Child Poverty Action Group Inc v Attorney-General [2013] NZCA 402; [2013] 3 NZLR 729; (2013) 9 HRNZ 742 (30 August 2013)

	Home \| Databases \| WorldLII \| Search \| Feedback Court of Appeal of New Zealand

BETWEEN	CHILD POVERTY ACTION GROUP INCORPORATED (CPAG) Appellant
AND	THE ATTORNEY-GENERAL Respondent

Hearing:	28 and 29 May 2013
Court:	Ellen France, Randerson and Stevens JJ
Counsel:	F M Joychild QC and J M Ryan for Appellant C R Gwyn, J Foster and C I J Fleming for Respondent R M Hesketh and S A Bell for the Human Rights Commission as Intervener
_Judgment:	_{30 August 2013 at 11.30 am}

(i) First question: Did the High Court correctly state and apply the test for a breach of s 19 of the New Zealand Bill of Rights Act

1990? Answer: No. All beneficiaries were the subject of prima facie discrimination because of the off-benefit rule.

(ii) Second question: Did the High Court correctly state and apply the test for s 5 of the New Zealand Bill of Rights Act 1990? Answer: Yes. The off-benefit rule is a justified limit on the right to freedom from discrimination on the ground of employment status.

CHILD POVERTY ACTION GROUP INC (CPAG) V THE ATTORNEY-GENERAL CA457/2012 [2013] NZCA

Para No Introduction [1] The context of the claim [8] The genesis of the Working for Families package [13] The Working for Families package [19] Family support [24] The in-work tax credit and its predecessor, the child tax credit [26] The other tax credits [30] The abatement regime [33] Other aspects of the package [35] The relevant statutory provisions [36] Is the in-work tax credit discriminatory? [43] Differential treatment as between groups in comparable situations? [44] Our assessment [47]
A material disadvantage? [67]

Discussion [72] Is the in-work tax credit a justified limit? [76] The approach to s 5 [79] New Zealand [80] United Kingdom [83] Canada [85] Conclusions [91] Minimal impairment [94] The decisions below [95] The competing contentions [99] The relevant principles [102] Application of the principles to this case [104]

A proportional response? [130] Submissions [133] Analysis [136] Result [154]

[1] In 2004, the then Labour Government introduced a package of reforms for social assistance known as the Working for Families package. One component of that package is the in-work tax credit. The in-work tax credit supplements the income of those in work by up to $60 a week for families with three or fewer

children.¹The tax credit is available to any family that meets its statutory criteria for eligibility. These criteria exclude persons who are receiving an income-tested benefit.²This exclusion is called the “off-benefit” rule. Persons who do not meet the “full-time earner requirement” (working at least 20 hours a week for a single person and at least 30 hours for a couple) are also excluded.³

[2] The appellant, the Child Poverty Action Group Inc (CPAG), maintains that the in-work tax credit constitutes discrimination on the basis of employment status. In terms of s 21(1)(k) of the Human Rights Act 1993, it is unlawful to discriminate against persons on the basis of their employment status which includes the fact that the person is in receipt of a social security benefit. CPAG made a complaint about the in-work tax credit and was granted legal representation by the Office of Human Rights Proceedings to pursue its case in the Human Rights Review Tribunal.

[3] The Tribunal concluded that the eligibility rules for the in-work tax credit were prima facie discriminatory on the grounds of employment status. However, the Tribunal also found that this was a justified limit under s 5 of the New Zealand Bill of Rights Act 1990 (the Bill of Rights).⁴Section 5 of the Bill of Rights provides that rights may be limited only to the extent reasonably necessary in a free and democratic society.

[4] CPAG appealed unsuccessfully against the Tribunal decision to the High Court.⁵The High Court found that the off-benefit rule was prima facie discriminatory but only in respect of the small (1,267) group of persons who, while on a benefit, would meet the full-time earner criterion for eligibility for the in-work tax credit. The High Court concluded that the off-benefit rule was however a

¹For families with four or more children, an additional $15 per week is paid for each subsequent child. The evidence was that 92 per cent of families in New Zealand have three or fewer children.

⁴Child Poverty Action Group Inc v Attorney-General NZHRRT decision 31/08, 16 December

⁵Child Poverty Action Group Inc v Attorney-General HC Wellington CIV-2009-404-273,

25 October 2011 [HC judgment], Dobson J sitting with Ms J Grant and Ms S Ineson.

[5] After the High Court declined to grant leave to appeal,⁶this Court gave leave to appeal on two questions of law, namely, whether the High Court correctly stated and applied the test for a breach of s 19 and for s 5 of the Bill of Rights.⁷The Court said that the question about s 19 was to include the issues raised by the respondent, the Attorney-General, by way of cross-appeal, that is, whether the High Court erred in applying s 19 in holding that:

(i) those who are ineligible for the in-work tax credit on the basis of only s MD 8(a) of the Income Tax Act [2007] are the subject of prima facie discrimination; and

(ii) those who are ineligible for the in-work tax credit on the basis of both s MD 8(a) and s MD 9 of the Income Tax Act [2007] are not the subject of prima facie discrimination.

[6] We note that the Human Rights Commission gave notice of its intention to appear and be heard on this appeal as it is entitled to do under s 92H of the Human Rights Act.

[7] Before answering the questions of law, we first discuss the context of the appellant’s claim, the features of the Working for Families package, and the relevant statutory provisions.

[8] CPAG is an incorporated society formed in 1994 to advocate for a “better informed social policy to support New Zealand children” particularly those living in poverty.⁸CPAG undertakes research, publishes information and is a lobbyist for policy change. Its management committee includes Dr Susan St John, an Associate Professor in the Economics Department at Auckland University and Professor Innes Asher, Head of Paediatrics at Auckland University School of Medicine both of whom gave evidence before the Tribunal as did CPAG’s director,

2002 challenging forms of State assistance to families with children that are unavailable to families whose parents are in receipt of benefits on the basis of discrimination.

[10] As the submissions for the respondent record, State social assistance in New Zealand is provided in a number of ways. The evidence was that tax credits, like the in-work tax credit in issue, have become an increasingly common means both here and overseas of providing social assistance.⁹The in-work tax credit and other tax credits are delivered under the Income Tax Act 2007. The other principal forms of income support are provided via the Ministry of Social Development under the Social Security Act 1964 and under the New Zealand Superannuation and

[11] Income support is seen as comprising three tiers. The first, or main, tier includes the domestic purposes and unemployment benefits. These benefits are intended to provide for basic living costs and are subject to income tax.¹⁰

The domestic purposes benefit is the most significant in this case because the majority of children in single parent families reliant on a benefit are in households receiving the domestic purposes benefit. By contrast, relatively small numbers of couples in receipt of the unemployment benefit have children.¹¹Second tier assistance is directed towards persons in particular situations and/or for specific ongoing costs such as accommodation, disability, and child care. Third tier benefits are income- and asset-tested and provided generally to assist in times of hardship. The temporary additional support benefit introduced in place of the special benefit as

[12] We turn then to the development and features of the Working for Families package.

⁹In Humphreys v Revenue and Customs Commissioners [2012] UKSC 18, [2012] 1 WLR 1545 at [30] Baroness Hale noted that the concept of “integrating the tax and social security systems, ... to smooth the transition from benefit to work and reduce the employment trap, has been attractive to policy makers for some time”.

¹⁰New Zealand Superannuation and Veteran’s pensions have rates that are before the deduction of tax. “Income-tested” benefits like the Domestic Purposes and Unemployment benefits are defined in s YA 1 of the Income Tax Act.

¹¹The evidence of Donald Gray of the Ministry of Social Development was that as at the end of

events leading up to the package which we draw upon in the material that follows.¹²

[14] The starting point for these purposes is 1996. By that time the universal family benefit had been replaced by family support. There are four aspects to note of the position as at that time. First, family support was paid in the form of a per child/per week tax credit based on joint parental income. Importantly, it was available to all low income families regardless of the work status of the adult(s). Secondly, for the benefits in issue here, the domestic purposes benefit and the unemployment benefit, there was a child component, that is, these benefits were paid at a higher rate for adults raising children. The third point that we highlight in terms of the position as at 1996 is that there was a family tax credit paid to a relatively small number of working families to guarantee a certain minimum after-tax income level. Finally, the special benefit was payable to families irrespective of the adults’ work status as a form of hardship assistance.

[15] In 1996 the then National Government recognised that the family support rate needed adjustment because of inflation. In addition to increasing the rates by an additional $5 per child a distinction between working families and families on benefit income was introduced for the first time. An additional $15 per child per week was paid to families where the adults were not in receipt of a main benefit. This new payment was called the “independent family tax credit” but in this judgment we call this “family support”.

[16] In 1999 there was a change of government and the new Labour-led Government took office. As the Tribunal said, “[a]ssessment and reform of social assistance programmes were priorities”.¹³In 2000, the relevant Minister commenced a programme of work to achieve a number of goals recognising that “the primary factor for most will be activity through paid employment”.¹⁴This work

_{programme led to a report to the Cabinet Social Policy and Health
Committee in}

September 2000 recommending draft goals for the social assistance system, a detailed work programme as well as the indicators against which the current system was to be assessed.

[17] Various papers followed as a result. As the Tribunal noted, an important output of the work programme was a report from the Minister of Social Services and Employment of June 2001 entitled Pathways to opportunity: Social assistance reform.¹⁵The report stated that under the proposed new approach, the focus would be on “providing opportunities and the financial encouragement to move off benefit into work and to stay in work (making work pay)”.¹⁶Six main areas of policy development were identified, including simplifying the system, “making work pay and investing in people”, “supporting families and children” and “tackling poverty and social exclusion”. Officials were directed to advance work accordingly.

[18] As the Tribunal records, there is “ample evidence” of the “considerable work” undertaken to realise the “Pathways” initiatives.¹⁷This work culminated in a Cabinet paper of 31 March 2004 which contained the various elements of the Working for Families package.¹⁸We will come back to some of the features of the policy development process later in this judgment.

[19] We preface our discussion of the main features of the package by noting, first, that while some of the changes resulted in an increase in assistance other changes reflected a decrease in support. For example, the new criteria for the temporary assistance allowance were less flexible than the largely discretionary special benefit that it replaced. However, in general terms, as noted in the Cabinet paper, the package represented the application of more money in total to this area. Further, the clear direction to officials was that overall there should be “no losers” as a result of the package. The evidence confirmed that in absolute dollar terms no one

_{was worse off. Secondly, the package, while introduced as part of the
2004 Budget,}

¹⁵Minister of Social Services and Employment Pathways to opportunity: Social assistance reform

¹⁸Cabinet Paper “Future Directions: Working for Families” (31 March 2004).

1 April 2008. The total cost of the package was estimated at $1.14 billion for the first year following full implementation. However, and this is the final general point we make, in 2005 before the legislation enacting the changes was in force, it became apparent more money was available to the Government. The decision was made to alter the family credit abatement rate and threshold which meant families earning higher incomes were able to receive family assistance tax credits including both family support and the in-work tax credit.

[20] The objectives of the package are expressed in this way in the Cabinet paper:

make work pay by supporting families with dependent children, so that they are rewarded for their work effort. This involves better alignment of benefits and in-work support (including Family Income Assistance, Childcare Assistance and Accommodation Supplement) so that people are better off as a result of the work they do

ensure income adequacy, with a focus on low and middle income families with dependent children, to significantly address issues of poverty, especially child poverty. The package also addresses housing affordability problems by responding to the increased cost of private housing for low income people, and

achieve a social assistance system that supports people into work, by making sure that people get the assistance they are entitled to, when they should, and with delivery that supports people into employment. This involves steps to streamline the social assistance system so that it is easier for people to understand and access, and initiatives to improve take-up and enhance the effectiveness of delivery.¹⁹

[21] The intention was that the package as a whole would achieve these goals. In other words, the various aspects of the package were seen as working together to achieve the objectives. That said, the Cabinet paper identified particular initiatives as achieving one or other of the objectives. For example, one of the key initiatives linked to the objective of improving income adequacy as at 1 April 2005 was the

¹⁹Those objectives are in turn reflected in the Future Directions (Working for Families) Bill.

Section 1A of the Social Security Act 1964 sets out a number of purposes including e nabling provision of financial and other support to help people to support themselves and their dependents while not working and to help people find or retain paid work, and to “enable in certain circumstances the provision of financial support ... to help alleviate hardship”: s 1A(b).

[23] For these purposes, “family income assistance” relevantly covers family support (now, the family tax credit), the in-work tax credit and its predecessor, the child tax credit, the minimum family tax credit and the parental tax credit. We outline the relevant features of each in turn.

[24] Family support, as the respondent submits, is the primary method of government contribution to income adequacy for families. It is available regardless of income source.²⁰In terms of Working for Families, the key changes to family support were an increase in the rates (in April 2005 and 2007) and an increase in the income level that could be reached before the family income assistance abated. The

	Rate for first child		Rate for subsequent children
	0–15 yrs	16–18 yrs	0–12 yrs	13–15 yrs	16–18 yrs
Prior to WFF reforms	$47	$60	$32	$40	$60
From 1 April 2005	$72	$85	$47	$55	$75
From 1 April 2007	$82	$95	$57	$65	$85

[25] The legislation also provided for family support to be adjusted regularly on the basis of the Consumer Price Index.²¹The quid pro quo for these changes was the

_{removal, for the second and subsequent children, of the amount
previously paid as}

the “child component” of the main benefits. The evidence was that the increases to family support more than offset this change.

[26] The in-work tax credit is available to any family that meets its criteria for eligibility. Those criteria are provided for in ss MD 4–10 of the Income Tax Act. Those sections state that a person must:

(b) be the principal caregiver for a child who is financially dependent on them;²³

(d) not be receiving an income-tested benefit (the “off-benefit rule”);²⁵

(e) be a full-time earner²⁶(a full-time earner is defined in s MA 7 as a single person employed for at least 20 hours per week or a couple employed for at least 30 hours per week in total).

[27] We interpolate here that some benefits, relevantly the domestic purposes benefit, are available when the recipient is earning a low income. Unlike the unemployment benefit, a person on the domestic purposes benefit has no work hours limitation.²⁷Their benefit will, however, abate.²⁸Accordingly, a person on the domestic purposes benefit may meet all of the criteria for eligibility for the in-work

_{tax credit except for the requirement to be off-benefit. The number of
persons in this}

category at the time of the High Court judgment was 1,267. To put that figure in context, the evidence was that at the end of December 2007, there were some 88,700 persons in receipt of the domestic purposes benefit. Over 17,800 of these declared they were working and 1,538 declared earnings of more than $300 per week.

[29] The child tax credit, introduced in 1996, can be seen as the predecessor to the in-work tax credit. It is now available only to those persons ineligible for the in-work tax credit who were eligible for this credit immediately before the in-work tax credit came into force. The evidence was that, as at the year ended 31 March

2007, the child tax credit was paid to 22,000 families at a cost of $25 million.

[30] The minimum family tax credit provides a minimum after-tax income to working families with dependent children.²⁹Its aim is to help to see that families are better off in work than on a benefit. This goal is achieved by topping up net non-benefit income to an amount set annually. At the time of the hearing before the Tribunal, the threshold was $18,044 annually. This tax credit has a full-time earner requirement like that of the in-work tax credit and an off-benefit rule. There was evidence that, because this is a top-up, for every extra dollar earned (net) under the threshold, a family receives $1 less of the credit.

[31] The evidence of Michael Nutsford from the Inland Revenue Department was that for the year ended 31 March 2007, this tax credit was paid to 2,800 families at a cost of $8 million.

[32] Sections MD 11–12 deal with the parental tax credit which is available to families not in receipt of paid parental leave for 56 days after the birth of a child. The evidence showed that for the year ended 31 March 2007, the parental tax credit

[33] Section MD 1 deals with abatement for the in-work tax credit, family support (the family tax credit), parental tax credit and child tax credit. They abate according to this formula:

Family tax credit + (in-work tax credit or child tax credit) + parental tax credit – credit abatement.

[34] The family tax credit abates first and then the in-work tax credit. As the parties submit, this means the in-work tax credit is available to some families on a relatively high income. We annex, as Annexure A, the schedule showing the abatement at the time of the Tribunal hearing.³⁰

[35] In terms of the other aspects of the Working for Families package, there were increases in the amount of childcare assistance available and increases to accommodation supplements. The special benefit was replaced by the new temporary assistance allowance. Again, payment for existing recipients of the special benefit was grandfathered for a period. There were also changes to invalids’ benefits and consequential changes to other social assistance programmes.

[36] The provisions relating to discrimination are found in the Bill of Rights and in the Human Rights Act.

[37] Section 19(1) of the Bill of Rights protects the right to freedom from discrimination on prohibited grounds.³¹The prohibited grounds are those set out in the Human Rights Act. The prohibited grounds include:³²

³¹Affirmative action measures do not constitute discrimination: Human Rights Act 1993, s 19(2).

(ii) being a recipient of a benefit under the Social Security Act 1964 or an entitlement under the Accident Compensation Act 2001:

[38] We interpolate here that employment status became a prohibited ground of discrimination on the enactment of the Human Rights Act.³³The report of the Department of Justice to the select committee considering the Human Rights Bill said the Department was unaware of any overseas legislation including this ground but noted that the Human Rights Commission’s “experience is that discrimination on this ground occurs here”.³⁴The Department’s report also noted that initially “beneficiary status” was to be a separate ground but because of a concern about the number of grounds, this ground was subsumed into the employment status ground.

[39] The rights in the Bill of Rights may be limited only in the manner provided for in s 5. Section 5 states that these rights:

[40] The Human Rights Act applied immediately to the private sector but its application to the Government was initially limited.³⁵Part 1A dealing with discrimination by government was introduced via the Human Rights Amendment Act 2001. The effect of the Human Rights Act is that a declaration may be made that a statutory provision is in breach of pt 1A of the Act if the provision is inconsistent with s 19 of the Bill of Rights. For these purposes, a statutory provision is

³³The Human Rights Act brought together and revised the Race Relations Act 1971 and the Human Rights Commission Act 1977. It appears that at the time of the 1977 Act, the approach was to start with a less ambitious measure in terms of the number and scope of the grounds of prohibited discrimination: (7 July 1977) 411 NZPD 1245 and (23 August 1977) 413 NZPD

³⁴Department of Justice Human Rights Bill – Report of the Department of Justice (28 May 1993) at 16. There are some broadly equivalent provisions in some of the provincial human rights legislation in Canada: for example, Human Rights Code SS 1979 c S-24, s 2(1); Human Rights Act RSNS 1989 c 214, s 5(1) (refers to discrimination on the basis of source of income); Human Rights Act RSA 2000 c A-25.5, preamble (source of income); Human Rights Code CCSM 1987

c H-175, s 9; Human Rights Act RSPEI 1988 c H-12, s 1(1); Human Rights Code RSBC 1996

³⁵It was envisaged that there would be an examination of the compliance of acts and regulations with the Act prior to 31 December 1998. This became known as the “Consistency 2000” project, see ss 151 and 152 of the 1993 Act.

(b) is not, under section 5 of the [Bill of Rights], a justified limitation on that right.

[41] We add that New Zealand also has obligations relating to the family under the International Covenant on Economic, Social and Cultural Rights and under the United Nations Convention on the Rights of the Child.³⁷

[43] There is no issue that the correct test to be applied in the s 19 analysis is the two-stage test adopted by this Court in Ministry of Health v Atkinson.³⁸The first step is to ask “whether there is differential treatment or effects as between persons or groups in analogous or comparable situations on the basis of a prohibited ground of discrimination”.³⁹The second question is whether the treatment “viewed in context,

... imposes a material disadvantage on the ... group differentiated against”.⁴⁰

[44] The issue under this heading is that raised by the Crown’s cross-appeal. It affects the choice of comparator. The High Court said that in determining whether there was differential treatment the appropriate comparison was between those whose income from wages is for sufficient hours to entitle them to the in-work tax credit and those excluded because of the off-benefit rule, irrespective of their

15; United Nations Convention on the Rights of the Child 1577 UNTS 3 (opened for signature

_{20 November 1989, entered into force 2 September 1990), arts 2, 3(2),
6(2), 26 and 27.}

[45] The Crown says that in analysing the appropriate comparator, the High Court should have taken into account all of the eligibility criteria for the in-work tax credit. If the Court had taken that approach, it would be apparent that the difference in treatment was not as a result of employment status but because those represented by the appellant group do not meet the full-time earner requirement. The latter requirement is not challenged by CPAG. The further submission is that the Court should have considered the purpose of the in-work tax credit, that is, to encourage people into work. That would be another reason for concluding that the different treatment was not as a result of status as a beneficiary.

[46] CPAG supports the choice of comparator adopted by the High Court and the Tribunal. CPAG further says that the existence of other statutory criteria which beneficiaries cannot meet is not fatal to its claim. That is especially so where the work hours criterion is inextricably linked to or is a proxy for employment status. CPAG argues that the Crown approach would require the adoption of the “mirror image” comparator approach, that is, the selection of a comparator in exactly the same circumstances as the claimant group except for the prohibited factor. That approach is now rejected by the authorities. Alternatively, CPAG contends that employment status is at least a material ingredient of the difference in treatment between the two groups via the off-benefit rule. The Human Rights Commission, similarly, submits that employment status is “a, if not the, determinative factor” in the situation of those in the group represented by CPAG.

[47] We consider that the High Court and the Tribunal were correct to conclude that there was differential treatment as between groups in comparable situations on the basis of their employment status. Our reasons follow.

[48] The first point we make is that the Crown’s approach to the comparator would impose too high a threshold and effectively cut out the inquiry into potential discriminatory action too soon. The intention of the Human Rights Act is to take what has been described as a “purposive and untechnical” approach to whether there

is prima facie discrimination and so to avoid artificially ruling out discrimination at the first stage of the inquiry.⁴²

[49] The Crown approach also leads to the types of problems that resulted in a move away from the “mirror” comparison analysis.⁴³The high point of the move away from a mirror comparator group is seen in the Supreme Court of Canada decision in Withler v Canada (Attorney-General).⁴⁴The Court outlined concerns arising from the search for a mirror comparator, first, that this search may mean that “the definition of the comparator group determines the analysis and the outcome”.⁴⁵

Secondly, the Court stated that the search for a “precisely corresponding” comparator becomes “a search for sameness, rather than a search for disadvantage, ... occluding the real issue.”⁴⁶Finally, McLachlin CJ and Abella J, in delivering the judgment for the Court, also suggested that another concern was that:

[58] ... allowing a mirror comparator group to determine the outcome overlooks the fact that a claimant may be impacted by many interwoven grounds of discrimination. Confining the analysis to a rigid comparison between the claimant and the group that mirrors it except for one characteristic may fail to account for more nuanced experiences of discrimination. ...

[50] The danger apparent in the Crown approach is that a range of criteria could be established for eligibility but with the knowledge that one of those criteria will effectively cut out and so discriminate against, for example, all those of a particular ethnic group.

[51] It is also necessary to come back to why it is that a comparison is being undertaken. The need to consider this exercise arises, at least in part, because legislation and policy decisions all involve to a greater or lesser extent differential treatment or the making of distinctions of some sort. What the Court is trying to do by reference to the comparator is to sort out those distinctions which are made on the

_{basis of a prohibited ground. The Court is looking at the reality of
the situation not,}

⁴³See discussion in Atkinson, above n 38, at [60]. The observation by Tipping J in McAlister suggesting the need for a close coincidence between the characteristics of the two groups preceded this line of authority.

as Iacobucci J said in Law v Canada (Minister for Employment and Immigration), “in the abstract”.⁴⁷It is necessary also to be comparing apples with apples and hence the inquiry focuses on analogous or comparable situations. The comparator exercise, as has been said on earlier occasions, is simply a tool in that analysis. In some cases, particularly those where there is a single criterion, the comparator analysis will effectively answer the first stage of the inquiry.

[52] Where there are multiple statutory criteria as here, or indeed where effects-based discrimination is being considered, further analysis may be required. There may be questions about how the multiple criteria impact on the choice of comparator and whether the discrimination is on the basis of the prohibited ground. The latter question may raise further questions about the causative link between the treatment and the prohibited ground. We agree with the parties and the Commission that consideration of this aspect is a part of this limb of the test. Accordingly, we do not consider the High Court was correct to consider questions of causation at the second stage. As this Court said in Atkinson, the second step is focused on whether the differential treatment has a discriminatory impact although we acknowledge that the two issues may overlap to some degree. Our approach is consistent with that

[53] The concept of “material ingredient” has been utilised in cases where the legislation refers to discrimination “by reason of” the prohibited ground. In the Human Rights Commission v The Eric Sides Motor Co Ltd, the Equal Opportunities Tribunal stated that for treatment to be “by reason of” a prohibited ground, the prohibited ground had to be “a substantial and operative factor” for the treatment.⁴⁹

CPAG points out that in McAlister, Tipping J said that this description of the test

required “too strong a link between the outcome and the prohibited ground”.⁵⁰

⁴⁷_{Law v Canada (Minister for Employment and Immigration)
[1999] 1 SCR 497 at [57].}

⁵¹See also at [48]–[50] per Tipping J and at [40] per Elias CJ, Blanchard and Wilson JJ.

[49] The correct question raised by the phrase “by reason of” is whether the prohibited ground was a material ingredient in the making of the decision to treat the complainant in the way he or she was treated.

[54] The Human Rights Review Tribunal in Winther v Housing New Zealand Corp did not see the different wording, that is, “by reason of” rather than “on the grounds of” as warranting a different approach to these issues under pt 1A of the Human Rights Act.⁵²Rather, the Tribunal said:

[59] ... [T]he problem is essentially the same. To adapt the approach articulated by Tipping J for use in this case – and bearing in mind the need to keep an eye on the difference between motives and reasons – the question we have to decide is whether the prohibited ground relied on (here, family status) was a material ingredient in HNZ’s decision to issue the plaintiffs with 90 day notices.

[55] In that case the Tribunal concluded that Housing New Zealand had not acted because of its belief as to the nature of the plaintiffs’ relationship with the men involved (that is, the prohibited ground of family status) but rather because of its belief that those men had been involved in serious acts of antisocial behaviour. Accordingly, the action was taken to terminate the connection between the men and the properties rather than on the family status of the plaintiffs.

[56] Turning to the overseas authorities, the point is well made by Iacobucci J in Law. The Supreme Court of Canada in that case was looking at whether an age requirement in the Canada pension plan was discriminatory. The Court described the plan as follows:

[90] The [Canadian pension plan] grants benefits to surviving spouses over the age of 35 immediately following the death of the contributor. However, those benefits are not available to able-bodied spouses without dependent children who are less than 35 years of age at the time of the death of the contributor, until they reach age 65 or unless they should become disabled in the interim. In addition, while those over age 45 are entitled to receive benefits at the full rate, those between the ages of 35 and 45 receive a reduced sum.

_{among the factors determining eligibility for survivor’s benefits
and the amount that}

was provided. The Crown responded that entitlement depended on the “interplay” of

age, disability and responsibility for dependent children.⁵³Iacobucci J stated:

[91] ... In my opinion, it does not follow from the fact that any one of several criteria, including age, might determine entitlement to a survivor’s pension, that the legislation does not draw a distinction on the basis of age.

[92] As an able-bodied woman without children, the appellant does not suggest that the [Canadian pension plan] discriminates by denying her equal benefits as compared to surviving spouses who have disabilities or dependent children. The appellant submits that the issue in dispute is whether age is properly included among the factors which determine eligibility ... and the amount that is provided. Had the appellant been able- bodied, without dependent children, and over age 45 at the time of her spouse’s death, she would have been immediately entitled to receive full benefits. However, as an able-bodied, childless woman who was 30 years of age at the time of her spouse’s death, she is denied any benefits until she reaches age 65, provided she does not subsequently become disabled. Similarly, for surviving spouses aged 35 to 45, it is their age alone that serves to reduce the amount of benefits they receive as compared to those over age 45. In my view, the survivor’s pension provisions of the [plan] clearly draw distinctions on the basis of the enumerated ground of age.

[58] Similar issues were considered by the United Kingdom Supreme Court in Patmalniece v Secretary of State for Work and Pensions.⁵⁴In that case the Supreme Court was considering whether State pension credit criteria were compatible with European Union law.⁵⁵In order to obtain a State pension credit, a person had to be “in Great Britain” as defined. The definition of being “in Great Britain” relevantly included being “habitually resident” and no person could be habitually resident if they did not have the right to reside in Great Britain.

[59] The link to the right to reside was challenged. The challenge was on the basis that nationals of other member states do not qualify for the same treatment unless they have a right to reside in the United Kingdom which they do not have solely on the grounds of their nationality. A citizen of the United Kingdom, by contrast, has such a right.

[60] The Supreme Court followed a European Court of Justice decision in finding that this requirement comprised indirect discrimination although the majority

⁵⁴_{Patmalniece v Secretary of State for Work and Pensions
[2011] UKSC 11, [2011] 1 WLR 783.}

⁵⁵The criteria in question were listed in the State Pension Credit Regulations 2002 (UK), reg 2.

considered that this limitation was justified. The purpose of the limit was to ensure an applicant was sufficiently socially or economically integrated in the United Kingdom before receiving social assistance. This reflected the principle of European Union law that persons dependent on social assistance would be looked after by their own member State.

[61] Since Patmalniece considered whether the alleged discrimination was direct or indirect it is not entirely analogous. Further, the concern in that case related to the cumulative effect of the criteria rather than whether or not, as in this case, two of the criteria would exclude. It is useful nonetheless to consider, first, the observations of Lord Walker, in dissent. Lord Walker expressed the view that the fact there was more than one criterion made it necessary to focus on the criteria as a whole. That was so if it was only one criterion that produces unequal treatment. Lord Walker continued:

[65] ... The right to reside condition is not a sufficient condition for entitlement, but it is a necessary condition, and it is one that is automatically satisfied by every British national. The fact that there is another cumulative condition (actual or deemed habitual residence) is irrelevant ... . It might be different if there were alternative conditions, because neither condition would then be necessary (although one would be sufficient).

[62] Secondly, Baroness Hale, in the context of confirming that the discrimination was indirect, stated:

[93] ... [I]n essence it is the application of a criterion which is applied equally both to nationals and to non-nationals but which in fact places non- nationals at a particular disadvantage when compared with nationals. The right to reside criterion obviously places non-nationals at a particular disadvantage when compared with nationals and has in fact placed Ms Patmalniece at that disadvantage.

[63] Further, Baroness Hale discussed authority suggesting that conduct may be discriminatory where there is an “exact coincidence” between a criterion and the prohibited ground.⁵⁶

[64] The point that emerges from this discussion is that the existence of another criterion which may render the person ineligible for assistance does not of itself

mean there may not be discrimination on a prohibited ground. That is particularly pertinent where there is a close link between the two criteria. That point is illustrated by the facts of this case because the two grounds, that is, being off-benefit and meeting the full-time earner requirement, are factually interlinked. Indeed,

99.2 per cent of persons who are ineligible for the in-work tax credit are ineligible on both grounds. Whether described as a “material ingredient” or as an operative factor, in this case, the reality is that no matter what a beneficiary’s eligibility is under any of the other criteria, that person will never be able to qualify under the “off-benefit” rule so long as he or she exhibits the characteristic on which basis discrimination is prohibited. The criterion operates to exclude people on the basis of a prohibited ground.

[65] Returning to the Crown’s proposed approach, there are difficulties if the purpose of the measure is considered at this stage rather than at the s 5 stage because that conflates matters of justification with the question of whether or not there is prima facie discrimination. The approach reflected in Atkinson was that “matters of justification” should be dealt with at the s 5 stage not at the s 19(1) stage.⁵⁷

[66] For these reasons, we agree with the High Court and the Tribunal that the comparison is between those whose income from work is for sufficient hours to entitle them to the in-work tax credit and those excluded because of the off-benefit rule, irrespective of their full-time earner status. Once that comparator is utilised, it is clear that the differential treatment is on the basis of the prohibited ground.

[67] There are two issues under this head. The first issue is whether the High Court was right to find that only those (1,267) beneficiaries who met the full-time earner requirement are materially disadvantaged or whether, as the Crown contends, there was no material disadvantage. The second issue is whether, as

_{CPAG submits, all beneficiaries with children are at a material
disadvantage.}

[68] As we have foreshadowed, with one qualification, the High Court found that there was no material disadvantage for those on a benefit because the lack of comparable gain was not caused by the off-benefit rule. That was because, even without that requirement, those on a benefit not working the requisite minimum

20 or 30 hours per week would still not qualify for the in-work tax credit unless they began to meet the full-time earner requirement. The High Court said that the only persons who are directly disadvantaged are the 1,267 who are in full-time work but elect to remain on a benefit. This group are at a real disadvantage financially, namely, to the extent of the difference between the total net income as earned from a benefit and that if obtaining the in-work tax credit.

[69] On this point, the Tribunal had accepted CPAG’s submission that children in beneficiary families were disadvantaged by the Working for Families package “at least in the sense that they were left behind the children of families that were in work or for whom a move to work was a realistic possibility”.⁵⁸

[70] CPAG says that all beneficiaries are materially disadvantaged by the off-benefit rule because they are in this way excluded from a comparable gain. In support of the submission that there is a material disadvantage, the Commission emphasises that the effect of the Income Tax Act is that beneficiaries are excluded from claiming the benefit of the in-work tax credit. Further, the Commission says that the inability to claim the benefit amounts to a disadvantage for a significant

[71] The respondent cross-appeals arguing, first, that no-one in the group represented by the appellant is materially disadvantaged by the measure. A person on a benefit either remains on the benefit or receives the in-work tax credit. Accordingly, he or she simply has two forms of state assistance available, each with advantages and disadvantages and can elect one or the other. Secondly, the argument is that in considering the group who do meet the full-time earner requirement, the

_{High Court should have looked at a number of contextual factors. In
that respect, the}

⁵⁹The Commission supports the approach taken by Hart Schwartz “Making Sense of Section 15 of the Charter” (2011) 29 NJCL 201 at 217 which focuses on exclusion as a way of identifying discrimination; see the discussion of Hart Schwartz’s approach in Atkinson, above n 38, at [132].

Crown emphasises that over a person’s lifetime that person may move on and off the benefit and so may at various points in time be eligible for and receive the in-work tax credit. The point is also made that it is accepted that a gap between earnings on a benefit and those in work is permissible and that movement into work is beneficial to society.

[72] In our view, beneficiaries with children are materially disadvantaged by the lack of a comparable gain, namely, the ability to receive the in-work tax credit. The Crown submission is that a more subtle inquiry is necessary but we do not see a need to complicate this part of the analysis. The point of the exercise is to consider the impact on the claimant group in context and that impact must be material. The lack of comparable gain meets that test. Further, a number of the matters referred to by the Crown relating to the benefits of the measure are more relevant to the s 5 inquiry.

[73] In any event, the evidence suggests that the choice, particularly of those on the domestic purposes benefit who otherwise meet the criteria for the in-work tax credit, is a constrained one. Ms Joychild QC pointed to the evidence that those working 20 hours and receiving a benefit primarily continue in receipt of the benefit to maintain some security and certainty of income in a situation where employment is less stable. Further, there may be other reasons, such as childcare responsibilities, that restrict any election to go off-benefit.

[74] Ms Gwyn for the respondent is correct that employment status is not an immutable characteristic like ethnicity. The evidence showed that there is considerable movement on and off the benefit. Over 26 per cent of working-aged recipients of the domestic purposes benefit were in receipt of the benefit for less than a year. However, over 36 per cent of those in receipt of the domestic purposes benefit have been continuously receiving that benefit for between one and four years. Nearly 26 per cent have been continuously receiving the domestic purposes benefit for between four and 10 years. For those in either category, the period of disadvantage is material.

[75] For these reasons, we conclude the in-work tax credit is prima facie discriminatory. It takes as an operative characteristic a prohibited ground of discrimination and results in a lack of comparable gain to beneficiaries with children. Accordingly, we consider that the High Court was wrong to limit the finding of prima facie discrimination to the smaller group of beneficiaries who met the full- time earner criterion. We turn then to the question relating to s 5 of the Bill of Rights. As we have noted, that section provides that the rights in the Bill of Rights “may be subject only to such reasonable limits” as can be “demonstrably justified in a free and democratic society”.

[76] In Atkinson, this Court approached s 5 considering the headings set out by

(a) does the limiting measure serve a purpose sufficiently important to justify [curtailing the right]?

(b)	(i)	is the limiting measure rationally connected with its purpose?
	(ii)	does the limiting measure impair the right ... no more than is reasonably necessary for sufficient achievement of its purpose [minimal impairment]?
	(iii)	is the limit in due proportion to the importance of the objective [proportionality]?

[77] In this case the limiting measure identified by the High Court and accepted by the parties is the off-benefit rule. The purpose is incentivising relatively low income earners to pursue and remain in work. The High Court found, and the parties accept, that this purpose is sufficiently important to justify curtailing the right and that the off-benefit rule is rationally connected to the purpose. The focus of the appeal is therefore on the last of the two questions identified by Tipping J, namely, minimal impairment and proportionality.

[78] Before we address each of these limbs in turn, we discuss briefly some of authorities here, in the United Kingdom and in Canada on the question of the

respective roles of the court and decision maker. We focus particularly on the extent of latitude or leeway to be given to the decision maker, for example, as to the choice of measure adopted as that was the subject of discussion at the hearing and in submissions.

[79] The authorities suggest that how much choice will be afforded to the legislature or decision maker depends on the circumstances. It is generally accepted, and it is accepted in this case, that in matters involving social security and the allocation of spending, a greater degree of leeway will be afforded to the decision maker’s choice.⁶¹

[80] Tipping J in Hansen referred to a spectrum extending “from matters which involve major political, social or economic decisions at one end to matters which have a substantial legal content at the other”.⁶²Tipping J envisaged that the nearer to the legal end of the spectrum the more intense the review by the courts was likely to be. As his Honour said, though, particular matters may have a number of different elements involving different aspects of the spectrum. To illustrate, Tipping J said, “the allocation of scarce public resources can often intersect with questions which, from a different standpoint, may seem more legal than political”.⁶³

[117] Ultimately, judicial assessment of whether a limit on a right or freedom is justified under s 5 of the Bill of Rights involves a difficult balance. Judges are expected to uphold individual rights but, at the same time, can be expected to show some restraint when policy choices arise, as they may do even with matters primarily involving legal issues. ... [Depending on the circumstances] the Court should allow the decision maker

... some degree of discretion or judgment. If the decision maker is Parliament, and it has manifested its decision in primary legislation, the case for allowing a degree of latitude may well be the stronger.

⁶¹This Court in Atkinson, above n 38, at [172] said that the concept of affording a degree of latitude to the decision maker had a “particular resonance in areas such as social and economic policy”.

[82] Tipping J went on to develop the concept of a bull’s-eye, a concept relied on by CPAG in this case. The margin of judgement or leeway left to Parliament represents the area of the target outside the bull’s-eye. The idea is that the size of the target beyond the bull’s-eye will turn on the subject matter. But, and this is the aspect CPAG relies on, Tipping J made the point that Parliament’s view must not miss the target altogether. We come back to this aspect in considering the proportionality of the off-benefit rule.

[83] In the United Kingdom these concepts are well established. There is a helpful summary of the position in Lord Nicholls’ judgment in Ghaidan v Godin-Mendoza as follows:⁶⁴

[19] ... Parliament is charged with the primary responsibility for deciding the best way of dealing with social problems. The court’s role is one of review. The court will reach a different conclusion from the legislature only when it is apparent that the legislature has attached insufficient importance to a person’s [European] Convention rights. The readiness of the court to depart from the view of the legislature depends upon the subject matter of the legislation and of the complaint. National housing policy is a field where the court will be less ready to intervene. Parliament has to hold a fair balance between the competing interests of tenants and landlords, taking into account broad issues of social and economic policy. But, even in such a field, where the alleged violation comprises differential treatment based on grounds such as race or sex or sexual orientation the court will scrutinise with intensity any reasons said to constitute justification. The reasons must be cogent if such differential treatment is to be justified.

[84] CPAG relies on Burnip v Birmingham City Council, a recent decision of the Court of Appeal of England and Wales.⁶⁵That case dealt with the provision of assistance for disabled persons who needed carers throughout the night. The Council provided a housing benefit but quantified it by reference to the one-bedroom rate applicable to the able-bodied. The Court concluded the maintenance of the one-bedroom rule was discriminatory and not justified. Henderson J in a separate

_{judgment noted that the case concerned a benefit designed to meet a
“basic human}

need” for acceptable accommodation for a limited number of persons the “cost and human resource implications” of which were modest.⁶⁶

[85] A similar approach to the question of the nature of the review role has been adopted in Canada. In Alberta v Hutterian Brethren of Wilson Colony, the Supreme Court of Canada was considering a requirement for a photograph on drivers’ licences.⁶⁷The Wilson Colony believes that the Second Commandment prohibits them from having their photograph willingly taken and objected to having their licence photographs taken on religious grounds. Some attempts were made to try to lessen the impact of the requirement but a photograph was still to be taken for placement in the province’s facial recognition databank. These attempts did not meet the concerns of the Wilson Colony so the requirement’s constitutionality was

challenged. McLachlin CJ delivering the judgment for herself, Binnie, Deschamps and Rothstein JJ made the point that under the minimal impairment test, another way of putting the question was to ask:

[53] ... whether there are less harmful means of achieving the legislative goal. In making this assessment, the courts accord the legislature a measure of deference, particularly on complex social issues where the legislature may be better positioned than the courts to choose among a range of alternatives.

McLachlin CJ continued that the test at the minimum impairment stage was whether there was “an alternative, less drastic means of achieving the objective in a real and substantial manner”.⁶⁹

[87] To illustrate the point that the leeway is not without its limits, reference can be made to the earlier decision of Eldridge v British Columbia (Attorney-General).⁷⁰

The Supreme Court of Canada concluded in that case that the failure of hospitals and

_{the Medical Services Commission to provide sign language interpreters as
an insured}

benefit under the Medical Services Plan to deaf persons using their medical services was an unjustified breach of the anti-discrimination provision. The Court was critical of the failure to attempt to set up a scheme constituting a lesser limit.⁷¹In that case, the Court had evidence that the estimated cost of providing sign language interpretation for the whole of the province, British Columbia, was $150,000 or “approximately 0.0025 per cent of the provincial health care budget at the time”.⁷²

[88] Reference should also be made to Gosselin v The Attorney-General of Quebec as CPAG says that case is closest on its facts to the present.⁷³In 1984 the Quebec Government set up a new social assistance scheme. One regulation of that scheme provided people who were single, unemployed and under the age of 30 with a base amount of welfare approximately one third of that for people over 30. Under the scheme, if a person under 30 participated in an approved employability programme then their welfare payment was increased to be close to, or the same

amount as, that payable to those over 30. In 1989, new legislation removed this age- based distinction. The appellant brought a class action challenging the 1984–1989 scheme alleging breach of, amongst other rights, the right equivalent to s 19 in the Canadian Charter of Rights and Freedoms (s 15). The majority concluded that the welfare scheme did not breach this provision so did not reach the justified limits analysis. However, it is worth noting two comments made by McLachlin CJ for the majority, albeit in the context of considering whether the measure was discriminatory. The Chief Justice said:

[55] ... Perfect correspondence between a benefit programme and the actual needs and circumstances of the claimant group is not required to find that a challenge provision does not violate the Canadian Charter. The situation of those who, for whatever reason, may have been incapable of participating in the programs attracts sympathy. Yet the inability of a given social program to meet the needs of each and every individual does not permit us to conclude that the program failed to correspond to the actual needs and circumstances of the affected group. ... Crafting a social assistance plan to meet the needs of young adults is a complex problem, for which there is no perfect solution. ...

[56] Second, we cannot infer disparity between the purpose and effect of the scheme and the situation of those affected, from the mere failure of the government to prove that the assumptions upon which it proceeded were

correct. ... [T]he legislator [does not have a duty] to verify all its assumptions empirically, even where those assumptions are reasonably grounded in every day experience and common sense. ...

[89] CPAG relies on the approach taken by the dissenting judges, L’Heureux-Dubé and Bastarache JJ. L’Heureux-Dubé J in the context of considering whether the scheme was prima facie discriminatory stated:

[112] ... By necessary implication, the fact that a legislature intends to assist the group or individual adversely affected by the impugned distinction also does not preclude a court from finding discrimination. Nor is it determinative, where a distinction produces prejudicial effects, that a legislature intends to provide an incentive for the affected individuals to alter their conduct or to change themselves in ways that the legislature believes would ultimately be beneficial for them ...

[90] In terms of the minimal impairment test, Bastarache J agreed that the Court should avoid “second-guessing” government policy and that the government need not have chosen the less drastic means available. However, the Judge continued that the government must have chosen to infringe the right “as little as was reasonably possible”.⁷⁴In that case, Bastarache J examined the evidence and concluded that there were other available alternatives, for example, the creation of a universally conditional programme that would achieve the objective but in a way that was less

[91] The effect of these authorities is therefore, that in approaching the s 5 analysis, some latitude or leeway is given to the legislature or the decision maker particularly in a case like the present which involves the complex interaction of a range of social, economic, and fiscal policies as well as taxation measures. In addition, those policy factors relate to the overall social assistance measures with various tiers of benefits for the relief of poverty, as well as incentives to encourage beneficiaries to move into employment. That latitude or leeway to the legislature does not however alter the fact that the onus is on the Crown to justify the limit on

[92] It must also be kept in mind that the effect of the Human Rights Act and the Bill of Rights is that when a measure is prima facie discriminatory the courts have to decide whether or not the measure meets the s 5 threshold. As Lord Scott said in A v Secretary of State for the Home Department, the function of measuring compliance with human rights norms is not one “that the courts have sought for themselves” but it is nonetheless a function that has been “thrust” on the courts by the Human Rights

Act and the Bill of Rights.⁷⁶In that context, the term “deference” as used in the

authorities is not helpful if it is read as suggesting the court does not need to undertake the scrutiny required by the human rights legislation. The courts cannot shy away from or shirk that task. Rather, it is a question of recognising the respective roles of the courts and the decision maker, here, the legislature.⁷⁷

[93] With these principles in mind, we turn then to the minimal impairment limb.

[94] The issue under this heading is whether the High Court was correct to conclude that the Crown had shown the right to be protected from discrimination on the ground of employment status was minimally impaired given there were other means of achieving a gap between earnings on- and off-benefit. A related issue is whether the High Court gave undue deference to the legislature’s choice of method to incentivise work.

[95] The High Court in reaching this conclusion noted that the objective of the off-benefit rule was to create a sufficient margin between earnings whilst on a benefit and earnings whilst in work. Given the influence of other non-financial incentives to pursue employment and the various different individual circumstances affecting the decision, it was not reasonable to decide how much of a gap was justified. The High Court did nonetheless look at the approximate extent of the

_{difference in the financial position of the two groups. That is set out
in the chart}

annexed to this judgment as Annexure B.⁷⁸The Court said that the figures in the chart suggested a gain for moving off the benefit of some $23 for a couple and $35 for a sole parent. The Court continued:

[204] ... By far the largest State payment made in recognition of child care responsibilities is [family support], which is paid indiscriminately to all principal caregivers (included in the chart at $139 per week for two children under 12). For those earning at the lowest levels, the biggest contributor to the difference between earned income and beneficiaries is the [minimum family tax credit], which abates first.

[96] Accordingly, the Court was not persuaded that in a quantitative sense there was a greater disparity and therefore a greater impairment on the right than was reasonably necessary to achieve the objective of a work incentive.

[97] The Court said that when the matter is looked at in a more qualitative way another qualifying requirement for the in-work tax credit is that the recipient is the principal caregiver of dependent children. The High Court stated:

[211] ... Predictably, no issue was raised that that is discriminatory against those without children. Once the relevant category is confined to those caring for dependent children, then a distinction can appropriately be drawn between those in receipt of benefits, and those in work. From that perspective, the limiting measure does not impair the right to be free from discrimination on the grounds of employment status more than is reasonably necessary to achieve the objectives of the [in-work tax credit].

[98] The Tribunal said that “at a most basic level” it is difficult to see how the objective of making work pay could ever be achieved without creating or enhancing a gap between the income available from benefits on the one hand and the financial rewards for being in work on the other.⁷⁹The Tribunal continued:

[257] The business of creating or enhancing that gap is reasonably necessary for sufficient attainment of the legislative purposes; more than that, we think it was an indispensable element for achievement of the [Working for Families] package. No matter what other elements there were, the legislative initiatives would not have made much sense, nor could they realistically have been expected to be effective, otherwise. [Working for Families] was intended to leave families on benefit income with financial reasons to move into work.

[99] CPAG says the High Court, in concluding that the measure met the minimal impairment threshold, has given too much deference to the legislature. Ms Joychild in developing the submissions on this point notes, first, that there is already a gap between earnings on benefit and those in work because of the minimum family tax credit. The submission is that the problem the Government was trying to address is largely confined to the small group of beneficiaries who are working 20 hours a week and staying on the benefit for security of income. Secondly, CPAG relies on the evidence of Dr St John as illustrating there were many reasonable alternative ways of achieving the objective. Finally, Ms Joychild emphasises the relevant international obligations and the failure to consider these in developing the Working for Families package.

[100] The Commission is similarly critical of what it sees as the High Court’s

[101] The Crown supports the approach taken by the High Court and the Tribunal and says those decisions have not relied unduly on the concept of deference. Ms Gwyn adds that the Court should not be in the business of considering the other possible alternatives advanced by CPAG because that would be to engage in political debate absent any “legal anchor”. If this exercise is undertaken, the Crown says it is apparent that each of the options has other disadvantages. In any event, it is not the Crown case that those alternatives are too expensive but rather that they are alternatives to the in-work tax credit and not to the off-benefit rule.

[102] As was the case in Atkinson, this part of the test can be dealt with by considering whether the approach taken fell within a range of reasonable alternatives. This limb was discussed by the Supreme Court in R v Hansen.⁸⁰In that

_{case, Blanchard J said that “a choice could be made from a range
of means which}

impaired the right as little as was reasonably necessary”.⁸¹Tipping J preferred to phrase the question as whether or not the limit was “no greater than is reasonably necessary to achieve Parliament’s objective”.⁸²His Honour considered that approach built in “appropriate latitude to Parliament” and would not “unreasonably circumscribe Parliament’s discretion”.⁸³Finally, McGrath J stated that the inquiry was into “whether there was an alternative but less intrusive means of addressing the legislature’s objective which would have a similar level of effectiveness”.⁸⁴

[160] ... [T]he law must be carefully tailored so that rights are impaired no more than necessary. The tailoring process seldom admits of perfection and the courts must accord some leeway to the legislator. If the law falls within a range of reasonable alternatives, the courts will not find it over-broad merely because they can conceive of an alternative which might better tailor objective to infringement ... . On the other hand, if the government fails to explain why a significantly less intrusive and equally effective measure was not chosen, the law may fail.

[104] The starting point in considering whether it has been shown that the in-work tax credit meets the minimal impairment requirement is that it is accepted that there has to be a gap between earnings on- and off-benefit. There is no dispute then that there has to be a measure creating such a gap. Once that is accepted, the focus turns to the way in which the gap is achieved. In that respect, the evidence before the High Court and the Tribunal of the experience of other member countries in the Organisation for Economic Co-operation and Development (OECD) was that a measure like the in-work tax credit is a recognised alternative in other democratic societies.

[105] The Tribunal heard evidence from two experts who worked with the OECD. Their expertise was in the effects of tax and benefit systems on employment and

30 member countries of the OECD have some form of grants that are dependent on being in work. Their evidence was that although in-work benefit policies were not “some sort of panacea for all social and labour market problems” the international evidence was that these policies were effective in raising the employment rate of the target group and in reducing poverty. They said that if New Zealand had not already introduced such a policy, the OECD would recommend that it do so.

[106] CPAG says that the models elsewhere are different and that this one is less tailored. That may be so, but it must be relevant to an assessment of the off-benefit rule that the OECD experience supports this type of alternative.

[107] CPAG also relies on the fact a gap already exists by virtue of the minimum family tax credit. However, in considering whether the off-benefit rule minimally impairs the right, it is no answer to say there is already an incentive to stay off-benefit. The purpose of the in-work tax credit is to ensure it is not more economic to stay on the benefit. The evidence of Donald Gray from the Ministry of Social Development was that the gap created by the minimum family tax credit was not sufficient, so both forms of tax credit were necessary. Suzanne Mackwell from the same Ministry explained that the minimum family tax credit is withdrawn by $1 for every $1 increase in income so has little incentive to increase earnings. She said that “[f]or this reason [this tax credit] was designed to operate across a quite narrow band of income for a small group of families”. Further, there are a number of factors in play that influence the decision as to how big a gap is necessary.

[108] It is also important that this is not a case where the Government has latched on to one option without careful consideration of the alternatives. Reference was made in argument to the observation of Abella J dissenting in Quebec (Attorney-General) v A.⁸⁶Her Honour said that the key factor is the ultimate legislative choice and “the degree of legislative time, consultation and effort cannot act as a justificatory shield to guard against constitutional scrutiny”.⁸⁷That goes

implementation is always relevant. Its absence will be significant, for example, in considering the leeway to be afforded a particular choice.⁸⁸Its presence will often be helpful for government. It is one of the relevant factors.

[109] The policy process leading to the enactment of the Working for Families legislation was an extensive one. Two points can be made about that process.

[110] First, a range of options were explored and tested against various indicators. That process was assisted by advice and research, for example, as to the programmes in other OECD countries. In addition, a comprehensive review of living standards in New Zealand was prepared as well as an analysis of the interrelationship between parental income and outcomes for children. One of the papers prepared early on in the process identified and evaluated what were described as three possible

[111] Each of these options was assessed against a range of factors including “poverty alleviation” and “employment outcomes”. The “Pathways” paper also discussed various options such as a universal benefit and rationalisation of the current tax and benefit based assistance.

[112] Further, as the Tribunal notes, “early design” of the Working for Families package also considered the pros and cons of a “two payment” as against a “single payment” system.⁹⁰In a single payment system, there would be an income adequacy payment and no separate work incentive payment. The two payment systems

_{involved “one set of payments designed to ensure income adequacy
and another}

⁸⁸The Supreme Court of Canada in Health Services and Support v British Columbia 2007 SCC 27, [2007] 2 SCR 391 at [156]–[157] was critical of the absence of any consideration on the record of less intrusive means and, in a case involving collective bargaining, was also critical of the very limited consultation with the unions.

⁸⁹Ministry of Social Development Future directions for social assistance (19 December 2001) at

designed to make work pay”.⁹¹By October 2003, officials were favouring a two payment system primarily on the basis a single payment system was not seen as being likely to achieve a sufficient work incentive.

[113] The second point that emerges from a consideration of the policy process is that there was discussion about a balancing off between the amounts of money invested into poverty alleviation against that invested in work and how striking a different balance could impact on the approach taken to social assistance.

[114] There was one “unexpected development” in the process leading up to the Working for Families package which occurred in or around about November 2003 that we need to discuss.⁹²Prior to that point, it was understood that the level of expenditure that might be approved was relatively modest. However, in late 2003, the Government decided it had more resources available for the upcoming 2004

$100 million, $300 million or, at the upper end, $450/500 million, officials found themselves working within a $1.1 billion expectation. But by then there was very little time left in which to get the [Working for Families] package ready for the 2004 Budget announcement. Our sense of the evidence on this topic was that by the time officials learned of the extent of the available funding, it was simply too late for them to go back to reconsider the ‘one payment’/‘two payment’ alternatives – even though that might have been appropriate, and notwithstanding that if officials had known the level of funding at the outset they might have made different recommendations for [Working for Families].

[115] CPAG is critical of the failure to reconsider options at this point. However, in addition to the time constraints referred to by the Tribunal, Ms Mackwell expressed some uncertainty about whether officials would have “got to” a single payment approach. That was in part because of difficulties perceived in making the one-payment approach address both of the objectives of the Working for Families package, namely, making work pay and income adequacy. In any event, we do not see this as detracting from the overall proposition that a range of possible alternatives including more fundamental change such as a move to a universal

[116] CPAG also points to the lack of consultation, the absence of specific reference to human rights obligations and international obligations and the absence of any report of inconsistency with the Bill of Rights as required under s 7 of that Act. The submission is that these omissions should lead to a reduction in any deference or latitude given to the government’s choice of measure. CPAG’s focus in this respect is on the child’s right to a standard of living. That is obviously important but the key focus here is whether the right to protection from discrimination on the basis of employment status has been minimally impaired.

[117] The absence of consultation no doubt reflected the fact the Working for Families package was introduced as a budget initiative. CPAG did give a presentation to two officials at one stage. Further, a number of government departments with different perspectives were involved in the policy development process. Finally, as we have noted, the proposals were measured against a range of different indicators.

[118] Consideration of human rights issues over the course of the policy development process was limited. Mr Nutsford explains issues that may have had human rights implications were identified “quite early” in the process although he did not know a great deal about what work was done on this. Ms Mackwell said the Ministry’s lawyers and the Department of Justice were involved. It is fair to say not a great deal was said in the papers about this and that closer attention to these issues and the associated international obligations would have been beneficial.

[119] As to the s 7 report, the respondent’s position remains that there is no prima facie discrimination. Presumably, the advice to the Attorney-General would have reflected that. Certainly, it does not necessarily follow from the failure to report per se that this matter was not given any consideration. The Attorney-General did table a s 7 report stating that the effect of the Working for Families bill breached the right to be free from discrimination on the basis of sexual orientation in treating

[120] When analysed in the round, it can be said that the adoption of this particular method of work incentive reflected a careful analysis of a range of possible options each with differing social and economic impacts. The method chosen was at least broadly in line with the approach taken in other democratic societies. This Court in Atkinson made the point that the fact there is an alternative with less impact does not

mean the option adopted is not reasonable.⁹⁴However, in this case we can test the

availability of another, less intrusive option, by briefly reviewing the other alternatives advanced by Dr St John for CPAG. When that analysis is undertaken, it is plain that each of the other possibilities raises other issues, for example, as to the impact on effective marginal tax rates.

[121] The options advanced by Dr St John were as follows: (a) a low income rebate or initial tax free band; (b) increasing the minimum family tax credit;

(d) a tax credit that differentiates on employment status but not family status;

[122] As to the first of these options, it was explored in the policy development process. Ms Mackwell’s evidence was that the suggested changes to the low income rebate would have had a very small impact on work incentives in comparison to the in-work tax credit and no effect on income adequacy.

[123] Dr St John’s response was that there were ways of changing the low-income tax structure that could have a greater effect on returns from work for low-income

workers.⁹⁵However, Ms Mackwell’s more general point was that the “key issue” with using tax reform to make work pay was targeting the group most in need. That was because tax is assessed on an individual not a family basis. She also said that targeting those on low to modest incomes only via this method was difficult and so fiscally more costly.

[124] The minimum family tax credit, as we have noted, also includes an off-benefit rule. It is not clear why this would be less rights impairing. Further, there was evidence that larger increases to this tax credit would distort the labour market by creating larger poverty traps at the low end of the income scale. Dr St John captured the concern when she accepted Ministers were “worried about it because it is so ugly with its marginal effective tax rate”.

[125] The point made by the Crown submissions in relation to a lump sum or time limited payment was that this provides a work incentive only for that period. If the recipient has not increased his or her income sufficiently within that period, there may be low incentives to stay in work. There was also evidence that such payments may also incentivise people to leave employment to gain access to the payment. Finally, such lump sum payments would necessarily not be available to beneficiaries so, again, it is unclear that they would be less rights impairing than the in-work tax credit.

[126] We also agree with the Crown submission that it is unclear why a tax credit differentiating on the basis of employment but which is not child related (the United Kingdom model) would be less impairing of the right to be free from discrimination on the basis of employment status.

[127] As to changes in the minimum wage, Ms Mackwell made the point that these are part of the employment relations framework. She accepted that increases in the minimum wage complement the goal of making work pay but “because they must balance competing objectives and do not recognise family type and number of children”, they are insufficient to achieve the objective of making work pay for

[128] Finally, in terms of abatement changes, there were some changes to abatement rates as part of the Working for Families package. Ms Mackwell said that, on their own, these were insufficient to make work pay for most low income families. Further, the evidence was that relaxing the abatement for recipients of the domestic purposes benefit would not encourage families to move off benefit but in

fact would have the opposite effect.⁹⁶Mr Nutsford explained that adopting different

abatement regimes for family support and the in-work tax credit would have an effect on effective marginal tax rates.

[129] Accordingly, this is not a case where there is an obvious reasonable alternative less impairing of the right that has been overlooked. Of course, it is possible to envisage a combination of the various options advanced by Dr St John but for us to weigh up the pros and cons of the possible combinations would be to stray beyond our role. In an area where there are a range of legitimate policy choices, an approach has been adopted following conscientious consideration of those options. That approach achieves a gap in earnings on- and off-benefit but it is accepted that the creation of a gap is a legitimate part of a work incentive measure. In these circumstances, the High Court was correct to conclude that the minimal impairment test has been met and has not unduly deferred to the decision maker’s choice of method.

[130] The issue under this heading is whether, given the cost of the scheme, its benefits outweigh its adverse effects on beneficiaries. CPAG says the salutary effects do not outweigh the deleterious effects. The Crown responds that they do.

[131] The High Court accepted that the high point of CPAG’s argument was the “relatively modest levels of success” on the goals of encouraging beneficiaries into work and alleviating child poverty given the significant overall expense.⁹⁷However, the High Court said that argument did not deal with the fact that the exercise at this

_{stage involved looking at proportionality between the off-benefit
rule and the}

objective of encouraging beneficiaries into work. If looked at in this way, the extent of the harm to the right is not out of proportion with the objective.

[132] The Tribunal said that the reality was that any such scheme involves a level of harm to the right. Looking at the matter overall, the Tribunal was satisfied that the “practical benefits” were sufficient to outweigh the damage.⁹⁸

[133] The appellant says the High Court has not focused on the key question of whether the salutary effects outweigh the deleterious effects. CPAG emphasises that the benefit was for a small number of persons whilst involving a very expensive work incentive. The expected effect of the incentive in moving people off-benefit was about two per cent. The submission is that this does not represent sufficient social gain given the cost. In her evidence, Dr St John describes this as a very expensive way to achieve the objective and says that it is not well targeted.

[134] The Commission is similarly critical of what it describes as a failure of the High Court to “engage” in the exercise “as it did not take into consideration those affected by the exclusion”.

[135] For the respondent, Ms Gwyn submits that the question on this point is whether the harm to the right that arises as a result of the off-benefit rule is proportional to the benefit achieved. When viewed in this way, the submission is that the High Court and the Tribunal were right, particularly as the purpose is to lessen the barriers to beneficiaries moving off-benefit. CPAG’s claim is really a claim for an increase in family support.

[136] To put the issues under this heading in context, the effect of pt 1A of the Human Rights Act is that it is not permissible to discriminate on the grounds of employment status. Measures chosen to achieve a work incentive accordingly have

_{to be implemented in a way that meets the s 5 test. While incentivising
work is an}

important objective, the measure chosen to achieve it must be proportionate. As to how proportionality is assessed, the description of the questions for this part of the test in Canada (Attorney-General) v JTI-MacDonald Corp is helpful.⁹⁹In that case, the Supreme Court of Canada said:

[42] ... This inquiry focuses on the practical impact of the law. What benefits will the measure yield in terms of the collective good sought to be achieved? How important is the limitation on the right? When one is weighed against the other, is the limitation justified?

[137] The modelling used in the Working for Families papers forecast “modest” movement off-benefit as a result of the in-work tax credit in the order of two per cent. CPAG posits the comparison between the cost of a gain of that order ($593 million) and a benefit of about two per cent.

[138] We do not consider that is the correct comparison. First, no one disputes some expenditure on a work incentive is legitimate. Accordingly, on CPAG’s approach less may have been spent on the in-work tax credit but we can assume a sum would still have been incurred. For example, if as an alternative a combination of an in-work tax credit and, say, an increase to the minimum wage was adopted, a proportion of the $593 million forecast to be expended on the in-work tax credit would still have been spent on that tax credit. Accordingly, the comparison is not as stark as CPAG suggests.

[139] Secondly, the evidence showed that the effects were better than the two per cent forecast. An Inland Revenue/Ministry of Social Development report dated January 2010 considered the labour market effects of the changes to financial incentives and support.¹⁰⁰That report noted that the employment rate of sole parents aged 18 to 64 increased by 10 per cent from 48.3 per cent in the quarter ending June

_{2004 to 57.7 per cent in the quarter ending June 2007. The report
noted that the}

¹⁰⁰Inland Revenue/Ministry of Social Development “Employment incentives for sole parents: Labour market effects of changes to financial incentives and support” (January 2010), and see also Jacinta Dalgety and others “Employment incentives for sole parents: labour market effects of changes to financial incentives and support” in Iris Claus and others (eds) Tax Reform in Open Economies: International and Country Perspectives (Edward Elgar Publishing, Cheltenham,

employment rate for sole parents grew faster than that for all of the working age population over that period.

[140] The report also noted that the numbers of sole parents receiving the domestic purposes benefit were fairly stable from 1999 to March 2005 but since that time New Zealand experienced “the largest fall” in the number of persons receiving the domestic purposes benefit since the benefit was introduced in 1973.¹⁰¹Numbers of all domestic purposes recipients decreased from 107,900 at the end of March 2004 just prior to the introduction of the first changes to the accommodation supplement

12 per cent. There had been an increase in the numbers of those receiving the domestic purposes benefit in the latter part of 2008 but the numbers, according to the report, are still well below the level seen in the early 2000s.

[141] The paper accepts that these changes cannot be attributed solely to the policy changes. Other factors, such as changes in the economy, are also responsible. However, the report estimates that around two thirds of the increase in the employment rate for sole parents can be estimated to be due to changes in financial incentives and support for work. Accordingly, the report writers estimate that six per cent of the increase was due to the policy changes.

[142] The report also notes that the percentage of sole parents meeting the eligibility threshold for the in-work tax credit requirement increased from

35.9 per cent in June 2004 to 47.5 per cent in June 2007, an increase of 11.6 per cent. The analysis suggests that around 9.2 per cent of this increase was due to the policy changes. Finally, the policy changes have increased the rate of domestic purposes benefit sole parent recipients’ exit from benefit and decreased the rate of re-entry to benefit for sole parents.

[143] Hence, there is force in the High Court’s conclusion that the projections in the Cabinet paper were “relatively conservative”.¹⁰²The fact individuals may

_{receive the in-work tax credit at some point, reflecting the
movement on- and}

off-benefit, is also relevant in assessing the effects. Therefore, the High Court

referred to the impact on a “rolling basis” as higher than the two per cent forecast.¹⁰³

We interpolate here that the Supreme Court of Canada in the Hutterian case suggested that the legislature was not required to show that the law in issue will in fact produce the forecast benefits.¹⁰⁴

[144] Two further points can be made about these effects. First, whatever the means chosen to incentivise work, the state of the economy will be a factor in its effectiveness. That is one of many variables that the government has to work with in determining the appropriate levels of expenditure. Secondly, our impression of the evidence is that relatively small gains are of significance in this area. The evidence was that an increase in employment rates of single mothers of 4–7 percentage points in the United States was seen as “substantial”. Likewise, a 10 per cent increase in employment in the United Kingdom was also seen as substantial employment

gains.¹⁰⁵Mr Gray made the point that any movement in this area was likely to be

small but in itself that was not insignificant. Accordingly, there is a basis for regarding the positive effects as proportionate to the off-benefit rule. The measure has not missed the target.

[145] In any event, the logical extension of CPAG’s submission is that more money should have been spent on beneficiaries with children to alleviate child poverty. On that approach, the focus is on the comparative amounts spent on family support and the in-work tax credit. At the time of the introduction of the Working for Families package, 60 per cent of the funding was directed towards those in work and

40 per cent to those on a benefit. Ms Mackwell in cross-examination accepted that with the 2005 changes the percentages were probably more like 70 to 30 per cent.¹⁰⁶

[146] The evidence was of forecast government expenditure for 2005/2008 as follows:

	$ m
Family support	2,000
Child tax credit	13
Special benefit/temporary assistance allowance	69
In-work tax credit	593

When those figures are utilised the $2,000 million spent on family support is approximately 75 per cent of the total and the $593 million spent on in-work tax credit is approximately 22 per cent of that total.

$593 million. Dr St John in her evidence was critical of the amount spent on those earning over $45,000 per year. On the 2006/2007 figures that was some 48 per cent or $229 million.¹⁰⁷In that year, however, using the same figures, some 31 per cent was spent on families earning less than $35,000 a year. An alternative grouping is those earning $25,000 to $45,000 a year and some 37 per cent of the total figure was spent on this group. The latter is the group from which Ms Mackwell explains the government was looking for the most gain. Hence the 60/40 split. Mr Nutsford in his evidence said that the bulk of expenditure in the Working for Families package

[148] On either analysis, it cannot be said the government’s approach is out of

[261] The legislation we are concerned with marks a point somewhere between two extremes. At one end there is a solution in which no government spending is directed towards those who are out of work, and social spending is directed only to those who are in work. At the other end there is a solution in which no government spending is directed towards those who are in work, and it is all directed to those who are out of work. The closer one gets to either of these limits, the easier it would be likely to find a “florid” violation of rights of the kind contemplated by Laws LJ and Lord Walker. But within reasonable limits along that continuum, the issue as to where the balance should be struck at any given point in time is inescapably a political question.

[149] Drawing these threads together, it is apparent that there are a number of considerations at play here. First, the balance struck in the decision making will

reflect social and economic factors such as the state of the economy, assessment of levels of poverty and relative disadvantage. The ongoing political process will mean differing views over time as to where the balance should be struck as between those factors. The relative proportions of money spent for relief of poverty as distinct from the other objective of moving people off-benefit and into work is very much a matter of overall political judgment.

[150] Secondly, a decision about where the balance should lie in terms of the amount spent on family support vis-à-vis a measure like the in-work tax credit, will undoubtedly impact on other areas of spending. This is one of the points made by John Yeabsley, an economist working at the New Zealand Institute of Economic Research, who gave evidence before the Tribunal about the policy development process. His opinion was that “the business of government in the social policy area is a complex and practically difficult job”. He explained that often even “relatively small adjustments in policy settings can radically alter the choices that most suit some families”. Mr Yeabsley noted the difficulty in predicting such outcomes especially given the diversity of family situations and their economics. We observe that the evidence disclosed the complex interrelationship between the various forms of benefit, abatement provisions and taxation measures, to name a few of the considerations relevant to the overall package. A change in any one of these material considerations may have an effect on the others.

[151] Finally, in the present case, once it is accepted that the other limbs of the s 5 test are met, it inevitably becomes harder to say that the measure that results is not proportionate. In making this point, we should make it clear that we do not subscribe to Professor Hogg’s view that when an objective is sufficiently important to meet the s 5 test, is rationally connected and minimally impairs the right, the

proportionality analysis adds nothing.¹⁰⁸However, in this area of socio-economic

policy the reality is that the various limbs of the s 5 analysis overlap. That said, we have found this aspect of the s 5 analysis difficult.

[152] Our difficulty no doubt reflects the undisputed fact that the issue of child

_{poverty highlighted by CPAG’s present claim is a serious one.
CPAG points to the}

¹⁰⁸Peter Hogg Constitutional Law of Canada (looseleaf ed, Carswell) at [38.12].

impact of benefit cuts in 1991 and the problems of child poverty for beneficiary families since that time. The figures CPAG relied on in the evidence show that when the legislation implementing the Working for Families package was enacted in 2004, there were some 245,000 children living in beneficiary families. CPAG refers to evidence from Janfrie Wakim indicating that approximately 180,000 of those children were living in hardship and 150,000 of those are described as in

“significant” hardship.¹⁰⁹Nor is there any dispute about the consequences of

poverty for children in particular and for society more generally. The Tribunal and the High Court both recognised these issues and their importance. Accordingly, CPAG properly seeks to focus attention on the need for consideration of this issue. We have, nonetheless, concluded that the High Court was correct that the off-benefit rule is proportional to the work incentive objective.

[153] Accordingly, while our reasoning on this aspect differs in some respects from that in the High Court we agree that the s 5 test is met.

[154] For these reasons, we answer the questions of law on which the appellant was granted leave to appeal as follows:

(i) First question: Did the High Court correctly state and apply the test for a breach of s 19 of the New Zealand Bill of Rights Act 1990?

Answer: No. The High Court should have concluded that all beneficiaries were the subject of prima facie discrimination because of the off-benefit rule.

(ii) Second question: Did the High Court correctly state and apply the test for s 5 of the New Zealand Bill of Rights Act 1990?

Answer: Yes. The High Court correctly found that the off-benefit rule is a justified limit on the right to freedom from discrimination

¹⁰⁹Bryan Perry Household incomes in New Zealand: trends in indicators of inequality and hardship

1982 to 2008 (Ministry of Social Development, 2009); and see also John Jensen and others New Zealand Living Standards 2044: Ngā Āhuatanga Noho o Aotearoa (Ministry of Social Development, 2006).

[156] The parties are agreed that there is no issue as to costs. We accordingly make no order as to costs.

	Sole Parent				Couple
Tax credit and benefit examples	Working 20 hours			Not working	Working 30 hours			Not working	6 hours work
As at April 2007	DPB claimed (net)	Wages (gross)	IWP claimed (net)	DPB claimed (net)	UB claimed (net)	Wages (gross)	IWP claimed (net)	UB claimed (net)	IB claimed (net)

Court of Appeal of New Zealand

Child Poverty Action Group Inc v Attorney-General [2013] NZCA 402; [2013] 3 NZLR 729; (2013) 9 HRNZ 742 (30 August 2013)