Introduction
 
The starting point for any analysis of the application of the Human Rights Act 1993 (“Act”) must always be to remind yourself of its broad purposes.  Analysis of the provisions of the Act requires, in the words of Thorp J (Coburn v Human Rights Commission [1994] 3 NZLR 323), “an appropriate regard for the substantial body authority both in New Zealand and abroad, as to the special character of Human Rights Legislation and the need to afford it a fair, large, and liberal interpretation, rather than a literal or technical one”.

In depth analysis of the wording of particular sections within the Act, and any attempt to limit the scope of the Act’s application, is coloured by this judicial approach.  The analysis which follows is no exception. It is important to take a step back when considering any potentially discriminatory practice and ask yourself “does this practice conflict with the broad purposes of the Act?” If the answer to this question is yes, and there is no clear exception permitting the proposed practice, there is likely to be a problem.

The first aspect that I want to cover in this article is to consider the limits upon the concept of an insurance “facility” - how narrowly can a product be defined?  I then move on to look at possible limitations in the scope of Section 44 of the Act, the main operative provision of the legislation that determines whether or not a potential problem exists in the supply of facilities to the public.  Finally, I will look at the insurance exception - Section 48, and explore some of the limits on the ability of an insurer to rely upon reasonable actuarial or statistical grounds to sustain any discriminatory practices. The article focuses on life and disability insurance, being the area in which most of the unresolved issues reside.

Insurance Facilities (how narrowly can a product be defined?)

Section 44 of the Act is directed at unlawful practices of, amongst other things, persons who supply goods, services or facilities to the public or to any section of the public.  In this context, the term “facilities” expressly  includes facilities by way of insurance (Section 44(2)).  The unlawful practices in question are to:
- refuse or fail on demand to provide any other person with those facilities; or
- treat any other person less favourably in connection with the provision of those facilities than would otherwise be the case,
- by reason of any of the listed prohibited grounds of discrimination. 

Critical to this is the question of what will constitute an insurance “facility”.  Section One of the Guidelines provides some guidance in this area.  In particular:  “The Act does not compel insurers to provide products which they do not usually supply.  If a company does not routinely offer a product then the Act does not require that it be provided simply because an individual requests it” (page 6).

The facilities provided by an insurer to the public will generally comprise one or more separate products.  An argument that you are not in the business of providing, for example, income protection insurance to those over 75 does not hold up (R v M 10/4/96 Complaints Division C154/95).  Although the particular product offered to the public might only envisage persons aged 75 and under, the insurance “facilities” that are provided to the public are “income protection”.  In R v M there was a clear refusal to provide insurance facilities to a section of the public.  Restricting the availability of facilities by reference to the offerees’ individual characteristics, therefore, is not an appropriate approach.

The alternative approach to defining the scope of insurance facilities is to look at the extent of the risks covered.  Insurance facilities involving life insurance cover are clearly distinct from insurance facilities for home and contents.  Where things get more controversial is where an insurer limits its activities to covering particular risks within a general type of insurance.  For example, a health insurer might provide cover in the event of physical injury or illness, but limit the facilities it provides to the public by having no product available that will cover psychiatric afflictions. Is it open for the insurer to limit the facilities it offers to this extent?

Let’s say, Ms A takes out an Income Protection Insurance Policy with insurer B.  The policy does not provide any cover for psychiatric disability and no income protection product supplied by B includes such cover.  A is then totally and permanently disabled by a nervous disorder.  Can A claim unlawful discrimination?  From A’s point of view she is now in a position where the need for the cover provided by B’s Policy has arisen.  However, in terms of the policy, the manner in which that need has arisen excludes her from the cover.

Limiting insurance facilities offered in this manner can be likened to excluding particular risks from the products offered.  In terms of Section Two  B4 of the Guidelines, blanket exclusions that apply to all applicants for insurance, regardless of known characteristics at the time a policy is taken out, are not “discriminatory” as all applicants are treated the same.    In other words the same “facility” is provided to all.   Is such an analysis consistent with the need to apply a “fair, wide and liberal” interpretation to the Act? 

The extent to which exclusion clauses might be regarded as legitimate was the subject of a landmark Supreme Court of Canada decision Battlefords and District Co-operative Limited v Gibbs [1996] 3 R.C.S. 566.  In that case, an employee was unable to carry out the duties of her occupation due to a mental disorder and became eligible for benefits under an insurance policy. A term in that policy, however,  was that in the event of mental disability cover terminated after two years unless the person was institutionalised.  The employee in question was successfully able to cry “foul”!  In the words of Sopinka J:

“A contract that explicitly provides for distinctions on prohibited grounds, albeit distinctions that only potentially occur in the future, is contrary to the objects of human rights legislation”.     

At first glance, the Battlefords decision seems hard to reconcile with  Section Two B4 of the Guidelines, permitting blanket exclusion clauses.  However, there are two principal distinguishing features between the approaches: 

First, in Battlefords there was a partial exclusion only - psychiatric disability was not excluded altogether but rather a limited form of cover was allowed for that particular type of disability.  An implication of this for insurers in New Zealand is that attempts at coming half way to meeting particular risks that would not otherwise be covered may well expose an insurer to a claim of discrimination, where a full exclusion may not.  By limiting the level of cover provided for a particular risk, it is no longer a blanket exclusion that is being applied, but a restricted form of cover.  The two are not necessarily subject to the same legal analysis.  One limits the boundaries of the facility provided, the other limits the extent of cover provided within the boundaries of that facility. 

Second, and more fundamentally, the Battlefords decision centres on employment discrimination.  It was the terms and conditions of employment that were primarily at issue, as opposed to the insurance product itself - i.e. the Saskatchewan equivalent to our Section 22, as opposed to Section 44. Offering terms of employment that involve those suffering from mental disabilities receiving less than those suffering from physical disabilities is always going to be problematic.  In a public provision of facilities, it is the refusal to provide those facilities, and different treatment in connection with the provision of those facilities, that is at issue.

The difference?  The question of employment discrimination, based on the statutory formula, is a continuing one that must be satisfied throughout the period of employment.  When public provision of facilities is at issue,  the critical time for assessment of discrimination is the date of provision of the insurance facilities in question.  The Guidelines proceed on the basis that provision occurs at the date on which the contract of insurance is entered into.  (Note however that the questions of when “provision” of insurance facilities actually occurs has yet to be tested in a New Zealand Court.  It is arguable, at least, that such provision is in fact continuous, occurring throughout the term of the contract).

Although sanctioned by the Guidelines, the legitimacy of the use of exclusion clauses is an issue that cannot be regarded as settled, based on public policy grounds, even if legislative reform were required to alter the current position.  If all insurers were to impose blanket exclusions in respect of particular ailments, the effect of the practice would be to place any person who may come to suffer from the particular ailment at a clear disadvantage with others in society, regardless of actuarial or statistical justification, with potential to perpetuate society prejudices.

Looking at the objects of the Act, it is apparent that insurers should rely upon the ongoing effectiveness and availability of blanket exclusions with a degree of caution.  The Guidelines themselves are not necessarily definitive on the question of blanket exclusions.  The explanatory text to the relevant Guideline itself contains cautionary wording.  Amongst other things, the explanatory text notes that discrimination will not occur “until the excluded condition or event occurs”, and urges caution in the design of products having regard to the need for which the product is provided.  Blanket exclusions should be reasonably applied.

Refusal or failure on demand - Section 44

Let’s go back to the hypothetical example used above.  Insurer B has a blanket exclusion for psychiatric disability in all of its income protection products.  Is it open to A to demand that a product be supplied that will provide cover for psychiatric disability, relying on the terms of Section 44(1)(a) of the Act? 

On the basis that B’s use of a blanket exclusion in its policies is legitimate in accordance with the Guidelines and the analysis above, the answer would appear to be no.  B’s argument is that its insurance facilities do not include cover for psychiatric disability.  It is not open to potential insureds to take on product design and dictate the facilities an insurer will cover.

If A were already suffering from a psychiatric disability, Section 44(1)(a) would ensure that B could not refuse to provide her with its income protection facility.  As with every other applicant for insurance, however, that facility need not include cover for her actual disability (which could also have been excluded as a “pre-existing” condition).  What B remains able to do, if appropriate actuarial or statistical support exists, is to place a loading on A on the basis that she is a sub-standard risk.  In terms of product design and marketing, this may push A into a different product made available by B as part of its income protection facilities, but this is not unlawful.  Refusal to provide a particular insurance product is not the same as refusal to provide the insurance facility in question.

One feature of Section 44(1)(a) may be worth exploring at this point.  This is the fact that the section focuses upon refusal or failure on demand to provide the relevant facilities.  It does not extend to failing to offer particular facilities.  This would seem to support the practice of making targeted offers of new products to “desirable” risk categories. 

For example, insurer B may have an extensive data base of health and other information on individuals who have taken out a particular product.  By going through that data base and picking out “desirable” risks for the offer of a different product, has a breach of the Act occurred?

On a strict analysis section 44(1), the answer must be no.  Although there has been a targeted offer of a particular product based on discriminatory grounds, there has been no refusal or failure on demand to provide the product in question.  Insureds under the original facility who missed out on the offer of the new facility have not necessarily been treated any less favourably than anyone else in the provision of the original insurance facility, provided the two insurance facilities in question are clearly different. 

The above practice runs a fine line. The manner in which any targeted offer is put together would need to be handled with some care, to ensure that there is no actual discrimination, and other provisions of the Act such as those relating to advertising are not infringed.  Further, any person falling outside of the original criteria who asks for cover under the new insurance facility must not be refused.  Upon demand being made by such a person (assuming remaining non-discriminatory criteria for the particular offer are met), the insurance facility in question must be provided.  At this point, however, the insurer would be able to have regard to Section 48 of the Act, and treat the individual differently by reference to actuarial or statistical data upon which it is reasonable to rely relating to life expectancy, accidents or sickness.

The main area in which the practice of targeted marketing may come into difficulty is where the new insurance product offered might properly be regarded as simply constituting an addition or variation to the original insurance facility.  In other words, it is part of the same facility.  In that situation, failing to offer the enhancement to the entire data base in question by reason of any of the prohibited grounds of discrimination will result in less favourable treatment of a person who does not receive the offer when compared with others.

 Section 48

No discussion of human rights in the context of insurance would be complete without some analysis of Section 48 - the exception in relation to insurance.

Section 48 is quite limited in its application.  It relates only to the offer or provision of policies on different terms or conditions for each sex or for persons under a disability or for persons of different ages (note that unlike Section 44(1), Section 48 specifically includes offers of policies within its ambit).  Any such different treatment will only be permitted if requisite data upon which it is reasonable to rely relating to life expectancy, accidents or sickness exists.  In respect of persons with a disability only, reputable medical or actuarial advice or opinion upon which it is reasonable to rely, where no actual data is available, will also support different treatment.  In either case, the different treatment must be reasonable having regard to the applicability of the data or advice and other relevant factors to the particular circumstances.

The first issue (which has been dealt with to a certain extent within the Guidelines) relates to the question of how insurers are expected to deal with policies where no data is available for the particular sex in question (bearing in mind the extended definition of sex within the Act), or for persons of particular ages.  Guideline B6 in Section Two contemplates passing on the costs of researching one off, deeply sub-standard risks to the individual in question.  What the Guideline recognises is that, in the absence of existing actuarial or statistical data upon which it is reasonable to rely, it is open to insurers to research one-off risks to come up with a reasonable level of data.  If insufficient data is gathered upon which it would be reasonable to rely, the particular risk must be covered on the same terms and conditions as a standard risk.

One aspect of Section 48 is particularly anomalous. This is the fact that it restricts the ability to rely upon actuarial or statistical data to situations where the data relates to sex, disabilities or age.  The fact that persons of a particular race or ethnicity may pose a higher than standard risk, according to reasonably reliable actuarial or statistical data, it is not a factor which can lawfully have any bearing on an insurer’s risk loading.  The fact that a relative of a person may suffer from a particular hereditary ailment, meaning that statistically there is a higher than normal risk of the person in question also coming down with that ailment, is not strictly a factor to which regard may be had in pricing policies.  Different treatment on this basis technically amounts to family status discrimination, which is outside of the Section 48 exception.

This restriction on the ability of insurers to accurately reflect the risks they are covering may have one of two effects.  The insurer may ignore the additional risk posed, which may have an impact on the commercial viability of its business.  Alternatively, premiums may be increased across the board to cover any additional risk the insurer is unable to directly pass on to the insureds in question.  Effectively, those posing a standard risk for the insurer end up subsidising those who pose an additional risk which the insurer is unable to address through direct premium loading or other different treatment.

Various techniques have been developed by insurers to overcome these difficulties and enable specific risks to be addressed.  These techniques tend to involve treatment of the issue as indirect discrimination, or treatment of the individual as if he or she were disabled so that reliance may be placed on actuarial or statistical data.  The difficulty with such practices is that they tend to come up against the need to afford the Act a fair, wide and liberal interpretation in order to promote its objects.  Reliance upon semantic niceties is inconsistent with this approach.

Conclusion

The Insurance Guidelines when originally released have gone along way towards providing some degree of certainty in the application of the Act to the insurance industry.  However, some unchartered territory remains, and a number of the boundaries established are not set in stone.  The question of what constitutes a particular insurance facility is one which requires a more definitive answer.  The bounds of the “fair, wide and liberal” approach to human rights legislation is also yet to be fully explored in the New Zealand context and this may impact upon marketing practices.  Lastly, the extent of the Section 48 exception cries out for some reform to ensure a fair allocation of the risk and costs of providing insurance cover.  A number of other issues faced by the industry, such as those relating to genetic testing, also remain unsettled.

     Whether these issues will be addressed in the years ahead remains to be seen.  The original Guidelines issued in 1997 may have been a significant, positive step towards providing certainty in respect of the impact of the Act on the insurance industry but they are unlikely to prove to be the final step.  

      Author: David Ireland, Partner, November 2006
For more information contact: Kensington Swan Lawyers
Posted: 31 July 2008