AustLII Home | Databases | WorldLII | Search | Feedback

ALTA Law Research Series

You are here:  AustLII >> Databases >> ALTA Law Research Series >> 2010 >> [2010] ALRS 11

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Thornton, Margaret --- "Women and Discrimination Law" [2010] ALRS 11; In Patricia Easteal (ed), Women and the Law in Australia, LexisNexis

Last Updated: 23 September 2011

In Patricia Easteal (ed), Women and the Law in Australia, LexisNexis Butterworths, Sydney, 2010, Ch 8, pp 131-51 (pre-publication version)


Margaret Thornton

Professor of Law
Australian National University

The division between public and private life that has been historically etched onto the bodies of men and women lies at the heart of women’s struggle for equality.[1] As Patricia Easteal makes clear in Chapter 1, this division has been normalised through the assignation of women to caring and unpaid work in the home, leaving men free to dominate public affairs and the world of paid work. This chapter will examine the efficacy of sex discrimination legislation as a mechanism for breaking down stereotypical assumptions in the workplace,[2] particularly the notion that the ideal worker is unencumbered and committed to a full-time career.[3]
In the course of the evaluation, I will elaborate on the main conceptual, substantive and procedural problems posed by the legislation in order to highlight the difficulties for complainants. I will focus on the Sex Discrimination Act 1984 (Cth) (SDA) as it has national application, but I will also make reference to distinctive aspects of State and territory legislation;[4] a thoroughgoing analysis of the nine jurisdictions that presently proscribe sex discrimination in Australia is beyond the scope of this chapter.[5]

Endorsing the Status Quo
The objects of the SDA are ambitious, promoting the principle of equality between men and women and specifying the elimination of discrimination in certain areas.[6] However, there is a significant gap between the rhetoric and the reality as the legislation continues to uphold the problematic division between public and private life. As legal regulation is largely confined to the public sphere, a truly radical approach would be required to address properly the inequities of the private sphere. Even then, the legislation does not apply to the public sphere as a whole but is restricted to certain specified areas of public and quasi-public life, such as employment, education, accommodation, access to goods and services, and clubs, with minor variations between jurisdictions. Nevertheless, the dividing line between public and private is not rigid. The inclusion of grounds, such as the proscription of discrimination against workers with family responsibilities,[7] provides evidence of the dynamic nature of gender relations.
When we move beyond the rhetorical aims, it can be seen that the philosophy underpinning the legislation is one of formal equality, which involves treating like cases alike – regardless of sex or other characteristic. Thus, when Mary applies for a managerial position with XYZ P/L, a multinational company, she has to be judged on her merits according to the selection criteria in the same way as John and other male applicants. She cannot be rejected at the outset because of stereotypical assumptions about women in positions of authority, and she cannot be interrogated about whether she intends to have children, her method of contraception or her child care arrangements – unless the same questions are asked of male applicants.
Substantive equality is directed towards achieving equitable outcomes for an entire class. Thus, XYZ could take on board the under-representation of women in management by developing strategies to change the gender profile of the organisation, such as identifying talented women, creating opportunities and providing leadership programs.[8] Such measures represent a ‘soft’ form of affirmative or positive action, which does not conflict with the norms of anti-discrimination legislation. ‘Hard’ forms, such as quotas, could lead to a complaint of discrimination if a man missed out on a benefit because of a perception that women were being treated more favourably.[9]
Australian anti-discrimination legislation makes scant provision for proactive initiatives, apart from ‘special measures’ in certain circumstances, such as pregnancy or childbirth.[10] The underlying assumption is one of strict equal treatment.[11] This is despite the fact that the UN Convention on the Elimination of all Forms of Discrimination against Women, on which the SDA is based, is sex-specific, while the SDA itself, like other Australian anti-discrimination legislation, is sex-neutral, that is, it applies equally to men and women. Such contradictions point to the fact that the legislation consists of an amalgam of both radical and conservative provisions, which reflects its contentious history.[12] These contradictions contribute to the contested character of discrimination.
The legislation is designed to provide a course of action and a remedy for an aggrieved individual (the complainant) rather than punishment of the perpetrator (the respondent). The significance of the individualised focus is that the emphasis inevitably tends to be on the tip of Easteal’s iceberg: that is, complaints that are overt and close to the surface.[13] For example, if Bloggs, the CEO of XYZ, tells Mary that she was the ‘best qualified but the wrong sex’, we would assume that she had been treated less favourably than John, the successful applicant. If Bloggs did not make such a crass comment but proceeded to reject Mary anyway, it would be difficult for her to prove discrimination, however outstanding her qualifications, for an employer is all too often able to proffer a seemingly rational non-discriminatory explanation for its conduct. Thus, while claiming to have appointed John ‘on merit’, XYZ would probably assert that some aspect of Mary’s qualifications – or experience – or management style – was wanting. Even if Mary suspects that the reason given was pretextual, she will be unable to disentangle the discriminatory conduct from the social prejudice against women in positions of authority and prove that she was treated less favourably than John because of her sex. The systemic discrimination that lies deep within the social psyche remains resistant to exposure and remediation under individual complaint-based discrimination legislation.
Despite the problems of form and substance that inhere within the legislative framework, including the numerous exceptions and qualifications that evince deference to vested interests, the hope is that the legislation will ultimately effect a modicum of substantive social change by virtue of the cumulative effect of the resolution of myriad individual complaints. The fact that thousands of complaints are lodged annually, and show no signs of abating,[14] suggests that the aspiration may be unrealisable. Anyone proposing to have recourse to the legislation needs to be realistic about what is achievable.

Total Sex Discrimination Complaints 2007–08[15]
(including breastfeeding, pregnancy, carer or parental status, sexual harassment and family responsibilities)

% of
total complaints

Threshold Issues
(i) Lodgement of Complaints
The SDA operates concurrently with State and territory legislation, which means that Mary is free to choose under which Act to lodge her complaint. An agency such as the Australian Human Rights Commission (AHRC) will then attempt conciliation. If unsuccessful, Mary may proceed to a formal hearing before a tribunal or court. She will then carry the burden of proof and possibly face the legal costs of XYZ, as well as her own, if unable to make out her case successfully. This possibility represents a powerful incentive to agree to a settlement at the conciliation stage.
I stress that barely two per cent of all complaints proceed beyond conciliation to a formal hearing. However, our understanding of discrimination law and practice derives almost entirely from that small proportion of decisions because they are determined in public and the written judgments are reported. In contrast, conciliation is largely a confidential and opaque process (See Ch 2). It might also be noted that neither causation nor any other issue has to be formally proved at the conciliation stage.

(ii) Direct discrimination
The alleged discrimination must arise from a proscribed ground and complainant and respondent need to be connected by an unbroken linear thread; that is, the discriminatory conduct must be shown to have been caused by the respondent. Direct discrimination focuses on the tip of the iceberg and conduct that is overt, such as the comment by Bloggs that Mary was the ‘wrong sex’. Comparability is the essence of direct discrimination in all jurisdictions except the ACT, where the focus is on treatment that is unfavourable or effects disadvantage.[16] Comparability requires a complainant to demonstrate that she was treated less favourably than another in the same or similar circumstances. Issues such as the feminised character of precarious work,[17] and the dearth of women in management,[18] complicate the comparability requirement. This has become more contentious since Purvis,[19] a disability case decided by the High Court.
In Purvis, a majority of the High Court conceptualised the appropriate comparator as a secondary student without a disability who engaged in the same conduct as a complainant with the disability.[20] Discrimination could be established only if the hypothetical comparator had been similarly disruptive in the classroom but was not suspended. The narrow conceptualisation of the comparator apropos Purvis, who is now expected to evince the same characteristics as the complainant, has significant ramifications for direct discrimination complaints generally.[21]
Mary would have to establish that she was treated less favourably than John but, apropos Purvis, she would not necessarily succeed. In the case of a senior position, an applicant’s qualifications and experience need to be evaluated in light of the organisational culture;[22] the criteria do not speak for themselves. There is inevitably going to be deference towards the views of XYZ, which might choose to emphasise one criterion, such as overseas experience, to argue that Mary was less meritorious than John. The chances of success for a complainant at the recruitment level are also reduced because the respondent inevitably has a monopoly over the evidence.
In proving her case, Mary has to show on the balance of probabilities that XYZ caused the harm. However, it is notable that the normal civil standard has been subtly elevated in discrimination complaints because of the moral odium associated with naming someone ‘a discriminator’.[23] The standard of ‘reasonable satisfaction’, developed by Dixon J in Briginshaw v Briginshaw,[24] is often cited.[25] Indeed, de Plevitz argues that the Briginshaw standard has become the accepted test in the anti-discrimination jurisdiction.[26] There are numerous reported decisions in which a complainant has been unable to discharge the burden of proving that the alleged discrimination arose from a proscribed ground. Instead, some other seemingly rational explanation, such as the personality failings of the complainant[27] or ‘organisational change’ is accepted.[28] The burden of proof is heightened by the inequality of bargaining power between Mary, an individual, and XYZ, a powerful corporate respondent.

(iii) Indirect Discrimination

If overseas experience was a specified selection criterion, Mary might be better off conceptualising her complaint as a manifestation of indirect discrimination. This form of discrimination takes a halting step towards addressing systemic discrimination by considering practices that are neutral on their face but exercise a disparate effect by virtue of sex or other characteristic. Mary, like many other women in the company, has not had overseas experience as social norms expected her to follow her male partner and accept positions close to the family home.[29]
While indirect discrimination purports to recognise that men and women are rarely similarly situated because of their different life experiences, the complainant and respondent must still be linked to the impugned practice, but proving the nexus is not easy. Most legislation includes a complex four-part test, although the SDA now incorporates a somewhat simpler test.[30] Mary would be required to show that there was a requirement or condition with which she was expected to comply, which disadvantaged her by virtue of her sex and which was not reasonable. Reasonableness remains the key element, although the burden of proof shifts to the respondent,[31] which is not the case with most State legislation.
It would seem to be relatively uncontentious that the requirement or condition that Mary had to satisfy was to have had an overseas posting, either with XYZ or some other company, although a threshold question is whether an overseas posting was an absolute requirement or not.[32] The SDA has dispensed with the complex test of measuring the impact of the requirement or condition on the complainant’s sex by proving that proportionately more men than women could comply with it.[33] Even if this element had to be proved, it would probably not be difficult to show that a substantially higher proportion of men could comply,[34] although this would depend on the numbers of complying men and complying women in relation to the total pools. In undertaking the computation, care has to be taken not to perpetuate discrimination against women in light of their historic exclusion from senior management positions.[35]
While the concept of disadvantage, which has replaced proportionality in the SDA, is not defined, a court could draw on the notion of disproportionality without becoming enmeshed in statistical complexities. It might take judicial notice of the fact that men find it easier to accept overseas postings than women generally. Also of relevance to the picture being constructed would be the fact that women comprise only 10.7 per cent of executive managers in Australia.[36] If the evidence revealed that only those women without family responsibilities were at the managerial level in XYZ in contradistinction to the experience of male managers, the issue of disadvantage would be confirmed.
The key question on which Mary’s case would turn would be whether the requirement or condition of having held an overseas posting is reasonable or not. The slippery subjectivity of reasonableness is a vexed one and many complainants falter on this part of the test.[38] Despite the reversal of the burden of proof, employer prerogative enables the respondent to reframe and constrain what constitutes discrimination.[39] Even if criteria to help determine reasonableness are included in the legislation, they must still be evaluated in light of organisational norms. The SDA refers to the nature of the disadvantage, the feasibility of mitigating it and the proportionality of the disadvantage.[40] If the requirement is non-negotiable, the disadvantage is devastating. However, as reasonableness is a question of fact, all circumstances need to be weighed up, including cost, the respondent’s financial circumstances and available alternatives.[41]
XYZ would undoubtedly argue that overseas experience was essential for a senior managerial position with a multinational company. The question is whether there are alternatives or not. Could the training be done on-line or through simulations? Reasonableness is not just a question of convenience for the respondent,[42] but the standard is not particularly stringent.[43] The higher burden of proof attaching to complainants does not seem to carry over to the reversal of proof of an evidentiary burden. This means that, despite the specification of criteria for reasonableness, corporate power and administrative convenience continue to be privileged.[44] Generalist courts, including the High Court, have tended to adopt a more deferential stance towards respondents in discrimination cases, which does not augur well for women seeking to rely on indirect discrimination.[45]

Grounds of Discrimination
(i) Sex or Gender?
‘Sex’ is the term that is used consistently within the legislation, with the exception of Tasmania, which uses ‘gender’.[46] What is the difference? ‘Sex’ clearly has a biological dimension and is sometimes contrasted with ‘gender’ to refer to the socially constructed elements of masculinity and femininity.[47] On other occasions, sex and gender are used loosely and interchangeably so that the biological dimensions of ‘sex’ merge with the socio-cultural elements, as may be seen from the legislative inclusion of ‘a characteristic that is generally imputed’ to persons of one sex.[48] Such a provision would encompass gendered stereotypes, such as an assumption that men are better suited to authoritative positions.
Marital status, pregnancy or potential pregnancy, family responsibilities and sexual harassment are all included in the SDA, denoting the way they are interlinked, although one or more of these grounds may be raised in conjunction with the ground of sex itself. Breastfeeding is a discrete ground in a number of jurisdictions, but not the SDA where it is encompassed by ‘a characteristic that appertains generally to women’.[49] For the most part, each ground is treated as a mirror image of the other, although sexual harassment is treated separately, as will be seen.
Sex may also intersect with other proscribed grounds, such as race, disability, sexuality and age. The issue of intersectionality involving discrete grounds is theoretically easier to deal with at the State or territory level, where all grounds are dealt with in the one Act. At the federal level, there are separate Acts for race,[50] disability[51] and age[52], while sexuality is not proscribed at all. However, as AHRC administers all federal Acts and, if Mary happened to be an Indigenous or non-English speaking background (NESB) woman and she believed that race was a factor in her rejection, she would not have to choose between the grounds of race and sex in the lodgement of her complaint. Intersectionality, however, is not properly addressed by either the legislation or jurisprudence.[53] Indeed, as well as sex plus race (See Ch 13), sex plus sexuality (See Ch 15), sex plus disability (See Ch 14) and sex plus age,[54] are all problematic for women. In Wiggins,[55] sex, sexual harassment and disability were dealt with simultaneously, but not intersectionally, by the Federal Magistrates Court. In upholding the complaint on the ground of disability, McInnes FM observed of the complainant, a former Lieutenant in the Navy, that her ‘recollection of the claims relating to sexual harassment and unlawful discrimination should be viewed in the light of her then suffering from serious depressive illness and that she was then undergoing active treatment’.[56] In an intersectional analysis, the Court would have considered whether the incidents of sexual harassment and sex discrimination contributed to the complainant’s disability.
The intersection of sex plus a cognate ground, such as marital status, pregnancy, family responsibilities or sexual harassment is more familiar. However, it is notable that when sex is coupled with one of these grounds, attention tends to be deflected away from sex discrimination itself. Sexual harassment and reproduction appear to be somewhat more tractable to remediation under the legislation than the abstract notion of sex discrimination, at least at the conciliation level, although the number of complaints lodged remain high. The way managerialism and merit, for example, become entwined with masculinity and power cannot be easily captured, as suggested by the XYZ scenario. Despite the best endeavours of feminist scholars to change conventional understandings of sex and gender, female embodiment remains at the heart of discrimination against women in the workforce, which does little to challenge the power structures underpinning the discrimination.[57]

(ii) Marital Status

Marital status has attracted relatively little litigation in its own right although it is a ground that is often coupled with sex. What if Mary’s partner, David, worked for OPQ, a rival company of XYZ and she was refused the position as a result? Would this constitute unlawful discrimination against her on the ground of marital status? Again, the authorities are divided. In Reddrop,[58] the complainant was the best qualified applicant but she was cast aside when it was learned that her husband worked for a rival pharmaceutical firm. The tribunal found that the decision was made on the basis of a characteristic imputed to married women, namely, they are prone to tell secrets to their partners, but the decision was overturned by the NSW Supreme Court. The Court held that ‘marital status’ did not extend to cover discrimination based on the identity or situation of one’s spouse, which would significantly restrict Mary’s chances of success.
However, in Waterhouse,[59] another New South Wales case, in which the complainant had been denied a horse trainer’s licence because she was married to a rogue, the court distinguished Reddrop. It found that the complainant in that case had a close relationship with an employee of a competitor rather than because of a characteristic generally imputed to married women. This reasoning allowed the court to find for the complainant. More recently, a biologist who commenced proceedings in the Federal Court against the Australian Stem Cell Centre because she was allegedly dismissed for being married to the centre’s former chief executive, accepted an out-of-court settlement.[60] While the problem of Reddrop appears to have been partially resolved in New South Wales, it could re-emerge as a result of Purvis. Hence, Mary would be advised to lodge her complaint under the SDA.

(iii) Pregnancy or Potential Pregnancy
Despite the sustained efforts to address pregnancy in the workplace through discrimination and industrial laws,[61] (See Ch 16) employers are reluctant to accommodate either the pregnant or the post partum woman. There is a significant disjuncture between formal policies that operate at a high level of abstraction and their implementation. Pregnant women receive general societal approval – but not in the workplace. Psychological studies suggest that women, unlike men, trade perceived workplace competence for warmth when they become parents, a bias that extends to pregnant women.[62]
A first typical scenario involves the time a woman informs her employer of her pregnancy and seeks maternity leave. The employer, unwilling to accommodate the pregnancy, grant leave or guarantee the employee’s return to the same or a comparable job after maternity leave, alleges poor work performance by the woman, who may then be dismissed.[63] In determining whether direct discrimination has occurred or not, the woman’s treatment is compared with a real or hypothetical employee who was similarly situated to the complainant, but with one proviso: ‘the only impugned discriminant that needs to be excluded from the characteristics of the comparator is pregnancy’.[64] Even though the accusations and dismissal may well have been pretextual, HREOC commissioners accepted the respondents’ reasons for dismissal as bona fide in as many as one-third of reported cases in the first twenty years of the SDA.[65] This underscores both the bias theory as well as the difficulty of satisfying the burden of proof when the unlawful reason, viz, dismissal on the ground of pregnancy, becomes intertwined with the ostensibly lawful reason, namely, the unsatisfactory performance.
The second typical scenario for women commonly arises when a woman endeavours to return to work after maternity leave and finds that that she has been effectively demoted because the quality of her work has been downgraded.[66] A detriment may be found to have occurred, even if the woman is remunerated at the same rate as before.[67] Comparability still has to be established in the case of direct discrimination on the ground of pregnancy which means that the comparator is an ‘employee ...who was not pregnant and who was returning after nine months leave and with rights of the kind reflected in the maternity leave policy’.[68] In proving that the less favourable treatment arose from pregnancy, maternity leave has been accepted as a characteristic that appertains generally to pregnant women.[69]
The ground of potential pregnancy refers to discrimination against women of childbearing age arising from bias in the workplace against them. Wardley,[70] the first case of sex discrimination to go to the High Court, is a classic example. The complainant was rejected as a trainee pilot because she was a woman of childbearing age perceived to be someone likely to take extended leave. Although this complaint arose at the level of recruitment, the complainant was able to demonstrate conclusively that she had performed better on the aptitude tests than a number of successful male applicants. The complainant survived a constitutional challenge to win her case after a protracted struggle.[71]

(iv) Family Responsibilities
The ongoing tension between public and private life is apparent in regard to the struggle to effect a work/family balance, an issue that has been described as ‘the topic of the 21st Century for families, employers and government’.[72] This tension has engendered widespread ambivalence within discrimination discourses. For example, the proscription of discrimination in employment on the ground of family responsibilities under the SDA is limited to the dismissal of the complainant,[73] including constructive dismissal,[74] which places the onus on the individual rather than the state. The ground is then ostensibly sex-neutral and an appropriate comparator is another worker with similar workplace duties but without caring responsibilities. The legislative limitation in the SDA means that sex discrimination becomes the default position, although State legislation is less restrictive. Thus, while caring for a child, elderly parent or other family member is not a characteristic of being a woman in the way that pregnancy and breastfeeding are, a claim of direct discrimination on the ground of sex would have to establish that ‘the characteristic of having responsibility to care for their own children is a characteristic that appertains generally to, or is generally imputed to, women’.[75]
Conflation between family responsibilities and sex may occur via pregnancy discrimination, as problems in the workforce frequently arise following the return from maternity leave.[76] At that stage, a familiar scenario is that some women wish to change their conditions of work from full-time to part-time. This is where the limited nature of the proscription of discrimination on the ground of family responsibilities under the SDA commonly compels a complainant to resort to sex discrimination. Thus, a condition or requirement that a woman must work full-time to maintain her position has been found to constitute indirect discrimination on the ground of sex, for it is a condition ‘likely to disadvantage women’.[77] A finding of constructive dismissal that constitutes indirect sex discrimination, rather than discrimination on the ground of family responsibilities, may result from the refusal of part-time work because the condition impacts disproportionately on women. Nevertheless, the judicial response has been uneven.[78] Ambivalence persists because the question of whether the respondent has acted reasonably or not is once again the question.[79] The complainant’s argument was not accepted in Howe because it was held that Qantas did not impose a requirement on the complainant to work full-time;[80] the problem was that there were no part-time positions available. Driver FM made clear that an employer can always refuse part-time work if it is reasonable to do so.[81]
What constitutes discrimination is always viewed through the contestable lens of reasonableness. If the complainant objects to the work she is assigned after she returns from leave, it would be clearly unreasonable to assign her even less favourable assignments, conduct that may crystallise as victimisation.[82] In contrast, in Evans, the evidence revealed that the complainant had taken some time off as personal sick leave and carer’s leave. Although entitled to this leave, which was accepted by Raphael FM, Branson J qualified the support for it on appeal:

[I]t is not illegitimate for an employer, all other things being equal and provided indirect discrimination is avoided, to favour for re-employment an employee who takes limited leave over an employee who regularly takes a lot of leave, albeit that it is leave to which he or she is entitled.[83]

This view would seem to undermine the already tepid legislative commitment in the SDA to the non-discrimination principle in respect of family responsibilities. While Branson J rejected the finding of sex discrimination and constructive dismissal, she upheld the discrimination on the ground of family responsibilities, but reduced the amount of damages.
Ambivalence would seem to beset the work/family interface at every turn.[84] Jane, an accountant with XYZ, was unable to obtain permission to work part-time and then sought to work from home two days per week to care for an asthmatic child. When XYZ refused permission, she resigned and lodged a complaint of discrimination. The challenging question that emerges once again is what degree of accommodation for Jane’s caring responsibilities is reasonable on the part of XYZ? Should an accountant be required to be physically present in the workplace on a full-time basis when the technology is available to enable the work to be carried out at home for part of the working week?
The requirement or condition in Schou[85] was that the employee should carry out the work of transcription as a parliamentary Hansard reporter in the designated workplace, a requirement with which a higher proportion of the population who did not have the status of a carer were able to comply.[86] Reasonableness was once again the sticking point. Phillips JA evinced hostility towards the idea of accommodating the complainant, finding it ‘almost inconceivable that the attendance requirement for sub-editors to attend the house on sitting days should be regarded as not reasonable’.[87] His view was that employer prerogative should be unquestioningly upheld despite a workable alternative. The Court of Appeal decision and that of the Supreme Court[88] contrasted with the views of the Victorian Civil and Administrative Tribunal,[89] which makes it impossible to give firm advice to Jane. This is despite the fact that XYZ would carry the burden of proof if Jane lodged her complaint under the SDA, whereas Ms Schou bore the burden of proving that the requirement or condition was not reasonable under the EOA (Vic). The divided opinions regarding the work/family intersection highlight the gap between the legislative mandate and the conservative world view of those charged with administering what many still regard as novel laws.[90]

(iv) Sexual Harassment
Sexual harassment complaints are more likely to be successfully conciliated than complaints of sex discrimination, especially if the behaviour is overtly (hetero)sexed and unwanted; the law has no interest in consensual sexual activity between adults.[91] It may be that the greater willingness to accept sexual harassment as a discriminatory harm is because sex at work has the potential to disrupt productivity. In analysing the phenomenon elsewhere, I have argued that the more sexualised the harassment is in heterosexed terms, that is, where there is a male harasser and a female target, the greater are the chances of success for the complainant.[92] If the sexual elements are less overt, the more difficult it is for a complainant to succeed, as the complaint then merges into an abstract issue of sex discrimination.
In all cases, the burden of proof looms large at the formal hearing level, especially if the impugned conduct took place in private. Nevertheless, an employer cannot rely on the ‘private’ descriptor to evade vicarious liability easily, although a harassing employee may be joined as a respondent.[93]
The paradigm of sexual harassment involves a male employee importuning a female employee for sexual favours, although there are manifold variations on this theme. In Poniatowska v Hickinbotham,[94] the complainant had been propositioned by several co-workers, a lewd MMS photograph had been sent to her, sexually explicit remarks had been made by a female supervisor and no remedial action had been taken by management when the complainant made her grievances known. Instead, the complainant was sent warning letters for allegedly poor work performance and then dismissed. Mansfield J did not agree with the negative assessment of the complainant’s work performance: ‘In my judgment, the employer then determined that she was a person who did not “fit” its work environment because she was a female who would not tolerate sexual harassment and the robust work environment’.[95] Ms Poniatowska’s complaints in respect of both sexual harassment and sex discrimination were therefore made out. In terms of sexual harassment, the relevant test was found by the judge to be satisfied, that is, a reasonable person would have anticipated that the complainant would be offended, humiliated or intimidated.[96] In addition, the judge found that Ms Poniatowska had been discriminated against on the grounds of sex, being dismissed in circumstances that would not have occurred in the case of a male employee. What is remarkable about this case is not only the unequivocal acceptance of the complainant’s version of events but the amount of damages awarded. The sum of $466,000 was awarded, which is considerably more than the few thousand dollars the typical complainant receives.[97]
In the case of lewd conduct on the part of co-workers, all tribunals have not been as quick as Mansfield J in Poniatowska to find sexual harassment, especially if the respondent has a sexual harassment policy in place.[98] Nevertheless, sexually permeated workplaces involving the display of pornographic imagery and the normalisation of obscene language and crude sexist jokes may ground a finding of sexual harassment. They are frequently masculinist workplaces where the female complainant may be the first woman. Such a case was Hunt [99] in which the complainant was the first woman to be appointed as manager of the Train Crew Assignment Centre for the New South Wales Rail Corporation. The hostile sexually permeated work environment was exacerbated by poor management practices, which caused the complainant to go on stress leave and then resign.[100]
If Mary did happen to be the first woman appointed to a senior managerial position with XYZ, where there were disaffected male colleagues who were resentful of her exercising authority over them, she could find herself subject to harassment designed to drive her from the workplace. Bullying conduct and gender disparagement, however, will not amount to sexual harassment without the requisite sexual dimension.[101] Thus, if Bloggs were to prod Mary in the chest while telling her to do what she was told, that may not constitute sexual harassment.[102] The test of being ‘offended, humiliated or intimidated’ would probably not be met.[103] If Mary were prodded in an erogenous zone, such as the breast, she might succeed in respect of sexual harassment.
It nevertheless would be open for Mary to argue that the harassment constituted sex discrimination, although it might be hard to prove less favourable treatment, even though she might suspect that a male manager would not be treated in the same way. In Hill,[104] the NSW Equal Opportunity Tribunal found that ‘sex-based harassment’ of a petty vindictive kind constituted sex discrimination because the complainant was treated less favourably than a male manager. However, one suspects that a barrage of micro-inequities, including gender disparaging remarks, may not have quite the same effect when directed at a male manager. We see once again that gender specific conduct does not lend itself well to a strict notion of comparability.

Dispute Resolution
(i) Conciliation
Conciliation is an informal mode of dispute resolution that involves a range of practices on the part of an agency. A conciliation officer acts as a go-between, which means that a face-to-face meeting may not be necessary. Lawyers are usually not involved, but may be represented at a conciliation conference with permission. (For a fuller treatment of ADR, See Ch 2). A settlement can be effected on whatever terms the parties agree.
As mentioned, conciliation is the primary mode of resolving disputes under anti-discrimination legislation with no more than a tiny fraction of complaints proceeding to formal hearing. This does not mean that all complaints are resolved satisfactorily; many lapse or are withdrawn because the complainant is worn down by what seems like a war of attrition being conducted against her.[105]
Conciliation does have some very positive features as a mode of dispute resolution. It is cost-free, expeditious, flexible in terms of effecting a settlement and non-threatening. The parties might be represented at a conciliation conference with permission, but a face-to-face meeting may not be necessary. The downside of conciliation is that the entire process takes place behind closed doors. The terms of settlement are also confidential, which inhibits the dissemination of knowledge about discrimination to the wider community.[106]

Formal Hearings
While the idea of pursuing justice in a public tribunal or court and being declared to be in the right by a judge is appealing in theory, the reality can be a scarifying experience for a complainant, as there may be a protracted hearing, followed by challenges and appeals. In the case of a powerful respondent, such as XYZ, which has substantial resources, this could take years. The cocoon of privacy enveloping the conciliation process is replaced by the full glare of media publicity, which can be overwhelming in the case of sexual harassment complaints where aggressive cross-examination may be a characteristic of the formal hearing.[107] Mary and Jane would have to be very resilient to persevere with a complaint within an adversarial setting against a multinational corporation. A stigma may also attach to them for lodging a complaint about XYZ and pursuing it in a public forum. In the light of Mary’s senior executive status, it could even lead to possible blacklisting and the inability to obtain another position in the future.[108] A representative complaint initiated by a group of women who were all deleteriously affected by the requirement of overseas experience could deflect attention away from Mary if it were feasible.
Hearings in the first instance, such as those before the Federal Magistrates Court, are not bound by the strict rules of evidence.[109] A complainant may appear in person,[110] although her credibility would probably be enhanced if she were represented.[111] The acceptance of the Briginshaw standard of proof discourages self-representation as it exposes complainants to ‘unnecessary pedantic legal scrutiny’.[112] It would be especially difficult for Mary to run a case based on indirect discrimination unaided due to its complexity. XYZ would undoubtedly be represented by high status senior counsel who could choose to focus on technical points of procedure to deflect attention away from the merits of the complaint, as occurs in respect of constitutionalisation.[113] Nevertheless, both she and Jane would find that legal aid for civil litigation is limited.

In contrast to the flexibility of conciliation, the legislation sets out the bases on which a remedy might be effected at a hearing. The ‘make whole’ principle underlies the raison d’etre of any civil remedy. A payment of damages is the most familiar means of compensating for emotional harm and injury to feelings. In addition, economic loss, including lost wages – front pay, as well as back pay – medical expenses, and loss of all kinds may be compensated.[114] Any reasonable act may be ordered, such as an apology or retroactive seniority. Reinstatement is rarely practicable, as is an injunction enjoining the respondent not to persist with the discriminatory conduct. The possibility of encountering a hostile workplace would suggest that a damages award would be most appropriate for Mary, whereas flexible conditions of work might be more appropriate for Jane.
Policy changes may be negotiated at the conciliation level but are problematic at the quasi-judicial and judicial stage where the focus is on compensating the individual,[115] although the NSW Administrative Decisions Tribunal may order the development and implementation of a program in respect of a vilification complaint.[116] While the principle underpinning hearings before specialist tribunals in the past was that each party pay their own costs, a more legalistic approach arising from the shift towards general tribunals and courts has seen an increasing tendency to award costs against the unsuccessful party, which makes the litigation route risky business.

This chapter has outlined the contemporary approach towards issues of concern in relation to sex discrimination in the workforce. It has highlighted the problems arising from legislation that favours formal equality. The jurisdiction can be positive for the occasional woman who is able to establish that she was treated less favourably than a comparable male colleague in instances that are overt, but is problematic in cases involving pregnancy, maternity leave and caring responsibilities, where comparability is contestable. An indirect discrimination complaint may succeed in instances where the impugned practice is egregious but, for the most part, instances of systemic discrimination are not tractable to remediation. The heroic complainant requires fortitude, tenacity – and resources – to challenge any instance of discrimination beyond the level of conciliation.

[*] Thanks to Dr Trish Luker for research assistance and the Australian Research Council for financial assistance.
[1] Margaret Thornton, 1995, ‘The Cartography of Public and Private’, in Margaret Thornton (ed), Public and Private: Feminist Legal Debates, Oxford University Press, Melbourne, pp 2–16.
[2] In 2007–08, 87 per cent of complaints lodged with HREOC under the SDA related to the area of employment. See Human Rights and Equal Opportunity Commission, 2008, Annual Report 2007-2008, p 50.
[3] Joan Williams, 2000, Unbending Gender: Why Family and Work Conflict and What to do about it, Oxford University Press, New York.
[4] Sex Discrimination Act 1984 (SDA Cth); Anti-Discrimination Act 1977 (ADA NSW); Equal Opportunity Act 1995 (EOA Vic); Anti-Discrimination Act 1991 (ADA Qld); Equal Opportunity Act 1984 (EOA SA); Equal Opportunity Act 1984 (EOA WA); Anti-Discrimination Act 1998 (ADA Tas); Discrimination Act 1991 (DA ACT); Anti-Discrimination Act 1992 (ADA NT).
[5] Peter Bailey, 2009, The Human Rights Enterprise in Australia and Internationally, LexisNexis Butterworths, Sydney; Neil Rees, Katherine Lindsay & Simon Rice, 2008, Australian Anti-Discrimination Law: Text, Cases and Materials, Federation Press, Sydney; Chris Ronalds, 2008, 3rd ed, Discrimination Law and Practice, Federation Press, Sydney; Patricia Easteal, 2001, Less than Equal: Women and the Australian Legal System, Butterworths, Sydney; Rosemary Hunter, 1992, Indirect Discrimination in the Workplace, Federation Press, Sydney; Margaret Thornton, 1990, The Liberal Promise: Anti-Discrimination Legislation in Australia, Oxford University Press, Melbourne.
[6] SDA s 3.
[7] Australia ratified ILO Convention 156, Workers with Family Responsibilities, in 1990. In 1992, the ground was included in the SDA. All States and Territories, except South Australia, now include parental status, or a cognate term, as a ground in their anti-discrimination legislation: ADA (NSW) s 49(s); EOA (Vic) s 6(1); ADA (Qld) s 7(1)(d); EOA (WA) s 35A; ADA (Tas) s 16(i); DA (ACT) s 7(1)(e); ADA (NT) s 19(1)(g).
[8] See Equal Opportunity for Women in the Workplace Act 1995 (Cth).
[9] Eg, Regents of the University of California v Bakke [1978] USSC 145; 438 US 265 (1978).
[10] Eg, SDA ss 7D, 31 & 32. A women’s health service was upheld as a bona fide special measure in Proudfoot v ACT Board of Health (1992) EOC 92-417 (HREOC).
[11] Margaret Thornton, 2001, ‘EEO in a Neo-Liberal Climate’, 6(1) Journal of Interdisciplinary Gender Studies, pp 77–104; Glenda Strachan, John Burgess & Lindy Henderson, 2007, ‘Equal Employment Opportunity Legislation and Policies: The Australian Experience’, 26 Equal Opportunities International, pp 525–540.
[12] See the debates at <> accessed 7 August 2009. For commentary, see Susan Ryan, 2004, ‘The “Ryan Juggernaut” rolls on’, 27 University of New South Wales Law Journal, pp 828–832; Susan Magarey, 2004, ‘The Sex Discrimination Act 1984’, 20 Australian Feminist Law Journal, pp 127–134.
[13] Thornton, The Liberal Promise, n 5 at 6.
[14] Paula McDonald & Kerriann Dear, 2008, ‘The Incidence and Patters of Discrimination and Harassment affecting Working Women in Australia’, 22 Women’s Studies Journal, pp 37–48.

[15] Compiled from the Annual Reports of each jurisdiction, 2007–08.
[16] ADA (ACT) s 8(1). See Patricia Easteal & Susan Priest, [2006/2007], ‘Employment Discrimination Complaints at the ACT Human Rights Office: Players, Process, Legal Principles and Outcome’ Contemporary Issues in Law, pp 62–79.
[17] Judy Fudge & Rosemary Owens, 2006, Precarious Work, Women, and the New Economy: The Challenge to Legal Norms, Hart, Oxford.
[18] Amanda Sinclair, 1998, Doing Leadership Differently: Gender, Power and Sexuality in a Changing Business Culture, Melbourne University Press.
[19] Purvis v New South Wales (Department of Education and Training) [2003] HCA 62; (2003) 217 CLR 92.
[20] Purvis, 160 [220] (Gleeson CJ, Gummow, Hayne and Heydon JJ).
[21] Belinda Smith, 2008, ‘From Wardley to Purvis – How far has Australian Anti-discrimination Law come in 30 Years?’, 21 Australian Journal of Labour Law, pp 3–29. Since Purvis, the Disability Discrimination Act 1992 (Cth) s 5(2) has been amended to require ‘reasonable adjustments’ for people with disabilities.
[22] Margaret Thornton, 1985, ‘Affirmative Action, Merit and the Liberal State’ 2(2) Aust Journal of Law & Society, pp 28–40.
[23] Beth Gaze, 2002, ‘Context and Interpretation in Anti-Discrimination Law’, 26 Melbourne University Law Review, pp 325–354, at 335.
[24] [1938] HCA 34; (1938) 60 CLR 336.
[25] Jonathan Hunyor, 2003, ‘Skin-deep: Proof and Inferences of Racial Discrimination in Employment’, 25 Sydney Law Review, pp 535–554.
[26] Loretta de Plevitz, 2003, ‘The Briginshaw “Standard of Proof” in Anti-Discrimination Law: “Pointing with a Wavering Finger”’, 27 Melbourne University Law Review, pp 308–333.
[27] Eg, Fenwick v Beveridge Bldg Products Pty Ltd (1986) EOC 92-147 (HREOC); Dobrzynska v Aust Telecommunications Corp (1990) 92-305 (HREOC); Ho v Regulator Australia Pty Ltd [2004] FMCA 62; (2004) EOC 93-332 (FMC).
[28] Eg, Van Druten v Sheraton Pacific Hotels P/L [1996] HREOCA 23; (1996) EOC 92-855 (HREOC).
[29] New South Wales v Amery [2006] HCA 14; (2006) 226 ALR 196.
[30] SDA ss 5(2); 6(2); 7(2).
[31] SDA ss 7B; 7C.
[32] Secretary of the Department of Foreign Affairs & Trade v Styles [1989] FCA 342; (1989) 23 FCR 251.
[33] Australian Iron & Steel v Banovic [1989] HCA 56; (1989) 168 CLR 165. See also Hunter, Indirect Discrimination, n 5, 202 ff.
[34] Cf Kemp v Minister for Education (1991) EOC 92-340 (WA EOT).
[35] Banovic, 180 (Deane & Gaudron JJ); 191 (Dawson J).
[36] <> accessed 4 August 2009.
37 Eg, Commonwealth Bank of Australia v Human Rights and Equal Opportunity Commission [1997] FCA 1311; (1997) 80 FCR 78.
[38] Eg, Commonwealth Bank of Australia v Human Rights and Equal Opportunity Commission [1997] FCA 1311; (1997) 80 FCR 78.
[39] Sara Charlesworth, 2005, ‘Managing Work and Family in the “Shadow” of Anti-Discrimination Law’ in Jill Murray (ed), Work, Family and the Law, Federation Press, Sydney, pp 88–126, at 93.
[40] SDA s 7B(2).
[41] Waters v Public Transport Corporation [1991] HCA 49; (1991) 173 CLR 349, Dawson & Toohey JJ at [29].
[42] Styles v Secretary, Dept Foreign Affairs & Trade [1988] FCA 364; (1988) 84 ALR 408 at 409 (Wilcox J).
[43] Hunter, Indirect Discrimination, n 5, at 226.
[44] Eg, Amery; Commonwealth Bank; Schou.
[45] Margaret Thornton, 2008, ‘Sex Discrimination, Courts and Corporate Power’, 36 Federal Law Review, pp 31–56.
[46] ADA (Tas) s 16(e).
[47] Moira Gatens, 1991, ‘A Critique of the Sex/Gender Distinction’ in Sneja Gunew (ed), A Reader in Feminist Knowledge, Routledge, London, p 139.
[48] Eg, SDA s 5(1)(c).
[49] SDA s 5(1A).
[50] Race Discrimination Act 1975 (Cth).
[51] Disability Discrimination Act 1991 (Cth).
[52] Age Discrimination Act 2004 (Cth). See Patricia Easteal, Channy Hiu Tung Cheung & Susan Priest, 2007, ‘Too Many Candles on the Birthday Cake: Age Discrimination, Work and the Law’, 7 Queensland University of Technology Law & Justice Journal, pp 93–107.
[53] Elena Marchetti, 2008, ‘Intersectional Race and Gender Analyses: Why Legal Processes Just Don’t Get It’, 17 Social & Legal Studies, pp 155–174.
[54] In Hopper v Virgin Blue Airlines P/L [2005] QADT 28, the focus was on the youth of flight attendants (under 35), not sex, yet the eight complainants were all women in an industry where appearance is valued.
[55] In Wiggins v Department of Defence – Navy [2006] FMCA 800.
[56] Wiggins [120].
[57] Cf Beth Gaze, 2004, ‘The Sex Discrimination Act After Twenty Years: Achievements, Disappointments, Disillusionment and Alternatives’, 27(3) UNSW Law Journal, pp 914–925 at 917.
[58] Boehringer Ingelheim v Reddrop (1984) 2 NSWLR 13.
[59] Waterhouse v Bell (1991) 25 NSWLR 99.
[60] Leigh Dayton, 2009, ‘Science centre settles with ex-chief’s wife’, The Australian, 22 July.
[61] Eg, Fair Work Act 2009 (Cth) ss 7084. Paid maternity leave will be introduced in 2011. See Australian Government, Australia’s Paid Parental Leave Scheme: Supporting Working Australian Families <> accessed 23 July 2009.
[62] Barbara Masser, Kirsten Grass & Michelle Nesic, 2007, ‘“We Like You, But We Don’t Want You” – The Impact of Pregnancy in the Workplace’, 57 Sex Roles, pp 703–712.
[63] Eg, Chilcott v HPH Investments (1990) EOC 92-353 (HREOC); Kelly v De Mestre (1993) EOC 92-506 (HREOC); Hancock v TP & PM Caravan P/L (1993) EOC 92-539 (HREOC); Milevski v Boral Building Svs P/L (1995) EOC 92-645 (HREOC); Manley v East West Holdings P/L (1995) EOC 92-645 (HREOC); Dare v Hurley [2005] FMCA 844.
[64] Dare [104], per Driver FM.
[65] Margaret Thornton, 2005, ‘Auditing the Sex Discrimination Act’ in Marius Smith (ed), Human Rights 2004: The Year in Review, Castan Centre for Human Rights Law, Monash University, Melbourne, pp 21–56.
[66] Ilian v ABC [2006] FMCA 1500.
[67] Eg, Gibbs v Australian Wool Corporation (1990) EOC 92-327 (HREOC); Thomson v Orica Aust P/L [2002] FCA 939; Rispoli v Merck Sharpe & Dohme [2003] FMCA 160.
[68] Rispoli at [82] per Driver FM. Cf Thomson at 138 per Allsop J.
[69] Thomson.
[70] Ansett Transport Industries (Operations) P/L v Wardley [1980] HCA 8; (1980) 142 CLR 237.
[71] Smith, ‘From Wardley to Purvis’, n 21.
[72] Sarah Squire & Jo Tilley, 2007, It’s About Time: Women, Men, Work and Family, Human Rights and Equal Opportunity Commission, Sydney, p xi. See also Barbara Pocock, 2003, The Work/Life Collision, Federation Press, Sydney.
[73] SDA s 14(3A).
[74] Commonwealth of Australia v Evans [2004] FCA 654.
[75] Evans at [51], per Branson J.
[76] Belinda Smith & Joellen Riley, 2004, ‘Family-friendly Work Practices and the Law’, 26 Sydney Law Review, pp 395–426.
[77] Eg, Hickie v Hickie & Hunt (1998) EOC 92-137, per Evatt C; Escobar v Rainbow Printing P/L (No 2) [2002] FMCA 122.
[78] K Lee Adams, 2005, ‘Indirect Discrimination and the Worker-Carer: It’s just not Working’ in Jill Murray (ed) Work, Family and the Law, Federation Press, Sydney, pp 18–44.
[79] Mayer v ANSTO [2003] FMCA 209.
[80] Howe v Qantas Airways Ltd [2004] FMCA 242. Cf Kelly v TPG Internet P/L [2003] FMCA 584.
[81] At [130]. Cf Correy v St Josephs Hospital Ltd [2009] NSWADT 40.
[82] Correy v St Josephs Hospital Ltd [2009] NSWADT 40.
[83] At [71].
[84] Eg, French v Gosford City Council (2004) 93-318 (NSW ADT); Du-Bois Hammond v Raging Thunder P/L (2004) 93-341 (QADT); Reddy v International Cargo Express (2004) 93-351 (NSW ADT).
[85] State of Victoria v Schou [2004] VSCA 71; (2004) EOC 93-328 (VCA). For commentary, see Gaze, 2002, ‘Context and Interpretation in Anti-Discrimination Law’, n 23.
[86] The complaint was lodged under the EOA (Vic).
[87] Schou [24].
[88] State of Victoria v Schou [2001] VSC 321.
[89] Schou v State of Victoria (Department of Victorian Parliamentary Debates) (2000) EOC 93-101; Schou v State of Victoria Melbourne (Department of Parliamentary Debates) [2002] VCAT 375; (2002) EOC 93-217.
[90] Charlesworth, ‘Managing Work and Family’, n 38; Adams, ‘Indirect Discrimination and the Worker-Carer’, n 77, at 18, 32; Thornton, ‘Sex Discrimination, Courts and Corporate Power’, n 44.
[91] Burgiss v Clisby P/L [2004] VCAT 1817; (2004) EOC 93-357.
[92] Margaret Thornton, 2002, ‘Sexual Harassment losing Sight of Sex Discrimination’, 26 Melbourne University Law Review, pp 422–444.
[93] For analysis of sexual harassment cases under the SDA, 2000-07, see Patricia Easteal & Skye Saunders, 2008, ‘Interpreting Vicarious Liability with a Broad Brush in Sexual Harassment Cases’, 33(2) Alternative Law Journal, pp 75–79.
[94] [2009] FCA 680.
[95] At [312].
[96] SDA s 28A.
[97] Paula McDonald, Sandra Backstrom & Kerriann Dear, 2008, ‘Reporting Sexual Harassment: Claims and Remedies’, 46 Asia Pacific Journal of Human Resources, pp 173–195; Easteal & Saunders, ‘Interpreting Vicarious Liability’, n 92.
[98] Eg, Howard v Geradin P/l t/a Harvard Securities [2004] VCAT 1518; (2004) EOC 93-358. Cf Styles v Murray Meats P/L (Anti-Discrimination) [2005] VCAT 914.
[99] Hunt v Rail Corporation of New South Wales [2007] NSWADT 152.
[100] Cf Hopper.
[101] Margaret Thornton, 2004, ‘Corrosive Leadership (Or Bullying by another Name): A Corollary of the Corporatised Academy?’ 17(2) Aust Journal of Labour Law, pp 1–24.
[102] Malone v Pike (1997) EOC 92-868 (HREOC). Cf Hosemans v Crea’s Glenara Motel P/L (2000) EOC 93-062 (HREOC).
[103] For critique, see Jenny Morgan, 1995, ‘Sexual Harassment and the Public/Private Dichotomy: Equality, Morality and Morals’ in Margaret Thornton (ed), n 1, pp 89–110; Gail Mason & Anna Chapman, 2003, ‘Defining Sexual Harassment: A History of the Commonwealth Legislation and its Critiques’, 31 Federal Law Review, pp 195–224.
[104] Hill v Water Resources Commission of NSW (1985) 92-127 (NSW EOT).
[105] Charlesworth, ‘Managing Work and Family’, n 38; Thornton, The Liberal Promise, n 1, Ch 5: ‘The Equivocations of Conciliation’.
[106] Easteal & Priest, ‘Employment Discrimination Complaints at the ACT Human Rights Office’, n 16.
[107] Rosemary Hunter, 2000, ‘Evidentiary Harassment: The Use of the Rules of Evidence in an Informal Tribunal’ in Mary Childs & Louise Ellison (eds), Feminist Perspectives on Evidence, Cavendish, London.
[108] Dunn-Dyer v ANZ Banking Group Ltd (1997) 92-897 (HREOC).
[109] Human Rights and Equal Opportunity Commission Act 1986 (Cth) (HREOCA), s 46PR.
[110] HREOCA s 46PQ(1).
[111] Patricia Easteal & Kezia Judd, 2008, ‘“She Said, He Said”’: Credibility and Sexual Harassment Cases in Australia’, 31 Women’s Studies International Forum, pp 336–344, at 343.
[112] De Plevitz, ‘The “Briginshaw Standard of Proof”’, n 26, at 329.
[113] Eg, Wardley; Aldridge v Booth [1988] FCA 170; (1988) 80 ALR 1. See also Margaret Thornton, 1999, ‘Towards Embodied Justice: Wrestling with Legal Ethics in the Age of the New Corporatism’, 23 Melbourne University Law Review, pp 749–772.
[114] ADA (NSW) s 108(2)(a) sets the cap at $100,000.
[115] Leves v Haines (1986) EOC 92-167 (NSW EOT); Misso v City of Moorabbin (1989) EOC 92-254.
[116] ADA (NSW) s 108(2)(e).

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback