Tranter, Kieran ; Sleep, Lyndal; Stannard, John --- "The Cohabitation Rule: Indeterminacy and Oppression in Australian Social Security Law" [2008] MelbULawRw 21; (2008) 32(2) Melbourne University Law Review 698

THE COHABITATION RULE: INDETERMINACY AND OPPRESSION IN AUSTRALIAN SOCIAL SECURITY LAW

KIERAN TRANTER,^[*] LYNDAL SLEEP^[†] AND JOHN STANNARD^[‡]

[This article argues that the cohabitation rule in Australian social security law is uncertain and has, as a consequence, given rise to an oppressive administrative regime. It tracks the indeterminate nature of the rule as a constant feature throughout its history and argues that this imprecision remains within its current formulation in the Social Security Act 1991 (Cth). Drawing upon basic ideas about the functionality of rules, it is suggested that the administration of an undefined rule should be attended by resistance and challenge. However, the social security regime and the cohabitation rule appear to have been accepted by the community. This acceptance is explained as being the result of the oppressiveness of the current administration. Drawing upon analysis of Administrative Appeals Tribunal decisions and interviews conducted with Centrelink clients, this article argues that the cohabitation rule unfairly targets vulnerable clients, is implemented through the use of invasive surveillance and provides opportunities for intimidation by Centrelink officers.]

CONTENTS

I INTRODUCTION

Australian social security law distinguishes between ‘single’ and ‘partnered’ clients for the purpose of determining payments. While a single client’s payment is dependent on their own income and assets, a partnered client’s payment is calculated on the basis of the income and assets of both the client and their partner.^[1] The result is that payment scales for single clients are greater than those for partnered clients.^[2]

Both de jure and de facto marriages have been used by social security law to distinguish between single and partnered clients.^[3] As a client’s relationship status must be determined ‘objectively’,^[4] it has been necessary for administrators to develop a rule for determining when a relationship between two persons of the opposite sex^[5] could be deemed equivalent to a marriage in law. Within social policy literature, the resultant rule has become known as the ‘cohabitation rule’.^[6] In the Social Security Act 1991 (Cth) the rule operates in two ways. Its first purpose is to determine whether persons who are not de jure married should be treated as a ‘member of a couple’.^[7] Its second purpose is to determine when de jure married or previously de facto persons should be treated as single.^[8] In this second context, it has been established that the rule will allow persons who are ‘separated under the one roof’ to be considered single.^[9]

This article argues that the cohabitation rule is uncertain and that its current administration by Centrelink is oppressive. In Part II it is argued that throughout its history, the cohabitation rule has been marked by indeterminacy due to both its open-endedness and the subjectivity required in its application. This indeterminacy remains notwithstanding 60 years of administration and reform.

In Part III the article draws upon recent Administrative Appeal Tribunal (‘AAT’) decisions and semi-formal interviews with 16 Centrelink clients to argue that the current administration of the rule targets the vulnerable members of society. This administration utilises invasive surveillance and often results in intimidation of clients by Centrelink officers. The article concludes that the rule manifests these characteristics in its administration because the rule itself is undefined.

A Basis for This Article

1 Feminist Criticism of the Cohabitation Rule

This article is predicated on feminist criticism of the cohabitation rule. In Australia, the existence of the rule has been justified according to basic principles of equality between married and unmarried women. Essentially, unmarried women who are cohabitating with a male should not get the benefit of social security payments that are not available for married women.^[10] Unfortunately, this justification is predicated on the highly problematic model of social and economic organisation whereby in a house with a cohabitating male and female, the male is assumed to earn income and to then distribute that income to the female and child dependents.^[11] This basic premise has been heavily criticised. One line of criticism is that this assumption has not been the reality for many women and children in Australia.^[12] Another line of criticism has identified the rule as an attempt to control the lives of women because it subjects women to scrutiny and operates on an assumption that women should be dependent on men.^[13] Following the reduced emphasis on the breadwinner model within social security law since 1995,^[14]

some have considered the rule to be of historical interest only.^[15] Over the last three years, however, the community has raised concerns with the implementation of the cohabitation rule by Centrelink.^[16] These concerns have stimulated a new focus on the administration of the rule and have even resulted in a recent Commonwealth Ombudsman Own Motion Report.^[17]

2 General Theories Regarding the Nature of Rules

This article also takes as its foundation ‘general rules theory’ as it applies to the administration of the cohabitation rule. If one accepts the proposition that the character of the cohabitation rule is indeterminate and therefore causes oppressive administration of the rule, this would mean that the cohabitation rule and its administration operates contrary to Terry Carney’s claim that the law has had a minimal impact on the culture of welfare administration in Australia.^[18] Carney’s writings have focused on public law both as a means of empowering clients and as a mechanism for shaping a more respectful and caring society — a use of law that he regards as ‘at best an empty gesture or, at worst, a delusion’ when it is unaccompanied by wider support for change.^[19]

In contrast, this article focuses on the ability of law to provide the foundation for an oppressive administrative regime.

In making the claim that the form of the written law informs its administration, this article draws upon some basic theories regarding the expression and function of rules within a society. Writing after the Second World War, Lon L Fuller grappled with what he saw as the failure of law as posited power in authoritarian regimes.^[20] Drawing on American legal realism of the 1930s, he argued that ‘one starts with the obvious truth that the citizen cannot orient his conduct by law if what is called law confronts him merely with a series of sporadic and patternless exercises of state power.’^[21]

As a basic proposition, Fuller suggested that legal statements should clearly set out the limits of permissible conduct.^[22] He believed that in order for law to fulfil its social function (informing rational beings of what constitutes appropriate conduct), law must involve the application of simple criteria. According to Fuller, law should consist of a body of rules that are clearly known and capable of being understood. Thus, law should be internally consistent, able to be complied with, not continuously changing, not retrospective in its application and, finally, it should provide the actual basis for a decision-maker’s decision.^[23] Fuller’s thesis allows for criticism of a legal proposition, not because its content fails to measure up to political or ethical ideals, but because it does not measure up to this ‘inner morality of law’.^[24]

The consequence for authorities which produce and administer rules that do not comply with the ‘inner morality of law’ is, Fuller suggests, a lack of respect for the authorities and, in extreme cases, popular rejection of the regime.^[25] However, Fuller acknowledged that this would only be the response in a society where people are suitably empowered to defend their rights from irrational or arbitrary officials. He provided an alternative hypothesis to explain the situation where an indeterminate rule did not generate opposition and in this Fuller was in agreement with his mid-century interlocutor H L A Hart.^[26]

Both suggested that a veneer of acceptance can be maintained by officials through oppressive conduct that coerces compliance and silences dissent.^[27]

It is these two hypotheses concerning the relationship between an indeterminate law, its administration and the popular response to its administration which inform the argument below.

II THE INDETERMINACY OF THE COHABITATION RULE IN SOCIAL SECURITY LAW

This Part argues that throughout its existence, the cohabitation rule has been marked by indeterminacy. The scope of what a decision-maker should consider in order to determine whether a de facto relationship exists has been open-ended. From its origins in women’s payments to its legislative enactment in the 1970s and the current marriage-like relationship criteria, the rule has been criticised because it has allowed a decision-maker’s subjective evaluations to affect decisions. Parallel with criticism concerning indeterminacy, there has also been criticism that the administration of the rule has been oppressive.

A 1900–70: The Emergence of the Rule

The situation of de facto couples presented difficulties at the very origins of the Australian social security system. The phrase ‘good moral character, and … led a sober and reputable life’ in the Old-Age Pensions Act 1900 (NSW)^[28] was criticised in the Commonwealth Royal Commission on Old-Age Pensions in 1906 for providing administrators with significant discretion in regards to paying the pension to a member of a de facto couple.^[29] While the subsequent federal Act omitted ‘good moral character’ and ‘sober and reputable life’, it retained the phrase ‘good character’.^[30] This meant that the legislation still empowered the decision-makers with a discretion to refuse a claim, grant a claim at a reduced rate or cancel a pension for a member of a de facto couple.^[31]

The precursor to the current cohabitation rule was contained in Part III of the Widows’ Pensions Act 1942 (Cth), which provided payments to civilian widows whose husbands had either died or had deserted them.^[32] It also provided that a woman who had lived with a man as a de facto wife was also eligible for the pension.^[33] The term ‘de facto’ was defined as a woman who was maintained by a man ‘and, although not legally married to him, lived with him as his wife on a permanent and bona fide domestic basis’.^[34] In 1943 an amendment was passed which, while leaving the wording of the provision otherwise unchanged, replaced the phrase ‘de facto’ with the phrase ‘dependent female’.^[35]

In this way, the payment essentially recognised de facto relationships as the same as de jure marriage. This recognition was significant as it allowed women who otherwise would not be eligible for the payment to receive it.

However, the way in which the Widows’ Pensions Act 1942 (Cth) responded to the situation of a pensioner who entered into a de facto relationship created an anomaly for the Department of Social Security.^[36]

The Social Security Act 1947 (Cth) reproduced the eligibility requirements for the Widows’ Pension but provided no guidance concerning on-going eligibility of a cohabitating pensioner.^[37] The result of this anomaly was that those who were responsible for determining the validity of a client’s pension claim were empowered with discretion and subjectivity. Writing about departmental decisions of the 1950s and 1960s, Alan Jordan discovered that the existence of a de facto relationship was dealt with by the Department in a variety of ways. It was intermittently ignored; used as evidence of de jure marriage; treated as grounds for cancelling the pension; used as a reason for Centrelink to demand the repayment of wrongly paid moneys; and used as evidence in criminal prosecution for welfare fraud.^[38]

Jordan suggested that the different responses depended on the perceived moral culpability of the client in making false statements about the relationship.^[39]

The absence of a legislative basis for terminating the Widows’ Pension on the grounds of cohabitation caused disquiet within the Department during the 1960s,^[40] and within the community ‘[t]ales were told of suspicious field officers checking cupboards and beds for evidence of male presence.’^[41] The introduction of the Supporting Mothers’ Benefit in 1973^[42] provided the opportunity to remedy the rule’s legal status.^[43] The new provisions provided that in order for a woman to qualify as a supporting mother, one of the requirements was that she was ‘not living with … a man as his wife on a bona fide domestic basis although not legally married to him’.^[44] Responding to criticism by the Commission of Inquiry into Poverty,^[45] this phrase was inserted into the definition of ‘widow’ for the purposes of the Widows’ Pension in 1975.^[46] The purpose of the phrase was to ‘give specific authority for the exclusion of a woman living with a man on a bona fide domestic basis’.^[47] This meant that by 1976 the Department had the authority to cancel payments on the basis of cohabitation.

In 1975, Minister Francis Stewart told Parliament that cohabitation inquiries ‘must be conducted with tact and discretion.’^[48] According to Jordan, by the 1970s the Department had developed guidelines concerning how to investigate alleged cohabitants. The guidelines provided a list of non-conclusive criteria concerning sexual relations and financial arrangements.^[49]

In 1975, the Department issued a public policy statement listing seven factors to be considered when examining a client’s circumstances: co-residence; type of residence; how household duties compared with a husband–wife relationship; pooling of financial resources; how the ‘couple’ spent their leisure time; whether they presented as married; and whether they shared any children.^[50] This policy also directed officers not to question claimants about their sexual relations; to undertake inquiries with sensitivity to ‘avoid the possibility of subjecting a person to a humiliating and distressing experience’; and to ensure that house inspections only be carried out with the permission of the client.^[51] According to Jordan, the 1975 policy formalised the Department’s previous practice.^[52]

However, Mary Jane Mossman was of the opinion that the policy diminished the emphasis on sexual relations and promised greater ‘uniformity in Departmental decisions.’^[53]

Mossman’s optimism was short-lived. In 1976, the ‘Baxter’ case demonstrated that the Department, even after release of the policy, continued to place emphasis on the existence of a sexual relationship and the taking of the man’s surname as proof of cohabitation. This emphasis, and the resultant finding that Mrs Baxter was cohabitating with Mr Baxter, ignored alternate explanations including the possibility that some of the actions, particularly adopting her housemate’s name, were taken to hide the client’s address from her abusive ex-husband.^[54] Furthermore, officers humiliated the client by making unannounced visits to her home, questioning her directly about her sexual conduct and threatening her with ‘a less friendly visit’ from departmental inspectors if she did not sign a statement admitting cohabitation.^[55] These concerns with the administration of the rule were shared by other commentators examining cohabitation investigations during the late 1970s.^[56] Indeed, Ronald Sackville criticised the post-1975 cohabitation policy decisions:

There is undoubtedly some discrepancy between the officially designated procedures and what happens in fact … Moreover, the instructions are so vague on some matters, particularly the relevance to the cohabitation issue of sexual relations between the couple under investigation, that it would be surprising if significant variations in practice did not occur.^[57]

This failure of policy to limit the amount of discretion available to departmental officers in applying the rule, as well as the capacity for the rule to expose clients to humiliation by departmental officers, led to three reform proposals. The first was the codification of the criteria used to determine the existence of a de facto relationship in legislation.^[58]

The second proposal was that the rule focus solely on financial considerations.^[59]

Finally, the third proposal suggested exploring new merits and judicial review avenues as a way of challenging the use of departmental discretion.^[60]

B 1980s: Living with a Man as His Wife on a Bona Fide Domestic Basis

By 1976, the cohabitation rule had been codified in legislation as part of the eligibility for Supporting Mothers’ Benefits^[61] and Widows’ Pensions.^[62] In 1977, Supporting Mothers’ Benefits were extended to sole fathers and renamed Supporting Parents’ Benefits.^[63] In April 1980, the AAT was given jurisdiction to hear social security appeals^[64]

and the first AAT cohabitation decision, Waterford v Director-General of Social Services (‘Waterford’),^[65] was decided in November 1980. Waterford appeared to endorse Mossman’s suggestion that financial considerations were paramount.^[66]

However, this reading of Waterford was rejected by the Federal Court of Australia in Lambe v Director-General of Social Services (‘Lambe’).^[67] The AAT in Lambe read Waterford as providing for a holistic investigation to see

whether that relationship contains any of the indicia of a family unit; and … the question of financial support provided to a woman will be an important consideration but it is only one of a number of relevant matters …^[68]

The Full Federal Court approved of this reading and suggested that:

there is nothing in the definition of ‘supporting mother’ to indicate that the amount of financial support he provides is of crucial significance. … [I]t is far from being the only incident, and it would be impossible these days to see it as a sine qua non of such a relationship or such a domestic circumstance.^[69]

Lambe was expressly followed in Lynam v Director-General of Social Security (‘Lynam’), in which Fitzgerald J emphasised that:

What must be looked at is the composite picture. Any attempt to isolate individual factors and to attribute to them relative degrees of materiality or importance involves a denial of common experience and will almost inevitably be productive of error. The endless scope for differences in human attitudes and activities means that there will be an almost infinite variety of combinations of circumstances which may fall for consideration.^[70]

The approach that emerged from Lambe and Lynam was an open-ended examination of a client’s circumstances, where no one factor is determinative. In Re Tang and Director-General of Social Services (‘Tang’), the AAT set out a list of criteria that could assist in the assessment of a relationship.^[71] However, the ‘Tang criteria’ were presented as non-conclusive considerations that were not to distract a decision-maker from considering the relationship as a whole.^[72] This understanding was emphasised by the Federal Court in Staunton-Smith v Department of Social Security (‘Staunton-Smith’).^[73] Whilst this case was decided in 1991, the decision concerned the phrase in the 1947 Act: ‘living with a man as his wife on a bona fide domestic basis, although not legally married to him’.^[74] After recasting the Tang criteria in the context of determining whether a married couple were separated, O’Loughlin J added:

It is not suggested that this list is exhaustive nor will each of these subjects fall to be considered in every case. It must also be emphasised that a particular answer to a single subject will rarely, if ever, supply a final solution. The responsibility of the fact-finding Tribunal is to have regard to all the material facts of each case, treating the matters listed above only as indicators.^[75]

For the period 1980–90, the AAT and the Federal Court were of the view that the decision was to be made with discretion, having regard to open-ended criteria rather than determined by any preconceived indicia. It was to be a flexible approach respecting ‘common experience’.^[76] In substance, both the AAT and the Federal Court gave approval to the Department’s 1975 policy.^[77]

Feminist critics continued to challenge the rule’s subjectivity, the invasiveness of its administration and its discriminatory effect.^[78]

In response to criticisms of open-endedness and subjectivity as well as earlier calls for a statutory criteria,^[79] Parliament enacted the Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth). In doing so, it was acknowledged that ‘[w]ithout any clear legislative provisions staff ha[d] been left to administer a program without clear rules and no clear procedures. The results ha[d] necessarily been intensive investigation and arbitrary decision making.’^[80]

This amendment meant that after 1 January 1990 the cohabitation rule was formulated as follows:

In forming an opinion about the relationship between 2 people for the purposes of the definition of ‘de facto spouse’ … the Secretary is to have regard to all the circumstances of the relationship including, in particular, the following matters:

(a) the financial aspects of the relationship, including:

(i) any joint ownership of real estate or other major assets and any joint liabilities; and

(ii) any significant pooling of financial resources especially in relation to major financial commitments; and

(iii) any legal obligations owed by one person in respect of the other person; and

(iv) the basis of any sharing of day-to-day household expenses;

(b) the nature of the household, including:

(i) any joint responsibility for providing care or support of children; and

(ii) the living arrangements of the people; and

(iii) the basis on which responsibility for housework is distributed;

(c) the social aspects of the relationship, including:

(i) whether the people hold themselves out as married to each other; and

(ii) the assessment of friends and regular associates of the people about the nature of their relationship; and

(iii) the basis on which the people make plans for, or engage in, joint social activities;

(d) any sexual relationship between the people;

(e) the nature of the people’s commitment to each other, including:

(i) the length of the relationship; and

(ii) the nature of any companionship and emotional support that the people provide to each other; and

(iii) whether the people consider that the relationship is likely to continue indefinitely; and

(iv) whether the people see their relationship as a marriage-like relationship.^[81]

The reasons given for the insertion of this list of criteria was to increase certainty in decision-making, so as to reflect the decision in Lambe.^[82] Notwithstanding this, the Minister for Social Security, the Honourable Bob Howe, did note that the criteria (a)–(e) constitute a ‘non-exhaustive’ list and that when making a decision, ‘in the end, the Secretary must look at the whole relationship.’^[83] Hence, whilst the amendment introduced statutory criteria, it did not produce a change in cohabitation determinations, as the decision-makers continued to look at the whole relationship rather than be strictly guided by the criteria. At the time, this disparity was passed over by commentators.^[84]

The absence of commentary might be due to the controversial nature of the other amendment introduced by the Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth). In addition to the statutory criteria, this amending Act also introduced s 43A.^[85]

This section was a ‘reverse onus’ provision that authorised an automatic investigation and review of a sole parent’s eligibility if they shared a residence with a person of the opposite sex for a period of eight weeks or more and another ‘triggering factor’ was present.^[86] The introduction of s 43A attracted strong criticism as it targeted women. This is possibly the reason why the introduction of the aforementioned criteria went relatively unnoticed.^[87]

C 1990–2007: Marriage-Like Relationship Criteria

The Social Security Act 1991 (Cth) reproduced the statutory criteria as s 4(3), but replaced ‘de facto’ with the phrase ‘marriage-like relationship’.^[88] The reverse onus provisions were reproduced as ss 4(4) and (5), but were limited to sole parents until their repeal in 2000.^[89] Whilst s 4(3) has remained unchanged, two amendments over the 1990s have changed the administrative impact of the rule.

The first amendment was a series of changes during 1994 that paradoxically reduced the traditional ‘breadwinner’ focus of the Act, whilst at the same time expanded the application of the cohabitation rule. The amendments of 1994 also involved a ‘de-gendering’ of the Act through the alignment of male and female pension ages,^[90] a targeted withdrawal of the Widows’ Pension,^[91] and the paying of Partner Allowance and Parenting Allowance directly to an eligible partner of a client receiving income support.^[92] These changes addressed the substance of the feminist criticisms of the male breadwinner model as they meant that dependents were now paid directly. However, the effect of the changes was to universalise the rule. The issue of whether a client was partnered caused some payments, which originally resulted in a higher rate, to result in a reduced rate if the new requirement was satisfied.

The second amendment occurred in 1995 when the phrase ‘is living’ in s 4(2)(b)(i) was replaced with the phrase ‘has a relationship’.^[93] When this amendment was introduced, it was explained to be a change aimed at preventing the Sole Parent Pension from being paid to a client who was temporarily apart from their partner.^[94]

By removing ‘living together’, the amendment expanded the range of relationships that could be considered marriage-like.^[95]

Whilst s 4(3A) excluded the finding of a marriage-like relationship where persons were living ‘separately and apart from the partner on a permanent or indefinite basis’, the amendment expanded the rule to cover persons whose relationship could be located within the grey zone between ‘living together’ and living ‘separately and apart’ on a permanent basis.^[96]

Although these changes expanded the scope of the rule, the exercise of the decision-making power using the criteria in s 4(3) still reflected the earlier decisions made before the introduction of the criteria. The AAT continued to hold that an opinion formed by an officer using the s 4(3) criteria ought to involve an examination of the relationship as a whole, having regard to, although not determined by, the statutory criteria. In Re Cahill v Department of Family and Community Services (‘Cahill’),^[97] it was suggested that:

it is possible a decision-maker might decide the individual is a member of a couple even though she does not satisfy all or even the majority of the criteria. Conversely, many of the indicia … might be present yet the circumstances as a whole might justify a conclusion that the couple live separately and apart, albeit under one roof.^[98]

A glimpse of how this was administered can be seen in Meredith Wilkie’s study in 1993. She concluded that, due to ‘the complexity and uncertainty surrounding the cohabitation rule, arbitrariness in decision-making is widespread … The decision ultimately, however, will always be subjective.’^[99]

More recent evidence supporting Wilkie’s conclusion can be seen in Table 1, which details the Authorised Review Officer (‘ARO’) decisions which have overturned primary cohabitation decisions.^[100]

Table 1

: ARO Cohabitation Decisions and Total Change Rates 2002–06

14

03–04

39^[105]

32^[106]

7

02–03

40^[107]

29^[108]

11

Table 1 shows that ARO cohabitation reviews have resulted in a higher number of original decisions being overturned compared to the average. Although this data comes 15 years after Helen Goodman’s research, the evidence of a higher success rate for cohabitation appeals confirms her data from 1988–90 concerning the rates at which the Social Security Appeals Tribunal (‘SSAT’) set aside departmental decisions.^[109]

In a recent report, the Commonwealth Ombudsman explains the high success rate of cohabitation appeals as being a product of the rule being more open to subjectivity in the decision-making process as compared with other types of social security decisions.^[110]

This subjectivity is also evident in the comments of Centrelink Customer Service Officers (‘CSOs’) regarding cohabitation decisions. In a survey conducted by the Australian National Audit Office, CSOs expressed the view that they were ‘not comfortable in asking’ personal questions of clients.^[111] Furthermore, ‘most CSOs interviewed believed that they were unable to identify MLRs [marriage-like relationships] unless the customer admitted to it.’^[112] In line with Wilkie and Goodman’s findings, the ARO change rates and the disclosures made by the CSOs in the interviews provide strong empirical evidence that, from 1990, the cohabitation rule was open-ended and subjective.

However, the 2006 Federal Court decision in Pelka v Department of Family and Community Services (‘Pelka’)^[113] challenged the operational understanding behind the administration of the rule. Pelka was the first Federal Court decision on the rule as it is currently expressed in ss 4(2) and (3) of the Act. The decision is important for two reasons. At first instance, the AAT held that the phrase ‘any significant pooling of financial resources especially in relation to major financial commitments’ in s 4(3)(a)(ii) was satisfied by financial cooperation.^[114] On appeal to the Federal Court, French J determined that the AAT was in error in holding that financial cooperation in the form of a barter system was sufficient evidence to constitute pooling of resources toward major financial commitments.^[115] Rather, French J held that, by virtue of the requirement that the pooling of the resources be for the purpose of ‘major financial commitments’, s 4(3)(a)(ii) ‘plainly involves something more than financial cooperation’.^[116] This finding in Pelka clarified the types of evidence that decision-makers can consider under s 4(3)(a)(ii). In AAT decisions after Pelka, joint contributions to rent and the joint purchase of a motor vehicle have been held to satisfy s 4(3)(a)(ii).^[117] However, a finding that one party had purchased a motor vehicle for the other’s use has not been considered ‘financial pooling’.^[118] Likewise, the joint purchase of a washing machine has been held not to be a ‘major financial commitment’.^[119]

While the Federal Court’s decision on this point provided greater clarity to the criterion in s 4(3)(a)(ii), Pelka is significant for a second reason — it has challenged the emphasis to be given to the criteria in s 4(3). In the Federal Court, French J stated that a decision-maker:

(1) Must have regard to their interpersonal relationship as a whole not limited by the factors listed in s 4(3).

(2) Must have regard to each of:

(a) the financial aspects of the relationship;

(b) the nature of the household;

(c) the social aspects of the relationship;

(d) any sexual relationship between the people; and

(e) the nature of the people’s commitment to each other.

(3) In having regard to the preceding five matters, must have regard to all factors relevant to each and, in particular, must have regard to the factors listed under each heading in s 4(3).

(4) Must specifically consider the total picture of the relationship created by all of these factors bearing in mind that consideration must be given to those which weigh against a marriage-like relationship and those which weigh in favour of it.

(5) Must undertake the preceding consideration bearing in mind that a marriage-like relationship is not disclosed solely by any one of the following matters:

(a) financial cooperation;

(b) cohabitation;

(c) a sexual relationship;

(d) cooperative household arrangements; or

(e) mutual commitment.^[120]

French J’s use of mandatory language in framing these factors suggests that the correct approach involves an examination of each and every factor listed in s 4(3).^[121] Indeed, recent commentators, drawing upon this statement and other comments by French J, have argued that Pelka presents a more structured interpretation of s 4(3).^[122] For instance, in Marei v Department of Employment and Workplace Relations (‘Marei’) Barnes FM emphasised French J’s use of the word ‘must’ when considering the s 4(3) criteria.^[123] In Marei, the AAT decided that care for children was an ‘essential element of a marital relationship.’^[124] On appeal to the Federal Magistrates Court, Barnes FM decided that on this point the Tribunal was in error. Rather,

the Tribunal is not obliged to have regard to each of the matters in the paragraphs of s 4(3) or to the factors listed under each heading in s 4(3) that, in itself, amounts to a misconception of the law. The Tribunal ‘must’ have regard to such matters and factors.^[125]

Pelka and Marei established that the approach previously adopted by Centrelink and the AAT, which were guided by earlier authorities that were not concerned with statutory criteria, needed to be tightened. Furthermore, they established that the proper process for making a decision under s 4(3) ought to involve a structured investigation of each and every criterion in s 4(3). Whilst this process does not preclude consideration of other unlisted factors, the final determination must account for the factors both for and against a finding that a couple is cohabitating. This change appears to be reflected in the latest updates to the Guide to Social Security Law (‘the Guide’) which is used in the daily administration of the Act.^[126]

The Guide now specifies, under the heading of ‘Factors to be considered for investigating marriage-like relationships’, that:

Making a determination that a person is a member of the couple requires that the indicators for a marriage-like relationship outweigh the indicators that the person is not in a marriage-like relationship. … All 5 factors must be considered. However, no single factor should be seen as conclusive and not all factors need to be present.^[127]

This emphasis on decision-making structured around the statutory criteria might suggest that Pelka provides a remedy to the indeterminacy which had plagued the rule since its inception. However, on closer examination, such optimism does appear short-lived. Indeed, in Pelka itself, French J was of the opinion that:

The judgment to be made is difficult and, once out of the range of obvious cases falling within the core concept of ‘marriage-like’, will be attended by a degree of uncertainty. Indeed, it may be that different decision-makers on the same facts could quite reasonably come up with different answers.^[128]

The decisions in Pelka and Marei, as well as the recent changes to the Guide, are minor reforms.^[129] The rule itself still requires consideration of all ‘the circumstances of the relationship’, as the five criteria do not provide a checklist of factors that must be present for a finding of a marriage-like relationship.^[130] Furthermore, the type and form of evidence required for a decision about financial matters, social matters, the household, sexual relations and subjective commitment, remain open-ended. Although the Guide does give examples of evidential questions that should be asked in relation to each of the criteria,^[131]

it does not limit the sources of evidence. There is also a danger that a structured approach that requires investigation, evidence gathering and assessment of all criteria might increase the level of invasiveness.^[132]

D Consequences of an Indeterminate Rule

In summary, the cohabitation rule in Australian social security law has been consistently characterised by open-endedness and indeterminacy. Throughout each step of the rule’s evolution there were concerns that the rule was open-ended and that it allowed for subjectivity in decision-making. The second result, indeterminacy, also emerged out of this open-endedness. The concern with indeterminacy, articulated by both critics and law-makers, arose from anxieties surrounding the appropriateness of an open-ended and subjective rule in the context of determining relationship status. It is not beneficial for the social security system to have a rule which allows for such a large degree of discretion in the decision-making process. The rule is restrictive and has invariably been applied to reduce the amount of payment. Furthermore, the implementation of the rule necessarily involves intense scrutiny of highly personal matters. These long running problems with the rule, clearly observable throughout its history, can explain the long process of its formalisation, initially as a term in legislation and then later as a rule accompanied by detailed criteria.

However, attempts to reduce the rule’s indeterminacy have only ever been partially successful. In 1973, although it was made clear that a woman ‘living with a man on a bona fide domestic basis’ was precluded from receiving Supporting Women’s Payment, the content of ‘bona fide domestic basis’ was left to the discretion of the Department. The 1975 policy provided factors to be considered but was only framed as a guide — an approach that was adopted by the Federal Court in Lambe, Lynam and Staunton-Smith. Although the criteria were codified in legislation in 1990, this did not change the initial approach to cohabitation decisions. It was not until the decision in Pelka in 2006 that decision-makers were reminded that the criteria in s 4(3) were now statutory criteria and could not be ignored if it was contrary to the decision-maker’s general reading of the relationship as a whole. However, even if Pelka introduced increased structure into the decision-making process, the indeterminacy of the rule remained. Post-Pelka, there were still problems in the administration of the rule, such as the probative value of any evidence collected, the weight to be given to some forms of evidence as opposed to others, the weighting of each criterion and the fact that decision-makers could still consider factors outside of the statutory criteria. The rule remains, as French J expressed in Pelka, ‘attended by a degree of uncertainty.’^[133]

Fuller suggested that an indeterminate rule, particularly one such as the cohabitation rule, which affected a person’s material wellbeing and exposed them to investigation and judgements about their person, would be met with hostility. He anticipated that such laws would generate political pressure for change and result in noncompliance and the challenging of decisions.^[134]

With regard to the cohabitation rule, whilst there have been moments of political lobbying against the rule, it has not been a contested site of Australian parliamentary politics. The introduction of the statutory test in 1973 was supported by both sides of Parliament, and the subsequent changes in 1975, 1989 and 1994 were presented as technical amendments that generated little debate. From the perspective of those experiencing the harsh administration, aside from some specific individuals such as Ms Pelka who challenged the decision made against them, clients whose relationships have been interrogated and whose payments have changed appear accepting of the Department’s decisions. (This is despite the fact that cohabitation decisions are often changed when reviewed, as shown above in Table 1.) The empirical basis for this claim is found in two sets of data about the rate at which Centrelink clients challenge findings made against them.

The first set of data refer to the percentages of Centrelink clients who challenge cohabitation decisions through merits review.^[135]

Merits review involves three tiers,^[136]

each tier is mandatory and there are no application fees for Centrelink clients.^[137] The first tier is the ARO’s decision. If a client is dissatisfied with the ARO’s decision, then they can appeal to the second tier, the SSAT.^[138] If the client or Department disagrees with the SSAT, either can appeal to the AAT.^[139] Table 2 compares the relationship between cohabitation appeals and total appeals. Table 3 compares the progression of cohabitation appeals and the total number of appeals within the appeals system.

Table 2

: Comparison of Total Appeals with Cohabitation Appeals for ARO, SSAT and AAT 2002–06

/

37 38[142]

4.8

367^[143]

/

753[144]

4.9

56^[145]

/

148[146]

3.8

04–05

1528^[147]

/

34 65[148]

4.4

313^[149]

/

762[150]

4.1

72^[151]

/

173[152]

4.2

03–04

1439^[153]

/

37 01[154]

3.9

313^[155]

/

782[156]

4.0

49^[157]

/

188[158]

2.6

02–03

1212^[159]

/

38 04[160]

3.2

275^[161]

/

919[162]

3.0

37^[163]

/

170[164]

2.2

MLR = cohabitation decisions (‘marriage-like relationship’)

All = total appeals in period

Table 3

: Progression of Cohabitation and Total Appeals 2002–06

MLR = cohabitation decisions (‘marriage-like relationship’)

All = total appeals to tier in period

Table 2 shows that cohabitation decisions appear to amount to a consistently small percentage of appeals at each tier. Whilst the cohabitation rule applies to all Centrelink clients, Table 2 shows that less than five per cent of appeals involve the questioning of a cohabitation determination. This suggests an acceptance by clients of Centrelink’s administration of the rule, notwithstanding its identifiable indeterminacy. Table 3 adds weight to this suggestion.^[165] It identifies a general pattern in which the rate of cohabitation appeals is generally less than the total rate of appeals. Again, the fact that the rule is not generating inordinate numbers of appeals suggests that it is being accepted by clients.

The second set of data, which confirms that Centrelink clients tend to accept the application of the cohabitation rule to their case, is the prosecution rate arising out of cohabitation decisions. Clients who have been ‘caught’ cohabitating can be prosecuted for fraud.^[166] Table 4 examines prosecutions by the Commonwealth Director of Public Prosecutions (‘CDPP’) arising from cohabitation decisions between 2002 and 2006.

Table 4

: Cohabitation Prosecutions 2002–06

96.6^[172]

96.7^[173]

95.9^[174]

Conviction Rate

(%)

95.7

100.0

98.2

98.0

Dismissed

3

0

2

2

Convicted

88

105

111

98

Prosecuted

92

105

113

100

Year

2005–06

2004–05

2003–04

2002–03

The consequences of a conviction for clients can be a criminal record, a fine or imprisonment.^[175]

Whilst the exact number of clients who have pleaded guilty to fraud relating to a cohabitation decision is unknown, it is reasonable to assume that the general rate for convictions from a plea of guilty applies to cohabitation prosecutions.^[176] This being so, Table 4 suggests that even when it comes to the serious consequences of a criminal conviction, most clients do not challenge the assessment of their relationship.

Tables 2, 3 and 4 present evidence against Fuller’s first hypothesis that an indeterminate rule will invariably lead to resistance and challenge. One conclusion from this data is that Centrelink is doing an excellent job of administering the rule and particularly in mitigating the rule’s open-endedness. However, this article does not necessarily agree with this conclusion. Rather, it accepts Fuller’s alternative hypothesis that a veneer of acceptance can be maintained through coercion and oppression. Through examining the history of the cohabitation rule, it can be seen that its administration has often generated claims of invasiveness and intimidation. As early as the 1960s, there were concerns with the manner in which officers sought to locate a male’s presence in a widow pensioner’s house. More recently, concerns have been voiced about the intensity of surveillance in cohabitation investigations.^[177]

This suggests that the ‘acceptance’ of the rule by clients is not genuine — rather, their acceptance is coerced by the oppressive administration of the rule by Centrelink officers, illustrating Fuller’s alternative hypothesis. Drawing upon contemporary evidence of Centrelink cohabitation decisions, the next Part of this article will argue that the current administration of the rule is oppressive for three reasons. These are that the rule targets the vulnerable, that it uses invasive surveillance techniques and that it allows for intimidation by Centrelink officers in the investigative process.^[178]

III OPPRESSION IN CONTEMPORARY COHABITATION ADMINISTRATION

This Part examines the contemporary administration of the cohabitation rule, drawing upon two data sources. The first source was a sample of 79 AAT decisions concerning cohabitation for the period 2005–07.^[179] Table 5 sets out the client’s gender and the nature of the payment. Although 79 decisions were examined, a number of decisions involved joined parties (including both members of the alleged couple), meaning that 88 clients were represented in the sample.

Table 5

: AAT Cohabitation Decisions 2005–07 by Gender and Payment

The AAT decisions documented Centrelink’s decision before the appeal. In analysing the AAT decisions, what was important was not the reasoning process of the AAT or the weight given to the evidence before it, but rather the parts of the decision detailing how the client came to Centrelink’s attention and the glimpses of the client’s life provided within the written reasons.^[181]

The second data source was 16 semi-structured interviews that were conducted in late 2006 with Centrelink clients who had been involved in a cohabitation decision. Participants were recruited through the National Welfare Rights Network (‘the Network’) and identified as Centrelink clients who had approached a member of the Network for advice or assistance regarding the cohabitation rule in 2004–06. Some participants had appealed or had appeals pending with the AAT or SSAT, while others had just requested information without appealing the decision. Two participants knew that they had been referred to the CDPP and one was being prosecuted. Interviews were conducted in person or, for regional and remote participants, over the telephone. The interview schedule focused on the participants’ experiences of the cohabitation investigation and decision-making process.^[182] For the purposes of locating the participant’s story in relation to the administrative history of their case, participants were asked to give consent for the researchers to examine the participant’s file, as held by the Network. This file analysis revealed basic demographic and payment details, and also allowed triangulation with the participant’s interview to pinpoint, where possible, how the participant’s experience fitted within the different stages of the decision-making process. Details of participants by gender, age and type of payment are provided in Table 6. Table 7 summarises the frequencies of age, gender and payment type.^[183]

Table 6

: AAT Participants by Age, Gender and Payment Type

Amy

45–46

Female

Parenting Payment (Single)

Arthur

40–45

Male

Disability Support Pension

Elizabeth

75+

Female

Aged Pension

Isabel

65–70

Female

Aged Pension

Jay

25–30

Female

Parenting Payment (Single)

Jocelyn

25–30

Female

Parenting Payment (Single)

Julia

50–55

Female

Disability Support Pension

Mary

30–35

Female

Parenting Payment (Single)

Meredith

30–35

Female

Parenting Payment (Single)

Rachel

40–45

Female

Parenting Payment (Single)

Regina

60–65

Female

Disability Support Pension

Rosie

50–55

Female

Carers Payment

Sabrina

50–55

Female

Newstart Allowance

Verdant

55–60

Female

Widow’s Allowance

Violet

60–65

Female

Disability Support Pension

Zoe

50–55

Female

Parenting Payment (Single)

Table 7

: Frequency of Age, Gender and Payment Type

Together, the AAT decisions and the interviews provide a window into the contemporary administration of the cohabitation rule from the perspective of clients. From this vantage point, the rule is revealed as an oppressive regime that targets vulnerable clients, provides grounds for an invasive administrative culture and creates the space for client intimidation by Centrelink officers.

A Targeting the Vulnerable

It can clearly be concluded from the AAT decisions and the interviews that notwithstanding its gender-neutral expression, the cohabitation rule targets women. Seventy-three per cent (64/88) of clients involved in the AAT decisions and 93.8 per cent (15/16) of the participants in the interviews were women. Furthermore, 62.5 per cent (55/88) of all clients in the AAT decisions were women either with sole responsibility for children under 16 years of age and therefore eligible for the Parenting Payment (Single), or were suffering a significant disability and hence eligible for the Disability Support Pension. This divide was mirrored in the interviews, where 68.8 per cent (11/16) of interviewees were receiving either the Parenting Payment (Single) or Disability Support Pension. Caring for young children or living with a disability imposes a considerable burden on an individual’s life, as it requires significant resources whilst at the same time placing limits on their ability to earn money. This means that individuals receiving the Parenting Payment (Single) or Disability Support Pension are more vulnerable members of society than able-bodied persons without dependants.

The tendency of the rule to target vulnerable women is explicit in a new aspect — the ‘new child review’ policy, which requires that a woman on the Parenting Payment (Single) who gives birth be automatically reviewed.^[186] A woman is subjected to such a review on the basis of the assumption that if a child has been born, then there is likely to be a father present who may be secretly cohabitating with the woman. The notion that the rule targets vulnerability can be expanded through a closer examination of the AAT decisions. The decisions do not support the conservative cultural narrative — that women with children hide the existence of a cohabitating man in order to maximise social security payments.^[187] Instead, the AAT decisions reveal that clients, usually women, are participating in ambiguous relationships because of the need to meet care responsibilities with limited resources.

Although many appeals relate to mothers with young children,^[188] the AAT decisions show that it is not only this group who are targeted by the rule. Aged Pensioners and Disability Support Pensioners who are sharing a house to save resources^[189] have also been suspected of ‘cohabitating’.^[190] Of the AAT decisions, 78.5 per cent (62/79) involved the complexity of separation under the one roof. Being separated but living under one roof is not uncommon and occurs for many reasons including: maintaining a secure home and home-life for children;^[191] religious opposition to divorce;^[192] age;^[193] illness;^[194] and cultural and language barriers.^[195] These structural vulnerabilities also include specific difficulties associated with the alleged partner, such as aggression, alcoholism and drug use.^[196] Often the client has been the victim of domestic violence^[197] and either remains in an abusive relationship with the alleged partner^[198] or has sought to live with the alleged partner as a result of fleeing this violence.^[199] The former situation occurred in Re Department of Employment and Workplace Relations and George, where Centrelink used police records from a domestic violence incident as evidence of cohabitation.^[200] The manifestation of these vulnerabilities was made clear in Re Walters and Department of Family and Community Services (‘Walters’).^[201] In that case, the client had a young child, a disability and was involved in an abusive relationship. The client was coerced by her alleged partner into claiming the Sole Parent Pension, which he then spent.^[202] Centrelink then made a finding that she was cohabitating with her alleged partner and ordered her to pay back the money. The AAT affirmed the client’s indebtedness to Centrelink, notwithstanding that she ‘may well have been in a difficult relationship and that she was suffering from an illness.’^[203] The reason for the decision was that there were other s 4(3) factors present.^[204]

This targeting of vulnerable members of society was reflected in the interviews. One interviewee, Meredith, married a ‘friend’ to escape from her abusive ex-partner. This ‘marriage’ triggered a cohabitation review resulting in a debt of over $50 000.^[205] During her interview, Meredith said:

Yeah, I stayed married because he [the ex-partner] doesn’t know — he doesn’t know my last name. It was the only way I could get my kids into a school with that name. Um, otherwise I would have had to give them the birth certificate. I don’t have birth certificates, you know, in that name and, like I said, I had him turning up at [the child’s] school trying to pinch her from school.^[206]

Meredith reflected that marrying the friend was a bad decision. In response to the question from the Centrelink officer undertaking her cohabitation review, the following exchange occurred:

‘Why didn’t you change your name through deed poll, you know, you’re an intelligent woman[?]’ and I said ‘I didn’t even think about it’. It was just not a thing that came through my head. I said, ‘If you’d seen the way I was, how I was living’ you know, I was 43 kilos. I was sick, I was not well …^[207]

Meredith’s decision to embark on a ‘marriage of convenience’ with her friend was a panicked and opportunistic decision to change her name through marriage so as to escape her past and provide her children with a better life.^[208]

Another example was Rosie, who was receiving the Carers Payment for looking after her disabled husband from whom she was separated. When her daughter became sick, Rosie also took on caring for her daughter and grandchildren. To make this situation work, she bought a house that could accommodate her estranged husband, daughter and grandchildren. This triggered a cohabitation investigation. During her interview, Rosie said:

when we first moved and I told them that we were purchasing the house, and the circumstances of it because we had — my terminal daughter with me and the grandchildren. … That was when they cut the rate.^[209]

At the very point when Rosie took on additional care and financial responsibilities she was targeted by the cohabitation rule.

Meredith and Rosie’s circumstances reflect the harsh administration of the cohabitation rule. That is, in practice, the rule targets vulnerable members of society who are more likely to enter into the ambiguous relationships that attract cohabitation scrutiny. The rule operates in this manner as a result of its indeterminate nature. That is, in authorising a detailed investigation into a client’s life and their relationships with others, but at the same time providing little guidance beyond the instruction to consider the s 4(3) criteria, the rule leads to the targeting of certain types of clients. Furthermore, the rule disproportionately affects women who, as a result of having responsibility to care for others such as children or ex-partners, are limited in the resources available to them. These ambiguous relationships, created as a result of protection or care obligations, are the focus of the contemporary administration of this rule.

The targeting of vulnerable people by the rule explains why there has been an ‘acceptance’ of its administration. Clients who are targeted by the rule are less likely to challenge or appeal a cohabitation decision, as they have other significant problems in their lives.^[210] A good example is the client in Walters, who only appealed the decision 13 years later, by which stage her child had grown up and she had had an operation to address her disability, meaning that she was finally in a position to challenge the decision.^[211]

B Invasive Administrative Culture

Another dimension of the cohabitation rule is the invasive surveillance which currently occurs in its enforcement. Prior to 1997, there were four sources of information available to Centrelink: material provided by clients; information from data-matching; requests by Centrelink to persons to provide information;^[212] and information provided as anonymous ‘tip-offs’.

Since 1998, there has been a significant expansion of Centrelink’s capacity to gather information. In 1998, Centrelink undertook a pilot program whereby it hired private investigators to provide ‘optical surveillance’ services.^[213] This program was implemented as the ‘Enhanced Investigation Initiative’ in 1999.^[214] In 2000, Centrelink began a program of random reviews of clients and reached a more formalised agreement with the Australian Federal Police (‘AFP’).^[215] In 2001, focus was given to increasing public tip-offs through the ‘Support the System That Supports You’ media campaign and the establishment of a dedicated tip-off hotline and web address.^[216] In 2002, Centrelink concluded data-matching agreements with secondary and tertiary education providers.^[217] In 2003–04, data-matching was expanded to state corrective services departments and a new program of identifying a geographical area for intense data-matching and compliance review was established.^[218]

Additionally in 2003–04, Centrelink renewed the agreement with the AFP, allowing AFP officers to be based within Centrelink.^[219] Also in 2003, amendments to the Financial Transaction Reports Act 1988 (Cth) allowed Centrelink access to financial transaction data held by the Australian Transaction Reports Analysis Centre.^[220] In 2005–06, the ‘Support the System That Supports You’ campaign was renewed.^[221]

Paul Henman has tracked some of the implications of the increased surveillance of unemployed clients.^[222] However, it is in the administration of the cohabitation rule that the impact of Centrelink’s surveillance has had a particularly potent impact. Indeed, surveillance has been specifically focused on cohabitation since the establishment in 2003–04 of a dedicated program for reviewing the relationships of Parenting Payment (Single) clients.^[223]

According to AAT decisions documenting the deployment of these powers, investigations usually begin with information provided through data-matching, particularly with the Australian Taxation Office^[224] and from tip-offs.^[225] Centrelink officers tend to gather documents concerning clients and alleged partners from a diverse array of sources, including application forms,^[226] transaction records from banks and financial institutions,^[227] details from employers,^[228] superannuation beneficiary declaration forms,^[229] leases,^[230] patient and admission records from medical practices and hospitals,^[231] enrolment and associated forms from schools and childcare centres,^[232] private health insurance policies,^[233] vehicle registration records,^[234] children’s birth certificates^[235] and electricity accounts.^[236] Furthermore, the AAT decisions document Centrelink officers commissioning reports from third parties and neighbours,^[237] real estate agents^[238] and landlords.^[239] They also suggest that optical surveillance of a client or alleged partner’s residence is carried out, so that the Centrelink officers may gather evidence about the alleged partner’s movements, duration of stay and whether the client and alleged partner would arrive or depart together.^[240]

These decisions suggest that Centrelink clients who profess to be single are being closely watched, both through data-matching and by members of the public. Furthermore, if a client is suspected of cohabitating, Centrelink officers assemble extensive portfolios of personal information. In operating a regime that facilitates this, Centrelink can be seen to be following the mandates of the rule. First, by putting into place systems that detect alleged cohabitants, Centrelink is carrying out its duty to determine whether there is a marriage or marriage-like relationship and is thereby distinguishing between partnered and single clients.^[241]

Secondly, the indeterminacy of the rule mandates this very need for detailed evidence from many sources in order to determine cohabitation objectively. Ultimately, there are five specific criteria in s 4(3) that, according to Pelka, ought to be determined prior to the broader investigation of the relationship as a whole. The AAT decisions reveal that the rule interfaces with the recent enhancement of Centrelink’s surveillance capacities to establish an invasive administrative culture that appears insatiable.

Further evidence of the growing invasive culture of Centrelink can be seen through Centrelink’s attempt to gain search and seizure powers. Schedule 2 of the Families, Community Services and Indigenous Affairs and Veterans’ Affairs Legislation Amendment (2006 Budget Measures) Bill 2006 (Cth) was intended to allow Centrelink officers to apply for a search warrant to enter premises to record and seize evidence relevant to any social security offence.^[242] Advocacy groups and the Commonwealth Ombudsman identified that the primary use of these powers would be in cohabitation investigations.^[243]

At the Senate Standing Committee on Legal and Constitutional Affairs inquiry into the Bill (‘Senate Inquiry’), one of the critics of the Bill was the AFP, which disputed Centrelink’s claim that they had not adequately responded to past requests.^[244] The Coalition-majority committee concluded that: ‘the proposed powers in Schedule 2 are unsupported by clear evidence, and disproportionate to the likely degree of intrusion which is likely to result from the exercise of the powers.’^[245] In response, the government withdrew sch 2.^[246]

Centrelink’s attempt to gain more intrusive powers shows an invasive administrative culture that has little respect for the privacy of its clients. However, what is oppressive in this situation is not just the harm to privacy rights, but the way that the invasiveness of the surveillance, even without the trauma of a raid, makes clients feel ‘subhuman … [or] under the microscope’.^[247] This was a recurrent theme emerging from the interviews. One interviewee, Jocelyn, had been repeatedly ‘dobbed-in’, even over periods when she was not receiving a Centrelink payment. Each time she was reported Centrelink would commission an investigation. During her interview, she noted that:

People don’t realise what they put people through, and you feel like you’re being watched all the time, and, like, if people walk past — I’m really suspicious of people. … Honestly, you feel like you’re being watched when you’re not doing nothing [sic] wrong.^[248]

Another participant commented on the experience of responding to a demand by Centrelink for more information:

It’s like holding a gun to someone’s head. If you do not provide me with this information, I’m going to pull the trigger. If you do not provide me with this information, then I’m going to cut your payments, so therefore your kids could starve; you will starve; you will be in poverty …^[249]

However, it was the totality of the surveillance that was most often commented on by participants. This was articulated by a participant named Rachel, when she recounted her experience of an interview with Centrelink officers:

one asked the questions. … And one answered it. So one had a folder, he would ask me the question. Example: ‘Did you live with [ex-partner] at so and so address?’ And at first I said, ‘No.’, and they said, ‘Yes, you did.’ The other one answered the question. So whatever question they asked me, they already had an answer. … They basically know when I went to the toilet last. … And it’s very true big brother is above us somewhere because he, um — Centrelink knew every car I had registered, every house I lived in. Every electricity bill. They knew everything.^[250]

Like Rachel, many participants spoke of being ‘known’ as a result of this invasive regime. Whilst being watched was part of the problem, the oppressiveness of the regime came from the powerlessness that it generated in the clients. This helplessness, in combination with the rule’s targeting of vulnerability, can help explain the ‘acceptance’ of the rule. Powerless clients, who feel that they are being watched and have a ‘gun to their heads’ do not challenge decisions.

However, Rachel further recounted that the two officers in her Centrelink interview seemed to be role-playing as television police investigators: ‘They sat down. They had a big folder in front … like you see on QC and — you know, all those Law and Order shows.’^[251] The fact that Centrelink often appears to be both an agency that provides social security payments and a law enforcement agency was one of the primary justifications given by the Senators when they rejected the search and seizure powers.^[252] However, this blurring of administrative and criminal processes has already given Centrelink officers significant scope in intimidating clients.

C Intimidation of Clients

Clients found to be cohabitating when they were claiming to be single can be prosecuted for social security fraud.^[253]

This means that the administrative investigation by Centrelink can lead to criminal law sanctions. The Federal Court in Ridley v Department of Social Security held that the criminal process is separate from the administrative process.^[254] The Court decided that the AAT, when hearing an appeal by a convicted client, may find that there was no cohabitation.^[255] Indeed, the division between criminal and administrative aspects of the cohabitation rule can be seen in Re Hansell v Department of Employment and Workplace Relations.^[256] In that case, a client who was convicted of social security fraud for failing to disclose cohabitation and consequently served a jail term argued, unsuccessfully, that the jail term had ‘clean[ed] the slate’ and he did not have to repay overpayments.^[257]

Not all clients who have an administrative debt arising from cohabitation are prosecuted. In 2004–05, there were 2540 cohabitation debts claimed by Centrelink.^[258]

Of this number, Centrelink undertook 889 prosecution assessments, with 192 briefs referred to the CDPP and 105 eventual prosecutions.^[259] This means that in 2004–05, 35 per cent (889/2540) of clients assessed as having a cohabitation debt were also subject to the criminal process. An example of this is in Pelka, where the Department asked, unsuccessfully, for a stay of the AAT hearing until it had decided whether or not to refer the matter to the CDPP.^[260] In 2005, both the administrative debt and criminal investigation processes in Centrelink were combined under an overarching section named ‘Business Integrity’.^[261]

However, the result of this streamlining has been that further opportunity for the intimidation of clients has arisen.

The Senate Inquiry into the search and seizure provisions in the Families, Community Services and Indigenous Affairs and Veterans’ Affairs Legislation Amendment (2006 Budget Measures) Bill 2006 (Cth) provided a forum for the airing of concerns about the way that the Business Integrity section has discharged its functions.^[262] In that Inquiry, the Queensland Council for Civil Liberties submitted that:

the interview techniques employed by Centrelink Officers [compared] with those employed by police officers shows quite clearly that the Centrelink Officers have no appreciation of the basic principles of natural justice and fairness. From the record of interview that the writer has seen Centrelink Officers regularly engage in leading questions, putting half-truths and misleading slants on the evidence put to witnesses.^[263]

In a similar vein Michael Raper, President of the National Welfare Rights Network, argued at the hearing that:

the people involved in the area of debt recovery, fraud and investigations generally … tend to hold negative attitudes. They tend to be incredibly prejudiced, judgemental and heavy-handed. These are not just my assumptions about their personality; it is based on the way they undertake interviews and the way they gain evidence. They set out with an assumption that the person is guilty and then try to find evidence to prove that is the case.^[264]

In short, officers from Business Integrity have been described as ‘cowboys’ whose approaches to investigations ‘are very careless, very slapdash and very intrusive.’^[265]

The AAT decision in Re Department of Employment and Workplace Relations and Donnelly (‘Donnelly’)^[266] provides a concrete example of these criticisms, particularly as they relate to the interviewing of clients by Centrelink officers. In Donnelly, the AAT found that the client admitted to cohabitating by signing a prepared statement at an interview but that the signing was done ‘under duress’ because ‘she felt threatened and had no choice other than to sign’.^[267] Claims of being threatened and intimidated, particularly by investigatory officers, were consistent themes from the interviews.

Generally, the participants reported two types of threats. The first was the threat of the discontinuation of their payment. As in Donnelly, clients were told that they had to sign a ‘confession’ in order for their payment to be continued. According to a participant, Sabrina:

That was the first thing they said … that I would have to admit that I was in a marriage-like relationship just to keep my benefits — my normal benefits going — and if I was going to fight it they would cut off all my benefits and I would receive nothing. Which they did.^[268]

The second threat was imprisonment. In her interview, Mary noted that ‘[the Centrelink officer] basically kept saying what a bad person [she] was, that [she] was ripping off the system, that [she] was a criminal. And he basically told [her] [she] was going to jail.’^[269] These threats involved not just jail, but also the implications for dependants of the client going to jail. An interviewee, Jocelyn, detailed that the officer she dealt with threatened: ‘“You’re looking at gaol because this is over $20 000. You’ll be looking at doing jail. Now, who’s going to watch your kids?”’^[270]

Participants also recounted a general level of intimidation by Centrelink officers. Many participants stated that the officers operated on an assumption that the participant was dishonest. Sabrina reported that:

the lady that interviewed me over the phone was really rude … when I told her that I wasn’t in a relationship. She asked me a whole lot of questions over the phone. … [S]he accused me of lying. She accused me of trying to rip off the system. She told me that my word wasn’t good …^[271]

Others felt that there was an assumption of guilt — ‘you’re guilty with them until you’re proven innocent.’^[272] Several participants reported open hostility from officers. Mary said that:

As far as he was concerned, I was guilty. … That’s exactly what he said to me. He was rude, aggressive and arrogant. So, I said to him would it be possible if I got the bank to ring him and explain what had happened? … And he said it didn’t matter what they had to say because I was guilty.^[273]

The participants also generally felt that the investigators were not interested in them clarifying their circumstances; instead, they felt that they had ‘tunnel vision’.^[274] Another participant, Violet, thought that a female officer would be more understanding of ‘the family kind of circumstances here, but, no, she, — she didn’t seem — she just seemed to be on one track, and I found it very hard to communicate with her.’^[275] Another aspect of the investigatory process that intimidated the participants was the repetition with which they had to recount their circumstances (according to Violet, by ‘going over the same questions over and over again. I felt like I was — like I was being treated as less than honest’).^[276] The accumulative effect of this continual re-questioning was commented on by another participant, Verdant:

They will say to you, ‘Well, you’re doing this. Why haven’t you looked for accommodation elsewhere? And why are you still staying with this person?’ But in your mind you’ve already told them 10 000 times why you’re there.^[277]

From the interviews, it is difficult to link all of these actions directly to Business Integrity Officers. One of the reasons is that this section of Centrelink seems to tactically hide their identity and the nature of their investigation. Rachel attended the interview with the innocuous invitation ‘to discuss — to find out if things were accurate’ and she was ‘not 100 per cent sure’ which part of Centrelink the investigating officers were from.^[278] At this interview, Rachel signed a statement admitting to cohabitating. When we spoke with her, Centrelink had imposed a debt in excess of $100 000 and she was anxiously waiting for her Magistrates Court date, at which she expected to be jailed.

The accounts of investigations narrated by the participants elaborate the opinions aired at the Senate Inquiry that officers are exploiting the ambiguity generated by the administrative and criminal processes to threaten and intimidate clients. The participants tell of being threatened with loss of payments and jail if they do not confess to cohabitating; being faced with the assumption that they are lying; being slowly worn down by having to repeat their story; and, of being misled into attending interviews. Empowered clients, with the skills and resources to seek independent advice in their dealings with Centrelink, might not feel intimidated by this blurring of administrative and criminal processes. However, by targeting vulnerable clients and then overwhelming them with surveillance, it can be expected that threats and intimidation to accept Centrelink’s version of a client’s relationship would undermine a client’s resolve to resist.

Furthermore, like the tendency to target the vulnerable and the employment of expansive surveillance, the use of threats and intimidation is a product of the indeterminacy of the rule. As an open-ended rule disposed to the subjectivity of decision-makers and requiring detailed evidence to build a composite picture of the client’s relationship, there is considerable scope for clients to challenge an officer’s reading of particular evidence and their reasoning process. French J authorises as much in Pelka when he wrote: ‘it may be that different decision-makers on the same facts could quite reasonably come up with different answers.’^[279] In 2005–06, clients were very successful at having an unfavourable cohabitation decision changed by the AAT, with 66.7 per cent (12/18) of client appeals resulting in the AAT setting aside or varying the earlier decision.^[280]

This figure is higher compared with the average success of social security appeals in the AAT. In 2005–06, 48.0 per cent (315/656) of applications in the social security jurisdiction resulted in a changed decision.^[281] For officers involved in the administration of the cohabitation rule and attempts to determine undeclared cohabitation for debt recovery and fraud purposes, having a client confess to cohabitation must seem a convenient way to sidestep the complexity of the rule, whilst at the same time achieving debt recovery and prosecution targets. In summary, the indeterminacy of the rule mixed with the fact that Centrelink officers are playing the combined role of administrator and police interrogator has created the situation where clients are intimidated by the administration of the rule. This contributes to the explanation for the ‘acceptance’ of the rule, as the intimidation is directly aimed at coercing clients into accepting Centrelink’s assessment of their relationship.

D An Oppressive Regime: Options for Reform

It is tempting at this point to offer proposals for reform of the rule. However, how to reform the cohabitation rule and its administration is a vexed question, which is a subject of another article.^[282]

There are problems associated with simplistic reforms to the cohabitation rule because the rule expresses a strongly held principle of Australian social security policy that ‘partnered’ clients should receive less support than ‘single’ clients. Provided that this principle is maintained, there remains a need for the cohabitation rule in some guise. However, this article contributes to thinking about the reform process in two ways. First, the material gathered regarding the conduct of Centrelink officers investigating cohabitation decisions encourages immediate reforms directed at this element of the administration process. Greater training on correct interview procedures and possibly the removal of the fraud investigation task from Centrelink, with a view to this task being performed by the AFP, would address some of the intimidation disclosed in the interviews. However, at a deeper level, the article has argued that the oppressiveness of the current administration is a function of the indeterminacy of the rule. This is the second contribution to the reform process discussion — reforms that tighten the open-endedness of the rule and reduce the influence of decision-maker subjectivity should be able to address the concerns identified. However, whether that can be achieved through the introduction of a statutory checklist that must be present in a marriage-like relationship, abolishing the rule entirely and reformulating the social security system so as to be only concerned with ‘individuals’, or recasting the rule (as was advocated by Mossman in the 1970s to consider only financial ties) requires detailed and patient analysis.

IV CONCLUSION

This article has argued that the cohabitation rule is indeterminate and that its administration is oppressive. It has examined the history of the rule and found that the rule has been marked by open-endedness and discretion for administrators. Furthermore, in the current articulation of the rule this indeterminacy and subjectivity remains. It was suggested, drawing upon Fuller, that it would be expected that the administration of an indeterminate rule would be associated with resistance and challenge. However, examination of merit review rates and conviction rates revealed the opposite ― a tacit acceptance of the decisions. This conforms to Fuller’s alternative hypothesis: that oppressive administration of an indeterminate rule can manufacture the appearance of acceptance. In Part III the article drew upon AAT decisions and interviews with clients to argue that the current administration targets the vulnerable; that it located clients in an invasive surveillance regime; and that the conduct of some investigatory officers involved threats and intimidation. Faced with this sort of oppressive administration, the AAT decisions and interviews provide support for Fuller’s argument that a rule can be accepted because it has been administered oppressively. In contrast to Carney’s claim that law has had little impact on social security in Australia, the history and contemporary administration of the cohabitation rule has seen an indeterminate policy become an indeterminate law, which has, consequentially, provided the basis for an oppressive regime.

^[*] LLB (Hons), BSc (Griff); Senior Lecturer, Socio-Legal Research Centre, Griffith Law School, Griffith University.

^[†] BSc (Griff), MA (ECU); PhD Candidate, School of Politics and Public Policy, Griffith University.

^[‡] LLB, GDLP (QUT); Principal Solicitor, Welfare Rights Centre. This article comes from a research project between the Socio-Legal Research Centre and the National Welfare Rights Network. The authors would like to thank Charles Lawson, Bronwyn Statham, Robyn Wilkinson and the two anonymous referees for comments on earlier drafts. The authors would also like to thank Lisa Paul and Robbie Sykes for their research assistance. The opinions expressed are those solely of the authors and not of the National Welfare Rights Network nor of the Welfare Rights Centre.

^[1] In relation to how being partnered affects the calculation of the Newstart allowance, see Social Security Act 1991 (Cth) ss 1068(1)(a), (2). For the effect on the Aged Pension and Disability Support Pension, see Social Security Act 1991 (Cth) ss 1064(1)(a), (b), (4). For the difference between the Parenting Payment (Partnered) and the Parenting Payment (Single), see Social Security Act 1991 (Cth) ss 1068A, 1068B.

^[2] Single clients on Newstart receive a maximum fortnightly payment of $424.30 compared to $382.80 for partnered clients: Centrelink, A Guide to Australian Government Payments 20 March – 30 June 2007 (2007) 16 <http://www.centrelink.gov.au/internet/internet.nsf/file

stores/co029_0703/$file/co029_0703en.pdf>.

^[3] Social Security Act 1991 (Cth) s 4(2).

^[4] Re El-Hourani and Department of Employment and Workplace Relations [2007] AATA 1651 (Unreported, Senior Member Isenberg, 9 August 2007) [53], [59].

^[5] The cohabitation rule has always been limited to heterosexual relationships. Indeed, heterosexuality is a requirement of the rule in the Social Security Act 1991 (Cth) s 4(2)(b)(i). For recent criticism of this requirement, see Human Rights and Equal Opportunity Commission, Same-Sex: Same Entitlements: National Inquiry into Discrimination against People in Same-Sex Relationships: Financial and Work-Related Entitlements and Benefits (2007) 198–226.

^[6] See, eg, Zoe Fairbairns, ‘The Cohabitation Rule — Why It Makes Sense’ (1979) 2 Women’s Studies International Quarterly 319.

^[7] Social Security Act 1991 (Cth) s 4(2)(b).

^[8] Social Security Act 1991 (Cth) s 4(3A).

^[9] See Social Security Act 1991 (Cth) ss 4(2)(a), (b), (3). Department of Families, Housing, Community Services and Indigenous Affairs, Australian Government, Determining Separation under One Roof (11 August 2008) Guide to Social Security Law <http://www.facsia.gov.au/

guides_acts/ssg/ssguide-2/ssguide-2.2/ssguide-2.2.5/ssguide-2.2.5.30.html>. See also Commonwealth Ombudsman, Marriage-Like Relationships: Policy Guidelines for Assessment under Social Security Law, Report No 14 (2007) 11–12 <http://www.ombudsman.gov.au/common

wealth/publish.nsf/attachmentsbytitle/reports_2007_14/$file/report_2007_14.pdf>.

^[10] Bill Hayden made a classic defence of the cohabitation rule in Australia when he was Minister for Social Security (in Commonwealth, Parliamentary Debates, House of Representatives, 20 March 1974, 668):

The reason for granting a higher rate of pension to a single person is that a married couple can share the costs of day-to-day living whereas a single person needs a relatively higher rate in order to enjoy the same living standard. … Two single people could not beat the system by simply living together as man and wife. We could not countenance this. Our view, and that of our predecessors in office, was that a couple who live together in this way are in effect married, and they are treated accordingly by the Department of Social Security. I am talking about a male and a female in a de facto relationship.

^[11] Marcia Neave, ‘From Difference to Sameness — Law and Women’s Work’ [1992] MelbULawRw 18; (1992) 18 Melbourne University Law Review 768, 796; Meredith Edwards, ‘Individual Equity and Social Policy’ in Jacqueline Goodnow and Carole Pateman (eds), Women, Social Science and Public Policy (1985) 95.

^[12] See Meredith Ann Edwards, The Income Unit in the Australian Tax and Social Security Systems (DPhil Thesis, The Australian National University, 1983) 81–3.

^[13] Julia Cabassi, ‘Caught in the Poverty Trap’ (1990) 15 Legal Service Bulletin 72, 74; Helen Campbell, ‘Single Mothers Deemed to Be Prostitutes’ (1990) 15 Legal Service Bulletin 44; Jocelynne A Scutt, Women and the Law: Commentary and Materials (1990) 369–72; Robin Joy, ‘Single Mothers: Married to the State?’ (1991) 8 Ormond Papers 142, 149; Meredith Wilkie, Women Social Security Offenders: Experiences of the Criminal Justice System in Western Australia (1993) 40–1; Patricia Easteal, Less than Equal: Women and the Australian Legal System (2001) 69–74. For an example of a case in which this rule appeared to have this effect, see Mary Jane Mossman, ‘The Baxter Case: Social Security and Cohabitation Policy in Practice’ (1976) 2 Legal Service Bulletin 66.

^[14] See below nn 90–2 and accompanying text.

^[15] Sharyn Roach Anleu, Deviance, Conformity and Control (4^th ed, 2006) 401–2. Regina Graycar and Jenny Morgan also seem to treat this rule as having only historical value. In the first edition of their seminal text, The Hidden Gender of Law (1^st ed, 1990) 158, Graycar and Morgan examined and criticised the rule in detail. In the second edition the rule is a note for consideration following the introductory section of the second chapter, discussing the public–private dichotomy: Regina Graycar and Jenny Morgan, The Hidden Gender of Law (2^nd ed, 2002) 14–15.

^[16] See Senate Standing Committee on Legal and Constitutional Affairs, Parliament of Australia, Families, Community Services and Indigenous Affairs and Veterans’ Affairs Legislation Amendment (2006 Budget Measures) Bill 2006 [Provisions] (2006) 17; Welfare Rights Centre, ‘Raid Powers Dumped’ (2006) 24(4) Rights Review: News & Comment on Social Security Issues 7; Submission to Senate Standing Committee on Legal and Constitutional Affairs, Parliament of Australia, in response to Families, Community Services and Indigenous Affairs and Veterans’ Affairs Legislation Amendment (2006 Budget Measures) Bill 2006 [Provisions], 23 November 2006, Submission No 12 (National Council of Single Mothers and Their Children Inc); Evidence to Senate Standing Committee on Legal and Constitutional Affairs, Parliament of Australia, Canberra, 10 November 2006, 5 (Helen Fleming, Senior Assistant Ombudsman, Office of the Commonwealth Ombudsman).

^[17] Tamar Hopkins, ‘Divorcing Marital Status from Social Security Payments’ (2005) 30 Alternative Law Journal 189; Lyndal Sleep, Kieran Tranter and John Stannard, ‘Cohabitation Rule in Social Security Law: The More Things Change the More They Stay the Same’ (2006) 13 Australian Journal of Administrative Law 135; Commonwealth Ombudsman, above n 9.

^[18] Terry Carney, Social Security Law and Policy (2006) 196–203. For general commentary on the relationship of welfare law and social values, see Terry Carney and Peter Hanks, Social Security in Australia (1994) 249–51.

^[19] Carney and Hanks, Social Security in Australia, above n 18, 253; Terry Carney and Peter Hanks, Australian Social Security Law, Policy and Administration (1986) 239.

^[20] Stanley L Paulson, ‘Lon L Fuller, Gustav Radbruch, and the “Positivist” Theses’ (1994) 13 Law and Philosophy 313, 326–7; Lon L Fuller, ‘Positivism and Fidelity to Law — A Reply to Professor Hart’ (1958) 71 Harvard Law Review 630, 648–57.

^[21] Lon L Fuller, The Morality of Law (revised ed, 1969) 110.

^[22] Ibid 43.

^[23] Ibid 39.

^[24] Ibid 42, 96.

^[25] Ibid 38.

^[26] H L A Hart, ‘Positivism and the Separation of Law and Morals’ (1958) 71 Harvard Law Review 593; Fuller, The Morality of Law, above n 21.

^[27] Fuller, The Morality of Law, above n 21, 107–8. Hart famously suggested that a society comprising laws not accepted by the community ‘might be deplorably sheeplike; the sheep might end in the slaughter-house’: H L A Hart, The Concept of Law (1961) 114.

^[28] Old-Age Pensions Act 1900 (NSW) s 9(f).

^[29] Commonwealth, Royal Commission on Old-Age Pensions, Report on Royal Commission on Old-Age Pensions Together with Proceedings, Minutes of Evidence, Appendices, and a Synopsis of the Evidence (1906) 42–3, 208, 242.

^[30] Invalid and Old-Age Pensions Act 1908 (Cth) s 17(c).

^[31] Alan Jordan, ‘As His Wife: Social Security Law and Policy on De Facto Marriage’ (Research Paper No 16, Research and Statistics Branch of the Development Division of the Department of Social Security, 1981) 16–18.

^[32] Commonwealth, Parliamentary Debates, House of Representatives, 14 May 1942, 1236–41 (Edward (Jack) Holloway, Minister for Social Services and Minister for Health); Stephen Garton, Out of Luck: Poor Australians and Social Welfare 1788–1988 (1990) 135.

^[33] Widows’ Pensions Act 1942 (Cth) s 4.

^[34] Widows’ Pensions Act 1942 (Cth) s 4 (emphasis in original).

^[35] Widows’ Pensions Act 1943 (Cth) s 3(b). For a discussion of this change, see Jordan, above n 31, 18.

^[36] Jordan, above n 31, 20.

^[37] Social Security Act 1947 (Cth) ss 59–60.

^[38] See generally Jordan, above n 31, 23–35.

^[39] Ibid 26.

^[40] Ibid 32–5.

^[41] Brian Dickey, No Charity There: A Short History of Social Welfare in Australia (2^nd ed, 1987) 135.

^[42] The introduction of the payment was an election promise adopted by the Whitlam-led Labor Party after the payment was pursued by a ‘vigorous’ campaign by the Council for the Single Mother and Her Child: Susan Barclay, ‘Where There’s No Will There’s No Way: The Interaction of Community Attitudes and Government Policies for Sole Parents’ (1990) 34 Refractory Girl 33, 34.

^[43] In the second reading speech, Bill Hayden glosses over this reform. He observed that: ‘As with Widows’ Pensions, Supporting Mothers’ Benefit will not be payable if the women return to live with their husbands or if they are living with any man on a de facto basis’: Commonwealth, Parliamentary Debates, House of Representatives, 22 May 1973, 2382–3 (Bill Hayden, Minister for Social Security).

^[44] Social Security Act 1947 (Cth) s 83AAA(1)(b). Section 83AAA was an amendment made to the 1947 Act by the Social Services Act [No 3] 1973 (Cth) s 9 (emphasis in original).

^[45] Ronald Sackville, Commission of Inquiry into Poverty, Law and Poverty in Australia: Second Main Report October 1975 (1975) 192.

^[46] Social Services Act [No 3] 1975 (Cth) s 7.

^[47] Commonwealth, Parliamentary Debates, House of Representatives, 15 October 1975, 2118 (Francis Stewart, Minister for Tourism and Recreation and Minister Assisting the Minister for Social Security).

^[48] Ibid.

^[49] Jordan, above n 31, 42–3.

^[50] Mary Jane Mossman, ‘De Facto Marriage and the Social Services Act’ (1975) 1 Legal Service Bulletin 257, 257–8.

^[51] Ibid 258. On the administrative arrangements provided for by the 1975 policy, see M J Mossman and Ronald Sackville, ‘Cohabitation and Social Security Entitlement’ in Susan Armstrong, M J Mossman and Ronald Sackville (eds), Essays on Law and Poverty: Bail and Social Security (1977) 80, 82–3.

^[52] Jordan, above n 31, 37.

^[53] Mossman, ‘De Facto Marriage and the Social Services Act’, above n 50, 257.

^[54] M J Mossman, ‘The Baxter Case: De Facto Marriage and Social Welfare Policy’ [1977] UNSWLawJl 1; (1977) 2 University of New South Wales Law Journal 1, 6–7.

^[55] Ibid 4.

^[56] Peter Hanks, ‘Cohabitation and Natural Justice’ (1979) 4 Legal Service Bulletin 177.

^[57] Ronald Sackville, ‘Social Security and Family Law in Australia’ (1978) 27 International and Comparative Law Quarterly 127, 157. See also Mossman and Sackville, above n 51, 80.

^[58] Sackville, Law and Poverty in Australia, above n 45, 192; Sackville, ‘Social Security and Family Law in Australia’, above n 57, 158.

^[59] Mossman, ‘De Facto Marriage and Social Welfare Policy’, above n 54, 17–18.

^[60] Mossman, ‘Social Security and Cohabitation Policy in Practice’, above n 13, 69; Mary Jane Mossman, ‘Decision-Making by Welfare Tribunals: The Australian Experience’ (1979) 29 University of Toronto Law Journal 218, 229–30. Hanks documented in 1979 the situation of a client who had been disentitled to her pension on the basis of cohabitating, only to have it reinstated upon commencement of proceedings in the original jurisdiction of the High Court: Hanks, above n 56, 178–9.

^[61] Social Security Act 1947 (Cth) pt IVAAA.

^[62] Social Security Act 1947 (Cth) pt IV.

^[63] Social Security Act 1947 (Cth) ss 6(1), 83AAA. The inclusion of fathers was achieved by the Social Services Amendment Act 1977 (Cth) s 3, which amended the Social Security Act 1947 (Cth) to include s 83AAA. Prior to the 1977 amendment sole fathers were granted a Special Benefit at the Department’s discretion: Commonwealth, Parliamentary Debates, House of Representatives, 4 November 1977, 2917 (Ralph Hunt, Minister for Health).

^[64] Jordan, above n 31, 59.

^[65] (1980) 49 FLR 98.

^[66] Ibid 107–9 (Senior Member Todd, Members Oxby and McLelland); John Kirkwood, Social Security: Law and Policy (1986) 61–2; Carney, above n 18, 188.

^[67] [1981] FCA 171; (1981) 38 ALR 405.

^[68] Re Lambe and Director-General of Social Services [1981] AATA 24 (Unreported, Senior Member Hall, Members Pascoe and Billings, 8 April 1981) [47].

^[69] Lambe [1981] FCA 171; (1981) 38 ALR 405, 413 (Evatt, Fisher and Ellicott JJ).

^[70] [1983] FCA 249; (1983) 52 ALR 128, 131 (Fitzgerald J).

^[71] [1981] AATA 42 (Unreported, Senior Member Todd, Members Oxby and Wickens, 5 June 1981) [22]–[29].

^[72] Peter Johnson, The Annotated Social Security Act (5^th ed, 1989) 22.

^[73] [1991] FCA 513; (1991) 32 FCR 164.

^[74] Ibid 166 (O’Loughlin J) (emphasis in original).

^[75] Ibid 170.

^[76] Lynam [1983] FCA 249; (1983) 52 ALR 128, 131 (Fitzgerald J).

^[77] Jordan made this point clearly when, having considered the 1980 and 1981 decisions of the AAT and Federal Court, he suggested that the Department of Social Security’s 1975 policy had been ‘vindicated’ by those decisions: Jordan, above n 31, 68.

^[78] See, eg, Lois Bryson, ‘Women as Welfare Recipients: Women, Poverty and the State’ in Cora V Baldock and Bettina Cass (eds), Women, Social Welfare and the State in Australia (1983) 130, 141; Edwards, above n 12; Regina Graycar and Deena Shiff (eds), Life without Marriage: A Woman’s Guide to the Law (1987) 115.

^[79] Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth) ss 24–5.

^[80] Commonwealth, Parliamentary Debates, House of Representatives, 5 October 1989, 1606 (Brian Howe, Minister for Social Security).

^[81] Social Security Act 1947 (Cth) s 3A, amended by Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth) s 25.

^[82] Explanatory Memorandum, Social Security and Veterans’ Affairs Legislation Amendment Bill [No 3] 1989 (Cth) 18.

^[83] Ibid.

^[84] This was confirmed by textbooks on social security law which restated the existing principles within the statutory framework. The first edition of Peter Sutherland’s annotation of the Social Security Act 1991 (Cth) stated that the s 4(3) ‘criteria reflect the principles developed under the 1947 Act’: Peter Sutherland and Peter Johnston (eds), Annotations to the Social Security Act 1991 (1^st ed, 1992) 7. This approach continues in the latest edition, which regards the AAT and Federal Court decisions concerning the phrase ‘living with a man as his wife on a bona fide domestic basis’ as authority for the current rule: see Peter Sutherland and Allan Anforth (eds), Social Security and Family Assistance Law (2^nd ed, 2005) 10. Carney’s text also adopts this continuity: see Carney, above n 18, 189. Helen Goodman, whose study period corresponded with the introduction of the statutory criteria, was also of the opinion that the statute ‘reflect[ed] the previous understanding of relevant case law’: Helen Goodman, ‘Social Security Appeals Tribunal Decisions and De Facto Relationships: Some Pointers to Variance with the Department of Social Security Decisions’ (1997) 4 Australian Journal of Administrative Law 92, 93.

^[85] Goodman, above n 84, 101–2; Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth) s 28.

^[86] Social Security Act 1947 (Cth) s 43A(1). This section was inserted into the Act via Social Security and Veterans’ Affairs Legislation Amendment Act [No 3] 1989 (Cth) s 28.

^[87] See, eg, Cabassi, above n 13, 74; Campbell, above n 13, 44–5; Graycar and Morgan, The Hidden Gender of Law (1^st ed, 1990) 152–3; Joy, above n 13, 149; Neave, above n 11, 796.

^[88] Social Security Act 1991 (Cth) s 4(3A).

^[89] Social Security (Administration and International Agreements) (Consequential Amendments) Act 1999 (Cth). There was no mention in Parliament when this Bill was introduced as to the specific reason for removal of ss 4(4) and (5) except that it was part of the clarification of social security law associated with the introduction of the Social Security (Administration) Act 1999 (Cth): see Commonwealth, Parliamentary Debates, House of Representatives, 3 June 1999, 5936 (Warren Truss, Minister for Community Services). There are no similar provisions in the Social Security (Administration) Act 1999 (Cth).

^[90] Social Security Legislation Amendment Act [No 2] 1994 (Cth) s 29(b).

^[91] Social Security (Parenting Allowance and Other Measures) Legislative Amendment Act 1994 (Cth) sch 4.

^[92] Social Security (Home Child Care and Partner Allowances) Legislation Amendments Act 1994 (Cth) s 5, amending the Social Security Act 1991 (Cth) to include s 771HA(1); Social Security (Parenting Allowance and Other Measures) Legislative Amendment Act 1994 (Cth) sch 1(1), amending the Social Security Act 1991 (Cth) to include s 905.

^[93] Social Security (Non-Budget Measures) Legislation Amendment Act 1995 (Cth) s 14(c).

^[94] Sutherland and Anforth, above n 84, 16.

^[95] Sleep, Tranter and Stannard, ‘Cohabitation Rule in Social Security Law’, above n 17, 138.

^[96] See Commonwealth of Australia, Determining a Marriage-Like Relationship (2008) Guide to Social Security Law <http://www.facsia.gov.au/guides_acts/ssg/ssguide-2/ssguide-2.2/ssguide-2.2.5/ssguide-2.2.5.10.html> . See, eg, Re Perera and Department of Family and Community Services [2004] AATA 70; (2004) 78 ALD 739.

^[97] [2005] AATA 1147 (Unreported, Senior Member McCabe, 18 November 2005).

^[98] Ibid [22] (Senior Member McCabe). See also Re McKenzie and Department of Family and Community Services [2006] AATA 92 (Unreported, Senior Member Imlach, 6 February 2006) [15].

^[99] Wilkie, above n 13, 43.

^[100] ARO review is the first tier of the merits review scheme for decisions under social security law: Social Security (Administration) Act 1999 (Cth) s 129. Section 135(1) allows ‘authorised review officers’ to exercise the Secretary’s review power and s 235 allows the Secretary to authorise AROs. There exists a pre-formal review stage called the Original Decision-Maker (‘ODM’) Reconsideration process. On the relationship between the ODM and ARO stages, see generally Australian National Audit Office, Centrelink’s Review and Appeals System, Report No 35 (2005); Australian National Audit Office, Centrelink’s Review and Appeals System Follow-Up Audit, Report No 40 (2007).