in curves that vary significantly in shape as the probability approaches zero. The experts conclude that the Logistic dose-response model reported by Hermann et al (2005) is not appropriate when making low dose extrapolations beyond the experimental data (as is required in this risk assessment model). They conclude that alternative models widely used in food-safety modelling are more appropriate, and have used a Beta-Poisson model in their report and final model.
The fourth decision was justified on the basis that EpiX Analytics had emphasised that “the overall model structure reflects the objective of estimating likelihood incorporating uncertainty around the mean, rather than the variability across the system”:
Throughout the model the expert consultants [EpiX Analytics] have checked to ensure parameters use the available data appropriately in this respect. Their reports to the [expert working group] and accompanying their final revised model describe the scenario and sensitivity analyses they have undertaken.
Section 22A does not specify the point in the process at which an independent review may be requested. There is some suggestion in the language of s 22A that a review by an independent panel can only take place after an import health standard has been issued or at least released in draft – s 22A(1) identifies the purpose of a review as being to consider whether, in developing an import health standard, there has been sufficient regard to the scientific evidence. However, we consider that the language is sufficiently broad to permit a person consulted about a risk analysis to request an independent review of it, given that risk analysis is a critical step in the development of an import health standard.
The question is whether the independent review process under s 22A could give rise to a further obligation to consult under s 22(6). Clearly where the review process accepted that sufficient regard had been taken of the scientific evidence, no further obligation to consult would be triggered. But what is the position where a review concludes that there has not been sufficient regard to the scientific evidence or does not reach any conclusion about that but does recommend further work, to take account of new data, for example?
The majority in the Court of Appeal considered the terms of s 22(6) and (7) to be decisive on this issue.[122] The chief technical officer’s obligation to consult arises before he or she makes a recommendation to the Director-General concerning the issuance or amendment of an import health standard. In the present case, the first import risk assessment and provisional import health standards were issued in mid-2001. The Ministry undertook further work and consultation before releasing its further import risk assessment for consultation on 25 July 2006. On 12 November 2007, the Ministry released the draft import health standards for consultation. On 7 April 2009, the Director-General issued his provisional import health standards, after which NZ Pork requested the appointment of a review panel. The Court regarded what followed after the review panel had reported as essentially a continuation of what had preceded it, so that no new obligation to consult arose. There was no recommendation from the chief technical officer about a new import risk assessment or new import health standard to trigger an obligation to consult.[123] Rather, the import health standards approved by the Director-General in April 2011 were materially the same as the provisional standards published in April 2009.
[122]New Zealand Pork Industry Board (CA), above n 64, at [87] per Harrison and Stevens JJ.
[123]At [90] per Harrison and Stevens JJ, agreeing with Williams J in New Zealand Pork Industry Board (HC), above n 68, at [233].
The majority’s judgment is not clear on the question whether there is a further obligation to consult where the scientific basis for a chief technical officer’s recommendation has changed substantially as a result of work undertaken following a review panel’s report, even though the recommendation itself remains the same. We consider that there would be an obligation to consult further in such circumstances. To take an extreme example, assume that a chief technical officer recommends import health standards on the basis of a risk analysis prepared by the Ministry. Following consultation on the risk analysis, the independent review process is invoked. The review panel concludes that the Ministry’s risk analysis has been so poorly carried out that the Ministry should discard it and undertake a new risk analysis, on a different basis and with expert assistance. The Ministry does that. On the basis of the new risk analysis, the chief technical officer makes the same recommendation to the Director-General as that made previously, without further consultation. In these circumstances, the consultation requirements in s 22(6) would not have been met, the new risk analysis being wholly different from that originally consulted upon.
This does not mean, however, that whenever a review panel recommends further work, there must necessarily be further consultation under s 22(6) after that work is completed. Work of a “tidy-up” nature is unlikely to trigger a further obligation to consult, for example, nor is the updating of existing data where the updated data is consistent with existing data and supports the conclusion already expressed. Ultimately, whether the obligation to consult again is triggered will depend upon the nature, extent and impact of the further work, the focus being on whether the work has led to a substantial change in the scientific basis for the chief technical officer’s recommendation. The fact that further consultation is required will not necessarily mean that there must be a further independent review if one is requested, given that the Director-General has some discretion in that respect.[124]
[124]The Biosecurity (Process for Establishing Independent Review Panel) Notice 2008, above n 79, at cl 9.
In the present case, the Ministry did undertake further consultation following the independent review. It held a workshop with stakeholders to determine how it should proceed in light of the review panel’s report. More importantly, it established the expert working group to consider the result of that further work, namely the Ministry’s revised model. NZ Pork nominated Dr Neumann to that group, and he presented a reworking of the Ministry’s revised model, the Neumann EWG model, at the end of the expert working group process. This in turn led to the EpiX model.
In the Court of Appeal, the majority did not accept that the EpiX model was a new or fresh import risk assessment. Mr Cooke challenged that finding, relying in particular on Professor Morris’s affidavit evidence. Professor Morris expressed the opinion that the EpiX model was “a very substantial underestimate of the risk of a PRRS incursion in New Zealand under the new import health standards”. He said that the EpiX model contained several very influential parameters, three of which he used to illustrate what he considered to be its weaknesses:
(a)the amount of pork imported (from any source) – Import_P;
(b)the proportion of imported pork sold as consumer-ready cuts – Consumer Ready_P; and
(c)viral persistence after slaughter – Prop_After_Bleeding.
In his dissenting judgment in the Court of Appeal, White J placed particular weight on this material.[125]
[125]New Zealand Pork Industry Board (CA), above n 64, at [102] per White J.
Before addressing the three aspects identified by Professor Morris, we repeat two points. First, over the course of the development of the import health standards, four quantitative risk analyses were developed (in addition to the Ministry’s 2006 qualitative risk analysis) – the Neumann/Morris model in 2007, the Ministry’s revised model in 2010, the Neumann EWG model in 2010 and the EpiX model in 2010. Each of the 2010 models was a development of what preceded it. Second, as far as Drs Zagmutt and Groenendaal were concerned, the EpiX model did not make any fundamental changes to the structure or logic of the Neumann EWG model, so that it was not a new model. Rather, they saw it as correcting some mistakes and unjustified assumptions in the Neumann EWG model.
Import_P
In its revised model, the Ministry noted that in the year to June 2010, imported pork accounted for 43.7 per cent of the total supply of pork to the New Zealand domestic market. The Ministry said that its revised model accepted the 0.42 value used in the Neumann/Morris model and allowed for up to 50 per cent of consumed pork to originate from overseas to accommodate an assumption that the new import health standard would result in increased imports. Import_P was accordingly described in the Ministry’s revised model by “Uniform (0.42, 0.5)”.
In the Neumann EWG model, Dr Neumann utilised a new figure for the supply of imported pork to the New Zealand market, namely 60 per cent. This was based on usage estimates provided to the expert working group by Mr Glass, a pork processing expert employed by FreshPork New Zealand Ltd. Mr Glass considered that the 60 per cent figure would result from the volume and price competitiveness of imports. Accordingly, in Dr Neumann’s reworked model, Import_P was described by “Normal (.6, .05)”. It is apparent from its report that these estimates were discussed at some length by the members of the expert working group.
By way of background, among the issues commented on by the review panel was the likely impact of the introduction of the import health standards on the volume of pig meat imported into New Zealand and the volume and distribution of waste material (such as scraps and off-cuts) that would be produced. The review panel recommended that the Ministry consider collecting information on the potential changes in the volume and source of imported pork and associated waste. Mr Glass addressed these matters in the expert working group, providing estimates of the increased volume of imports that would result and the amount of waste material that would be generated. Dr Neumann utilised these estimates in the Neumann EWG model.
Drs Zagmutt and Groendaal commented on the use of “Normal (.6, .05)” for Import_P in the Neumann EWG model as follows:
[T]he mean of the Normal distribution was based on Mr Glass’s personal estimates, and the .05 standard deviation (standard error) was provided by Dr Neumann. No further references or explanation are given for the choice of parameters.
They said that this difference alone provided an explanation for the large differences between the Ministry’s revised model and the Neumann EWG model. Although they accepted other estimates provided by Mr Glass in the EpiX model (in particular, in relation to waste), they decided that they would use the existing parameters for Import_P rather than those suggested by Dr Neumann on the basis of Mr Glass’s estimates.
Drs Zagmutt and Groenendaal explained that they did not consider that Mr Glass had justified his 60 per cent estimate, whereas they considered that there was a legitimate basis for the figure adopted by the Ministry in its model. It is, of course, not the Court’s role to determine which of these two views is correct; and even if it were, it would be difficult to do so given that Import_P involves a prediction of future market behaviour following a change in regulatory settings.
What is important in this context is that the estimates provided by Mr Glass were discussed in the expert working group, and Mr Glass had the opportunity to explain and defend them. Ultimately no consensus was reached. But it is untenable to suggest that NZ Pork’s nominee on the expert working group, Dr Neumann, did not know what was proposed in relation to Import_P and did not have a fair opportunity to express his views. Equally, it is untenable to suggest that the views of Mr Glass and Dr Neumann were not considered in the expert working group process. Ultimately, the advice of Drs Zagmutt and Groenendaal was that the Glass estimate of Import_P should be rejected in favour of the lower figure based on the figure in the Neumann/Morris model and the Ministry’s revised model. The Director-General ultimately accepted that advice. But the adoption of an Import_P figure based on the figure used in earlier models did not render the EpiX model a new model.
Consumer Ready_P
Consumer Ready_P was described in the Ministry’s revised model by “Pert (0.0095,0.02645, 0.0434)” and in the Neumann EWG model by “Normal (0.2, 0.05)”. In discussing this difference, Drs Zagmutt and Groenendaal quoted Dr Neumann’s explanation as follows:
[The Ministry] has used a flawed procedure to conclude that less than 3% of the imported product would be utilised as consumer-ready, whereas [Mr] Glass concluded that it would be about 33%. To avoid unnecessary debate about the exact value which would be reached under the proposed import policy, I have used a figure of 20% in the results reported below.
Again, Drs Zagmutt and Groenendaal considered that insufficient support was provided for Dr Neumann’s revised Consumer Ready_P parameter. Given the sensitivity of the parameter, Drs Zagmutt and Groenendaal considered that the Ministry’s description of the parameter should be used as the Ministry had shown how it had been derived from existing data. Again, this parameter was discussed within the expert working group, although no consensus appears to have emerged.
Prop_After_Bleeding
In relation to Prop_After_Bleeding, the EpiX model used the same parameters as were used by Dr Neumann in the Neumann EWG model: “Uniform (0.002, 0.009)”. Drs Zagmutt and Groenendaal explained that they had some concern about the parameters, but having conducted a sensitivity analysis they concluded that they did not make a biologically significant difference to the model output. They therefore decided to keep the parameters in the Neumann EWG model. NZ Pork can have no valid complaint about this.
Conclusion concerning parameter changes
Like the majority of the Court of Appeal, then, we reject Mr Cooke’s submission that the EpiX model was a new import risk analysis which triggered a further obligation to consult. The EpiX model did not involve a substantially different approach from the Neumann EWG model, nor did it introduce substantially new data or other material. Rather, it involved some adjustments to individual parameters in the Neumann EWG model, parameters which were discussed at length in the expert working group, to which NZ Pork had nominated Dr Neumann.
In his dissenting judgment, White J considered, on the basis of Professor Morris’s affidavit, that there remained “a genuine and unresolved scientific issue as to the risks involved in the importation of raw pig meat …”.[126] It is true to say that there was a difference of opinion among the experts as to some of the parameters to be used in the risk analysis. That is hardly surprising. As we have said, risk assessments, even quantitative ones, inevitably contain subjective elements.[127] To the extent that they involve assumptions and predictions, they involve uncertainty – and with uncertainty comes scope for differences of view. But consultation does not have to lead to consensus. It requires the decision-maker to consider in good faith the views that consultees have expressed about what he or she proposes to do.
[126]New Zealand Pork Industry Board (CA), above n 64, at [103] per White J.
[127]At [136] above.
The parameters utilised in the EpiX model were discussed in the expert working group and the various arguments about them were aired. At some point in the process, a decision had to be made even though consensus was not forthcoming. In this connection, the observations of Drs Zagmutt and Groenendaal in their discussion of the EpiX model are pertinent:[128]
A typical situation occurring while performing quantitative risk assessments is that the more the analyst, decision makers, and stakeholders think about the risk and the limitations of the model, the more complexity and/or collection of further data the analysts are encouraged to pursue. However, at some point further collection of data or improvements to the modelling methods may not provide incremental information to make a biological difference that may affect the decision to be supported with the model. We have attempted to point out the parameters and data that could be further refined while also considering whether those refinements can make a significant difference in the model predictions.
From this work, it is our belief that the current model provides a conservative estimate of the risk of introduction of [PRRS] into New Zealand via the importation of fresh pork meat …, and therefore any further modelling work or collection of new evidence should only be considered if either the current risk is not acceptable for the decision maker, or if new evidence will show that in fact the model underestimates the risk. In our experience, the former is possible but the latter is less likely given all the time, auditing, and discussions that have revolved around this particular risk assessment model.
[128]Emphasis added.
The Director-General had all this material before him when he made his decision. He decided to accept the analysis in the EpiX model despite the different parameters used in the Neumann EWG model, which produced a different outcome. As the extract which we quoted from the Handbook indicates,[129] a decision-maker can bring some objectivity to an inherently subjective process by ensuring that the analysis is transparent:
All the information, data, assumptions, uncertainties, methods and results must be comprehensively documented and the discussion and conclusions supported by reasoned and logical discussion. The analysis should be fully referenced and subjected to peer review.
As we see it, this is what the Director-General was intending to achieve in the present case through the processes he implemented over a lengthy period. Against this background, we consider that the Director-General was entitled to make a decision at the time he did.
Sensitivity analysis
[129]See [136] above.
Finally, we address Mr Cooke’s submission that the combined effect of the changes to the parameters mentioned above was that the estimated frequency of PRRS outbreaks was much lower in the EpiX model than in the Neumann EWG model. Given the sensitivity of the parameters, Mr Cooke argued, further consultation on the EpiX model was required.
The members of the expert working group discussed sensitivity issues, as did Drs Zagmutt and Groenendaal. The impact of the changes to particular parameters was well understood. Again, we consider that the Director-General was entitled to make a decision on the basis of the material before him. There was no obligation to consult further on this account.
Conclusion
We summarise our conclusions as follows:
(a)The Director-General responded lawfully to the report of the review panel established under s 22A of the Act.
(b)The consultation obligations contained in s 22(6) of the Act were met.
Result
For these reasons, we dismiss the appeal. The appellant is to pay costs of $25,000 to the first and second respondents collectively, plus reasonable disbursements as fixed by the Registrar.
Solicitors:
Jonathan Kaye Law, Wellington for Appellant
Crown Law Office, Wellington for Respondents
Davys Burton, Rotorua for Intervener
7
1
1