By Ben Hoffman
Economic Justice Australia has asked me to reflect on the recent change of the Department of Human Services approach to now consider the presence of domestic violence when determining conjugal status for social security entitlement under the Social Security Act 1991 (Cth). While a welcome change, it greatly surprises and distresses me this was not already happening; in New Zealand, we settled that this was a required consideration in 1996, and it is arguable that the Act has always required it. It seems this was not previously the opinion of the Department nor Tribunal, so I am surprised that nobody took the issue to the Federal Court to have the interpretation clarified. That aside, our experience on this side of the ditch loudly warns that just having something in a policy or court decision is entirely inadequate and there must be continuous anxious scrutiny of its application. This is a sordid tale of our New Zealand experience of the executive’s failure to apply the law well (indebting vulnerable people, usually single parents, with alleged debts that are often incorrect), the barriers to challenging those decisions, and the importance of maintaining specialist legal help.
The New Zealand situation
Like Australia, relationship (conjugal) status is a fundamental factor affecting entitlement to New Zealand social security benefits. Entitlement to particular benefits, most commonly, assistance as a sole parent, the payable rate and means testing, all turn on relationship status. The most complicated provision is the ability to treat two unmarried people as having entered “a relationship in the nature of marriage” our equivalent phrase for “member of a couple.” Determining if such a relationship exists is one of the most complex decisions in all social security law. Our courts had traditionally held that this required the voluntary merging of lives in the same way as a de jure married couple driven by sufficient emotional and mental commitment. This was examined based on physical indicia, with the required commitment inferred from this. In the round, the traditional test did not require “going behind” the external veneer of the relationship to examine what underpinned it.
On 1 October 1996, a full bench of the Court of Appeal significantly altered the previous test in its landmark decision, Ruka v Department of Social Welfare. The District Court originally convicted Ms Ruka of benefit fraud for claiming assistance as a sole parent while being in what it found to be a relationship. The High Court initially upheld this on appeal. Both courts accepted that appalling violence was a central feature of the relationship and that she was trapped in it, but nonetheless found that the relationship was “in the nature of marriage.” The Court of Appeal quashed this finding and held it was not so due to the lack of financial support and Ms Ruka’s lack of commitment to the relationship. The majority held the test to require two contemporaneous features:
- Primarily, willingness of the parties to assume financial responsibility for the maintenance of each other if one had or developed inadequate income (termed “financial interdependence”). The Court held that this must go beyond the mere sharing expenses as flatmates may do, and went as far as holding that the lower courts erred in not primarily focussing on this aspect, particularly in the context of social security law; and
- A genuine mental commitment to cohabit through the voluntary merging of lives in the same way as a married couple.
Ruka significantly altered the previous test because it added the requirement for the willingness of the parties to assume responsibility for each other as the primary feature. The Court also required a much deeper enquiry by “going behind” physical indicia to examine the underlying commitment, rather than simply inferring it. It also expressly acknowledged that the presence of domestic violence can negate the required emotional commitment if the victim only stays out of fear and control. I pause to acknowledge the lawyers’ efforts in bringing this appeal as it resulted in a much more explicit test than the previous authorities.
Deficiencies in administration
Despite the significantly favourable ruling in Ruka, the chronic failure of welfare authorities to apply it accurately, or even at all, is of considerable concern and has been an issue for years. In 2000, the Labour government commissioned Frances Joychild QC to audit how the Department had implemented the fundamentally changed test. She was almost universally scathing and found a widespread failure to implement the test from October 1996 to mid-2000. Most appallingly, she found a complete failure to update instructions for staff and information for the public to include the revised test until nearly four years after it became law. She recommended the proactive reassessment of all overpayments established due to alleged undeclared relationships from 1 November 1996 to December 2000. The Department chose not to do this and instead shifted the onus to those affected to apply for reassessment; few did – only 5,700 of approximately 15,000 eligible (38%). Disturbingly, 63% of reassessments resulted in disestablishing the debt.
More recently, since 2017, the Department’s successor, the Ministry of Social Development (MSD) has disestablished $304,305.02 in alleged overpayments for my clients where it (or a predecessor) wrongly decided that the clients were in a relationship in the nature of marriage. In disestablishing the debts, it refunded $29,626.26 in payments made and paid $10,500 in compensation for harm caused by wrongful establishment and recovery. In all cases, MSD agreed that the evidence gathered did not support the finding a relationship existed. Very concerningly, one case relied on purported evidence that did not exist, and several cases contained evidence of violence, including medical documents with allegations of rape by the alleged partner. While these results of these reassessments are profoundly positive, they represent a tiny fraction of the thousands of relationship decisions since 1996, of which many will have an alleged debt attached. It is beyond argument that a subgroup of these people, who are likely to be single mothers, will spend years or decades paying off debts that should not exist.
The heart of this issue is how MSD trains, supports, and audits staff in making decisions. Staff are not legally qualified and are expected to use “policy” guidelines when making decisions. While staff have access to an internal staff helpline and departmental lawyers, it is questionable how well these are used. Staff investigating fraud allegations have slightly more training but are not required to undertake any formal legal study or fraud examiner certification. This is concerning as determining conjugal status is complex; it is a careful evaluative exercise requiring a correct understanding of the test, asking the right questions, gathering the right evidence, correctly weighing it, and doing all this in a procedurally fair manner. The ability to do all of this well takes significant knowledge and skill which takes time to develop – I have not seen these determinations done well by many staff. There is also the significant issue of staff simply accepting the proffered existence of a relationship in marriage without checking if the required features exist. I pause to make clear that this is not a criticism of any individual staff, and while I have seen some be rude or less than helpful, I have never seen any act in what I considered to be bad faith. I acknowledge that they do a very difficult and demanding job to the best of their ability.
Given this background, it is vitally necessary for MSD to explain the test to its staff and the public thoroughly, and to have good systems to ensure correct decisions as often as possible. Unfortunately, this is not so. As staff are not legally trained, MSD expects them to follow internal “policy” guidelines when making decisions. These guidelines are somewhat muddled and provide piecemeal, poorly written information, rather than providing a comprehensive decision-making tool. This problem is not unique to these instructions and is broadly a feature of many internal policies. Specific problems include:
- MSD principally direct staff to determine if a “de-facto relationship” exists, which is the term our Act now uses. However, this is just updated language, and the Ruka test still applies. The guidelines are initially silent on this important point but eventually explain it on one page if staff choose to view it, however, the explanation is unclear and the point may well be lost. This is significant because, as explained below, the common understanding of this term and test comes from the significantly different Property (Relationships) Act 1976.
- The guidelines are silent on the required depth of the assumption of financial responsibility, and that this is the principal required feature. Again, this is mentioned on one page if staff choose to view it, which lists joint bank accounts or mutual financial arrangements as examples showing that the required commitment exists. This framing is fundamentally wrong as neither of those features is a sound basis for presuming so. Worse, staff are not instructed to “look behind” the examples to examine the nature of any financial commitment between the parties.
- The guidelines list the Thompson indicia examples to help staff to examine if the required emotional commitment is present. However, the guidelines fail to provide any instructions on weighing any indicators present. For example, the lack of living together must be scrutinised closely and given due weight as it strongly suggests the absence of the required commitment to merge lives in the nature of a de jure married couple. Equally, the length of and parties’ future desire for the relationship ought to attract significant weight, whereas public perception, for example, should not. The lack of instructions of this type is a substantial omission as it lulls non-legally trained staff into the false perception that any one factor is as important as another. Indeed, in my experience, staff struggle to understand this is not so.
- The guidelines contain limited mention of domestic violence and there is no requirement for staff to proactively explore its presence. Instead, they caution staff about the need to carefully assess if the relationship contains the required features if violence is disclosed. Seldom have I seen vulnerable and traumatised victim-survivors of domestic violence actively disclose it to staff.
I have seen many examples of poor decisions, usually with muddled attempts by staff to explain the test. It is little wonder that this happens given the lack of well-written, comprehensive, accurate tools to guide staff on this highly complicated matter. I have equally seen many examples of public confusion. Many believe the test still only turns on physical indicia, the “he stays three nights a week” test, which has not been law since 1993. Many assume the (significantly different) de-facto relationship test for relationship property division applies. MSD has failed to explain the test simply and accurately to enable the public to soundly evaluate if their relationship is in the nature of marriage. It provides the following explanation online, in a brochure and in application forms:
- People [are considered] in a relationship for benefit purposes if they are in a “de-facto relationship.”
- There also needs to be a degree of commitment for the foreseeable future and financial interdependence.
And to help people “decide if they are in a relationship”:
- [They] “should think about whether their relationship includes some of” [the Thompson indicators, specifically including] “shar[ing] money.”
This is quite misleading and does not accurately explain the test because:
- Stating that people “are considered” to be in a relationship if a “de-facto” relationship exists misleads readers to the commonly known but significantly different and incorrect test for relationship property division. It is insufficient to state that the Ruka factors “also need” to be present as this infers they are two distinct concepts when, in fact, these features determine if such a relationship exists – it is incompatible with the closed “are considered” statement. Given these irreconcilable statements and lack of differentiation, the significance is likely lost.
- There is no explanation that “financial interdependence” requires the parties to have formed a willingness to assume responsibility for each other if one has inadequate income to support themselves. There is no explanation that it is the most important of the two required features.
- Asking people to examine if their relationship has “some of” the Thompson physical indicators misleads them that the test is one of strict liability: where, if their relationship has “some of” the indicators then it is made out, and that all factors are equally important. There is no mention of the caution against using this list in a mechanical, tick-the-box approach.
It is particularly concerning that I have seen investigators use this explanation to offer an explanation of the relationship test. Applicants for main benefits must also endorse they understand this ostensible explanation when selecting a conjugal status to declare. I have seen this as the basis of a criminal charge of dishonest use of a document. This poor wording could form a defence of honest belief as it places the defendant beneficiary in a precarious position.
Deficiencies in challenging decisions
Aggravating the effect of any relationship decision is the meagre rate of challenges to decisions, the seriously flawed appeal process, and the almost total lack of legal help. The Act provides a two-step dispute resolution scheme to challenge these decisions. First, internal panel review by a Benefits Review Committee (BRC) and, second, adjudication by a judicial tribunal, the Social Security Appeal Authority (the Authority). These are two distinct steps for which a dissatisfied person must apply separately. Beyond this, appeals to the courts are possible on questions of law. Only around 0.06% of all decisions are formally challenged, which includes relationship decisions. The scant research available articulates reasons for this extremely low rate: lack of awareness of appeal rights, lack of confidence, lack of ability, and fear of repercussion. As one participant aptly put it, “[beneficiaries] would no more [challenge a decision] on their own than fly to the moon.” MSD data shows that in the 17 years from 2003 to 2020, it received just over 1,000 reviews of relationship decisions. This almost certainly represents a small fraction of decisions. For example, during the entire 17-year period, MSD received fewer applications for review than overpayment debts established in one year of fraud investigations between 2014 and 2019. A breakdown of the outcome of these review applications is as follows:
|Withdrawn by applicant||345||33%|
|Overturned in full||230||22%|
|Confirmed to proceed to BRC||498||45%|
|Out-of-time review declined||56||11%|
|Overturned by BRC||50||10%|
|Partially overturned by BRC||34||7%|
|Confirmed by BRC||348||70%|
The internal review step is a significant barrier to effective and just dispute resolution and is almost certainly causative of the very low rate of cases coming before the Authority. Specific problems include:
- It forces those disputing MSD to first deal with the agency they are disputing and have that very agency determine the validity of the dispute (as the panel decides if it has jurisdiction, will accept a late review, or will change its own decision). This is unique among similar schemes; for example, the accident compensation and immigration schemes have independent first-instance decision-makers. Even in the most comparable example of child support disputes, which the Commissioner of Inland Revenue first reviews, the Commissioner contracts external, legally qualified Review Officers, rather than using non-legally trained departmental staff such as those sitting on MSD panels.
- Despite the Supreme Court confirming that panel review is entirely administrative in nature, MSD conducts itself as if the process is adversarial, adjudicative litigation. MSD letters use litigious language and its “submissions” (called “reports”) are often dozens of pages long; both are often of poor quality and lack legal rigour. These features significantly disadvantage an unrepresented lay applicant. MSD also does not make clear that a review is not the substantive step (which would be an appeal to the Authority). Its public information and letters barely mention the possibility of appeal to the Authority and couches the panel in terms and phrases more akin to an independent adjudicator.
- While applicants can attend the panel meeting if they wish, these meetings often lack legal rigour and are very slow to schedule. It is common for them to take many weeks or months to arrange.
- There is no right of removal or appeal to the Authority if MSD (via its panel) decides it has no jurisdiction to determine the review or, if a decision is lodged outside the standard three-month timeframe, MSD refuses hear a late application. Despite couching these matters as decided “by the BRC”, this is just MSD. Where there is an associated conviction, MSD always asserts that res judicata (issue estoppel) prevents the BRC (and thus the Authority) from hearing the matter. It is also common for applicants to lodge reviews outside the three-month timeframe. Occasionally, the issue of whether an application is in fact lodged outside three months arises. These can be complex legal matters and are beyond the ability of non-legally trained panel staff, and indeed, some lawyers. It is grotesque and unacceptable that there is no ability to appeal these decisions. The only remedy is judicial review or an Ombudsman complaint, and both are not ideal. I acknowledge MSD has improved its guidance on whether to accept a late review, but the lack of appeal rights remains dire.
With all the above considered, it is unsurprising that around half of all applicants do not attend review meetings or MSD confirms its original decision approximately 85% of the time. Sadly, no political interest exists in abolishing or significantly reforming this process. A Green Party bill to introduce independent Review Officers at first instance failed in 2009. Despite receiving detailed submissions, a recent “expert” report to the government on welfare overhaul almost completely ignored the dispute resolution scheme, recommending only a “simpler, [faster] and more accessible” system. To date, there has been no progress. A potential solution does, however, exist within the current law. The Act requires two MSD staff sit on panels, which could easily be casually employed Review Officers, ideally with formal legal qualifications. This would be a major improvement, particularly in legally complicated decisions such as relationship status, given that MSD currently has non-legally qualified staff undertaking this function.
Administrative review aside, the substantive right of appeal is adjudication by the Social Security Appeal Authority. Sadly, the number of cases before the Authority is small. Around 1,100 review decisions are eligible for adjudication each year, but only around 150 (or 13%) are appealed to the Authority. Between 1996 and 2022, the Authority decided just 67 relationship in the nature of marriage cases, or an average of just two per year. It quashed 18 findings of such a relationship, or 26%. Of these, 62 were one of the eligible 382 cases between 2003 and 2020, and of these, it quashed 13 (21%). That means only 16% of eligible cases across these 17 years reached the Authority. Despite these low numbers, the Authority has not shied away from criticising the methods of MSD (or a predecessor) in making the decision under appeal or handling the review process. Specific examples include:
- In decision  NZSSAA 62, the Authority heavily criticised the investigator for “most concerning” behaviour; specifically using selective evidence, asking leading interview questions, and making conclusions with no rational basis. For example, the investigator only selected text messages of the appellant that supported his view and determined the alleged partner must have assumed responsibility for the appellant due to her low food spending. He decided the latter “without reference to any recognised standard for measuring food costs.” The Authority quashed the decision and disestablished the $68,726.82 overpayment.
- In decision  NZSSAA 32, while finding the appellant’s relationship was one in marriage, the Authority wrote off the $67712.92 debt on the basis it was not satisfied the investigator had the applied Ruka test or had done so incorrectly when the department first investigated the appellants in 1999. It specifically referred to the fact this decision dated from three years after Ruka became law. This is during the period that Ms Joychild found the Department failed to update any material for staff on the changed test.
- In decision  NZSSAA 90, while finding the relationship was in the nature of marriage, the Authority criticised the conduct of the BRC as “totally unacceptable.” The appellants had applied for review more than three months after being notified of the decision. In such a situation, MSD (via the BRC) must give leave to consider a late review. It had done so and then immediately considered the substantive matter without giving the applicant appellant any time to prepare.
- Finally, concerningly and reflected in its 2017 criticism, in decision  NZSSAA 136, the Authority criticised what it observed as a lack of open-mindedness during the investigation that led to the decision under appeal. The Chair, Judge Middleton, stated:
“We make it the observation that during this week we have heard appeals of this nature rising in different districts. The department has received an allegation from an outside person against the appellant and has thereafter appeared to pursue an investigation with a view to establishing the correctness of that allegation. In neither case did we consider that the department’s investigations have been pursued with the purpose of obtaining a fair decision. It appears to us that in each case the department has accepted documentary evidence which appears to support the allegation and has acted on that without investigating further the explanations given by the respective appellants. We consider that where the department has to exercise a discretion which can result in a very heavy penalty for the beneficiary who loses a benefit, the evidence must be examined with great care and all explanations should be investigated and considered before a decision is made.”
Lack of legal assistance
Unfortunately, a near-total lack of social security legal assistance means that very few New Zealanders can get advice and representation, so even if they do challenge MSD, they must do so alone or with basic advice of questionable quality. While we have 24 Community Law Centres, none specifically specialises in social security law and most lack any knowledge in this area. A small number of centres do good work in this area, but they operate only in particular geographic areas, so it is simply a matter of luck whether someone lives in that area. Even if a centre wanted to develop in this area, most lack the capacity, or a lawyer with sufficient litigation experience, or funding to employ one. There is also the practical problem of the absence of an easy-to-understand, comprehensive and practically focussed textbook or other training material. The funding environment means that the ideal solution – establishing a specialist social security CLC – is fraught with difficulty. CLCs were established in New Zealand in the 1980s and 1990s, with no significant additions since Auckland Disability Law became a stand-alone CLC in 2007. Establishing a new CLC risks diverting already limited funding from existing centres, which is obviously undesirable. Undeterred, however, the possibility of creating some form of specialist service – even initially as a pilot programme within an already established CLC – is something I am actively exploring.
Given this bleak situation, the reality is that most of the small amount of social security work in New Zealand is undertaken by a small group of non-lawyer advocates, either independently or within a formal service. Independent advocates may simply do social security work, for example, as part of budget advice or social work. It is not a reserved area of law in New Zealand meaning non-lawyers can work in it, including at the Authority. However, it is questionable how ideal this current situation is, given that independent advocates have none of the professional obligations of lawyers despite doing essentially the same role. The services available all operate on some form of limited basis, either in terms of time availability, geographic area covered or level of knowledge and skill. Not all services have the level of knowledge and skill required to challenge relationship status or other complicated decisions, particularly at Authority level.
Legal Aid is also no real solution. Authority appeals are one type of proceeding eligible for civil Legal Aid, but in reality, it is nearly impossible to access. There is a severe shortage of lawyers as the low pay rates make this work uneconomic (we do not have public provision of civil Legal Aid). Our Chief Justice recently remarked that the situation is so dire and civil Legal Aid so broken that the system may collapse. Even if a sympathetic lawyer wanted to, it is doubtful they could gain an adequate grasp on this complex area of law.
The conjugal status test plays a significant role in New Zealand’s social security entitlement. Incorrect application of this test can have significant civil and/or criminal liability, including significant debt, which statistics show has a high likelihood of being potentially incorrect. Since the introduction of the current test in 1996, welfare authorities have repeatedly been found to be misapplying the test. Deficiencies in staff training, public information, and the review and appeal process create an environment that places the most vulnerable in society at further disadvantage. Our experience provides a significant warning to Australia on matters requiring anxious scrutiny in this area.
Appendix: Social Security Appeal Authority relationship in the nature of marriage decisions
 LLB (Dist.) Barrister and Solicitor.
 Primarily the Domestic Purposes Benefit (now Sole Parent Support from 15 July 2013). Social Security Act 1964 (NZ), s 63(b) prior to 26 November 2018. On 26 November 2018, the Social Security Act 2018 came into force. Section 8(4) carries over the ability to treat two unmarried persons as married if they have entered a “de-facto relationship,” which Sch 2 defines as having the meaning in the Legislation Act 2019. Section 14(1)(a) of this Act defines such a relationship as meaning one “in the nature of marriage.”
 Thompson v Department of Social Welfare  2 NZLR 369 (HC).
 At 374.
 The Court of Appeal is New Zealand’s intermediate appellate court. It sits either as a full bench (five members including the President in cases deemed significant enough) or a standard divisional bench (of three Justices). The Australian equivalent is a full bench of the Federal Court. Richardson P, Blanchard and Thomas JJ agreeing and Gault and Henry JJ dissenting.
 Ruka v Department of Social Welfare  1 NZLR 154 (CA). On 1 October 1998, the Department’s Benefits and Pensions Division and the Employment Division of the Department of Labour merged into the Department of Work and Income. On 1 November 2001, DWI and the Ministry of Social Policy merged to form the Ministry of Social Development. The Ministry maintains the “Work and Income” name for the unit administering social security benefits.
 Ruka v Department of Social Welfare HC Auckland AP 82/95, 6 December 1995. His Honour Judge Andrew Becroft of the High Court (founding lawyer of Mangere Community Law). Her Honour Karina Williams of the District Court, Patrick Driscoll, Denise Bates QC and Frances Joychild QC (Auckland Women Lawyers Association).
 Report to the Minister for Social Services: Review of [the] Department of Work and Income Implementation of the Court of Appeal decision, 18 June 2001, Frances Joychild QC, at p i.
 At p iii (recommendation [A1]).
 The Ruka Review: How a government department ignored the law and its reluctance to put things right: Tina McIvor Scoop Politics 21 October 2005 at https://www.scoop.co.nz/stories/PO0510/S00182/the-ruka-review-by-tina-mcivor.htm.
 An ex-gratia payment. Not all clients applied for a payment.
 Social Security Act 2018, s 8(4) uses the term “de-facto relationship,” which sch 2 defines as having the meaning in the Legislation Act 2019. Section 14(1)(a) of that Act defines this as being a relationship “in the nature of marriage,” which is the text in the former 1964 Act.
 Excell v Department of Social Welfare  (HC).
 Property (Relationships) Act 1976, s 2D. While this lists many of the same features as Ruka, subs 3(a) states none are mandatory for a “de-facto” relationship to exist, whereas in Ruka, both requisite factors are mandatory.
 Ministry of Social Development “Work and Income: Are you in a relationship?” webpage at https://www.workandincome.govt.nz/on-a-benefit/tell-us/are-you-in-a-relationship.html.
 Crimes Act 1961 (NZ), s 229A.
 Community Law Canterbury (2014) Access to justice for beneficiaries: A Community Law Response, at p 61.
 MSD Official Information Act response of 20 September 2019. MSD established 11,469 overpayments in its fraud investigations between 2014 and 2019. While not all debts will have been for alleged relationship status, a significant portion will be. For each of those years, the number of overpayment debts established is consistently more than the number of reviews of relationship debts for the entire 17-year period between 2003 and 2020.
 MSD will first perform a low-level intake assessment to attempt to correct obviously wrong decisions; it refers to this as an “internal review.”
 An independent, predominantly legally qualified Review Officer (Accident Compensation Act 2001, ss 137-138); Immigration and Protection Tribunal, s 217.
 Child Support Act 1991, ss 90B, 96B, 96Q; Tax Administration Act 1994, s 7.
 Arbuthnot v Chief Executive of the Department of Work and Income  NZSC 55.
 As examples, MSD letters inform a review applicant it will first conduct an “internal review,” and if unfavourable, the matter continues to the BRC “for a hearing,” to which MSD will “send a report.” In the round, this implies the BRC is an external adjudicator. The public webpage refers to the ability to challenge a decision through the “review” process, which may “end up” at the BRC. There is detailed information about the review process but only one small point about the right to appeal to the Authority. This strongly gives a false impression that the BRC is an external body and the substantive process. See Ministry of Social Development, “Review of a Decision” page. By comparison, the Department of Human Services (Australia) “Reviews and appeals of a Centrelink decision” page clearly states the two-step process and provides detailed information about the Administrative Appeals Tribunal.
 Community Law Canterbury, above n 14, at p 98.
 At p 99.
 Welfare Expert Advisory Group (2019) Whakamana Tāngata: Restoring dignity to social security in New Zealand, p 196 at .
 Section 392(1) and cl 2(b) of sch 7.
 The number of published decisions hosted by the New Zealand Legal Information Institute with a primary decision of conjugal status found by searching (separately) for “relationship in the nature of marriage” and “Ruka.” Excluded collateral decisions, for example, the quantum of debt from a correctly found relationship.
 Lawyers and Conveyancers Act 2006, s 27(1)(b)(i).
 “Legal Aid system ‘broken and may collapse’ – Chief Justice” Radio New Zealand, 12 October 2021.