The difficulties in construction of s 18B
152 Bearing in mind the mischief to which s 18B was directed, the use of the phrase "the insured shall not be disentitled to be indemnified" accords with the notion that the provision is designed to deal with circumstances whereby cover, to which the insured was otherwise entitled, might be lost by the operation of some clause of the policy.
153 Paragraph (a), however, contemplates a definition of cover of the circumstances in which the insurer is bound to indemnify the insured. Paragraphs (a) and (b) provide for that definition of cover excluding or limiting the liability of the insurer on the happening of particular events or on the existence of particular circumstances if the happening of those events or the existence of those circumstances is, in the view of the insurer, likely to increase the risk. Thus, in the words employed by Parliament, the exclusion or limitation of cover may be in the definition of cover.
154 Where these events or circumstances occur or exist, the insured is not "disentitled" to be indemnified, by reason only of such provisions if the loss for which the insured seeks indemnity was not caused or contributed to by the happening of the events or the existence of the circumstances.
155 The section operates (if the happening of the events or the existence of the circumstances did not causally contribute to the loss) to remove the effect of provisions of the policy and, thus, to allow the policy to operate otherwise according to its terms.
156 Thus, essential to the operation of the provision is the ascertainment as to whether there is exclusion or limitation of the liability of the insurer as contemplated by paras (a) and (b). If there is, and if that would provide a partial or whole answer to the claim, the provision will not operate to disentitle the insured (in whole or in part) if it can prove the matters in s 18B (2).
157 The section does not operate by imposing statutory liability on the insurer. Rather, it operates to remove the effect of certain provisions of the policy which were part of the definition of the obligation of the insurer to indemnify, in certain circumstances, thereby allowing the policy otherwise to operate, without the effect of such provisions.
158 The section does not employ concepts such as (dominant) "scope of cover" and (servient) "exclusion" or "limitation" clauses. It operates according to its terms in the manner I have described.
159 It is important also to recognise that this was a beneficial provision intended for the protection of insureds. That benefit was in respect of the curing of the mischief which I have identified. That does not mean, however, that a limitation should be read into the provision to confine it to the precise circumstances illustrated by Kolokythas or Bazouni. Nor is the Court engaged in construing the NSWLRC report, or the Second Reading Speech. They identify the purpose, object or mischief. The words drawn to fulfil that purpose or object should be construed to give the fullest relief which a fair meaning of the words allow: Bull v Attorney-General (NSW) [1913] HCA 60; 17 CLR 370 at 384. With these considerations in mind, it can be readily accepted that the question as to whether one finds a provision of a character described by s 18B(1)(a) and (b) is a matter of substance (to which form will be relevant) and not a matter of mere form.
160 The difficulty posed by this appeal and which was before the arbitrators can be expressed in a manner referable to the argument of the reinsurers: whether s 18B can operate to extend beyond its effect on so-called exclusion and limitation clauses leaving the so-called true scope of cover to operate and to have an effect on clauses that truly reflect the so-called scope of cover, thereby extending the intended substantive reach of the policy.
Was there strong evidence of error?
161 The question as to whether there was strong evidence of relevant legal error by the arbitrators should first be viewed from the perspective of the reasons of the primary judge. Was his analysis (though principally directed to manifest error) correct? In my view, it was not. The proper construction of s 18B is not without its difficulties. I have set out its pre-enactment history and its context and apparent purpose. Nevertheless, the imposition upon s 18B, or the reading into it, of a distinction between "scope of cover" and "limitation" or "exclusion" clauses is, arguably, not warranted.
162 Such a distinction based on the characterisation of the clauses in a policy does not find its source in the words used by Parliament. To the contrary, s 18B(1)(a) contemplates the scope of cover to be defined, in part, by an exclusion or limitation clause: "the circumstances in which the insurer is bound to indemnify … are so defined as to exclude or limit the liability …".
163 The absence of the distinction suggested by the reinsurers and accepted by the primary judge in the words used by Parliament is not surprising: it is one, if based only on the text of a policy, that promotes over-reliance upon form, rather than substance. For example, if an event or a circumstance is perceived to increase a risk, such can be contained in a limitation or exclusion clause with an otherwise ample insuring clause, or the insuring clause can be framed to exclude the risk. The matter is one of drafting. The point is well illustrated by the facts here. The first response of the reinsurers to what they saw as the inadequately expressed treaty wording by its omission to deal with policy reporting periods of up to three years, was to have an exclusion inserted into the treaty (in Article 5: Exclusions) excluding policies issued for periods longer than 36 months: see [47] above. They could equally have required the insuring clause to be amended.
164 If the distinction between "scope of cover" and "limitation" or "exclusion" clauses depended upon the form of the policy, the operation of the provision, avowedly for the benefit and protection of insureds would depend on the drafting policy and acuity of insurers.
165 Mr Pembroke SC disavowed such an approach based on form. He submitted that the matter was one of substance, not form. Properly understood, this was the approach of the primary judge. What was required in the application of the section was to characterise the policy and its terms to identify its true nature (the class of business) and its true scope or essence. Once that had been done, clauses limiting or removing matters from the cover could be identified. It was only to the latter provisions that s 18B spoke. Thus, here, Mr Pembroke submitted that it did not matter whether the question of policies for periods beyond 36 months was in the insuring clause or an exclusion; it went to the substance of the cover and was characterised as related to scope of cover to which s 18B did not speak.
166 Whilst it can be accepted that the notion of scope of cover and class of business are important working concepts in insurance, especially reinsurance, s 18B is concerned with all types of insurance and not merely (on the present hypothesis) reinsurance. It is strongly arguable that the drafter of s 18B and Parliament did not intend to introduce such an unexpressed distinction between scope of cover and limitation or exclusion clauses into s 18B. The just and reasonable operation of s 18B can be ensured by the application of the requirements under (b) (if the loss was not caused or contributed to by the event or circumstance) and by the overall test of reasonableness in the proviso.
167 Extreme examples were posited in argument and by the primary judge as consequences of not imposing the scope of cover / limitation or exclusion clause distinction on the proper construction of the section; for example, a fire policy might be called upon to respond to a professional indemnity risk. Such examples ignore the sensible or practical working of the text of the section. Professional indemnity risk is not covered by a fire policy, but this is so because it is a different risk described and not because the liability of the insurer is defined as to exclude or limit its liability by an event or circumstance seen by the insurer as likely to increase the risk.
168 Reliance on cases such as Burton Lommers and QBE v Commercial Union is not helpful. Those cases concerned s 18. The inutility of the distinction between scope of cover and the condition in relation to the cognate section in the IC Act, s 54 was made clear in a number of cases: East End Real Estate Pty Ltd v CE Heath Casualty & General Insurance Ltd (1991) 25 NSWLR 400 at 403-404; FAI General Insurance Co Ltd v Perry (1993) 30 NSWLR 89; Antico v Heath Fielding Australia Pty Limited [1997] HCA 35; 188 CLR 652; Greentree v FAI General Insurance Co Ltd (1998) 44 NSWLR 706 (esp 722); FAI General Insurance Co Ltd v Australian Hospital Care Pty Ltd (1999) 10 ANZ Insurance Cases ¶61-445 and [2001] HCA 38; 204 CLR 641.
169 I cannot agree with the primary judge that there is an inconsistency between [78]-[81] and [90]-[93] of the award reasons and that [90]-[93] are somehow wrong. Read together, these paragraphs reflect the conclusion of the arbitrators that, properly understood in its context, the 1999 treaty contained a definition of the circumstances in which the reinsurers were bound to indemnify Gordian so as to exclude or limit the liability of the reinsurers on the existence of a circumstance that an underlying policy provided for a reporting period of more than three years, the liability being so defined, because that particular circumstance was in the reinsurers' view likely to increase the risk of loss occurring rather than if the policy had provided a reporting period of three years or less.
170 That approach that was employed by the arbitrators involves an entirely arguable construction and application of s 18B. It does not involve the requirement of any antecedent question as to whether the restriction on cover (to use a neutral term) was properly characterised as referable to "scope of cover" to which s 18B did not speak or "exclusion" or "limitation" of cover to which s 18B did speak.
171 The broad construction placed on s 18B by the arbitrators is supported in particular by at least two factors: first, it employs, without gloss, the words chosen by Parliament without the need either to rely on the form of restriction chosen by the insurer or to answer a question of characterisation not expressly found in the section; and, secondly, it gives the most complete remedy that such language permits in aid of a beneficial section.
172 For these reasons, I am of the view that the primary judge was wrong to conclude that there was strong evidence (in the sense of a strong prima facie case) that the arbitrators misunderstood or misconstrued s 18B.
173 If I be wrong about that, and there was a strong prima facie view that the arbitrators were wrong about their construction of s 18B, leave would have been correctly granted by the primary judge. I would accept that the proper construction and interpretation of the Insurance Act, s 18B was the determination of a question that may add or may be likely to add substantially to the certainty of commercial law. The contrary was submitted by Gordian, but in my view judicial exposition of the meaning of a provision affecting the operation of insurance markets in Australia plainly satisfies that criterion. This is so even if much of the work of the Insurance Act is undertaken by Commonwealth law: the IC Act, the Marine Insurance Act 1909 (Cth) and the Life Insurance At 1995 (Cth).
174 It would then (and only then) be necessary to deal with the question of the proper construction of s 18B on the appeal and the appropriate remedial consequences based on that construction. I will come to this in due course.