II The policy challenge and the policy response

In 2003, Syed Mumtaz Ali, retired lawyer and self-styled ‘patron-in-chief’ of the newly created Islamic Institute for Civil Justice (IICJ), announced that the IICJ would begin arbitrating in family and inheritance matters in accordance with Muslim law, which would then be enforceable under the terms of the Arbitration Act 1991.Footnote ⁵ These proposed tribunals were quickly dubbed ‘Shari'a courts’ in the media, and the prospect of their arrival generated considerable alarm over the importation of oppressive, patriarchal practices associated with Islamic law in other countries. Ali himself did little to dispel the anxiety when he declared that ‘good’ Muslims would be obliged to use such tribunals to the exclusion of the ordinary courts.

In response to the public outcry, the Premier of Ontario sought the advice of his Attorney-General and the Minister Responsible for Women's Issues, who in turn appointed former Attorney-General Marion Boyd to conduct a review of the use of arbitration in family and inheritance law. Her mandate, explained in her report ‘Dispute Resolution in Family Law: Protecting Choice, Promoting Inclusion’ (hereafter ‘Review’), was to ‘explore the use of private arbitration … and the impact that using arbitrations may have on vulnerable people’.Footnote ⁶ Over the course of six months, Boyd received submissions in person, by phone and in writing from over fifty groups, as well as many private individuals. She met with women's organisations, religious bodies, immigrant settlement organisations, family lawyers, arbitrators, mediators and concerned citizens.

Ultimately, Boyd endorsed the continued use of binding, enforceable arbitration of family law issues under the Arbitration Act: ‘The Review did not find any evidence to suggest that women are being systematically discriminated against as a result of arbitration of family law issues. Therefore the Review supports the continued use of arbitration to resolve family law matters’ (Boyd Review, p. 133).

Unsurprisingly, Boyd's report did not quell the controversy. In particular, the Canadian Council for Muslim Women (CCMW) intensified its campaign in opposition to Islamic family arbitration, in collaboration with the National Association of Women and the Law (NAWL) and a major Canadian non-governmental organisation, the Centre for Human Rights and Democratic Development (Rights and Democracy). They convened conferences and workshops to publicise the issue, orchestrated international letter-writing campaigns to provincial politicians, and invited activists from around the world to publicise the negative impact of Islamic law on women's rights in various countries. The fate of ‘Shari'a tribunals’ under Ontario law managed to garner the attention of individuals, activists and governments in Europe, the Middle East, South Africa and Asia. It also generated considerable academic commentary (Bader, Reference Bader and Grillo2009; Bakht, Reference Bakht2006; Bhabha, Reference Bhabha2009; Gaudreault-DesBiens, Reference Gaudreault-desbiens2005; Hirschl and Schacher, Reference Hirschl and Shachar2008–2009; Korteweg and Selby, Reference Korteweg and Selby2012; McGill, Reference Mcgill2007; Schacher, Reference Shachar2008; van Torne, Reference Von torne2007–2008; Weinrib, Reference Weinrib and Moon2008).

In September 2005, the Ontario government announced that it would not adopt the recommendations of the Boyd report. Instead, it would amend the law to deny legal recognition to faith-based arbitration of family law dispute under the Arbitration Act, and bring arbitration of family law disputes within the purview of the Family Law Act. The practical significance of these amendments will be discussed below.

III Surveying the legal terrain

The secular legal system is not generally regarded as implicated in the form or substance of whatever multicultural ‘problem’ between individual and community that it is called upon to resolve. Legal enactments are represented as the instruments through which the state manifests its public reason in regulating how actors ought to and shall behave. But the norms and processes of the legal order often shape the emergence of multicultural dilemmas; as such, the legal regime is as much a part of the problem as the solution. At a minimum, it should be apparent that the state and its legal system simultaneously configure and reflect the social meaning and political implications of culture, liberty and equality. Although this should caution against any attempt to assign an unreflexively prescriptive role to law as a neutral arbiter of competing claims, assumptions about the putative neutrality of law prove difficult to dislodge, especially when positioned against the avowed partiality of culture.

Ayelet Shachar's thorough analysis of the Shari'a arbitration debate describes the proposal as an attempt to ‘privatize diversity’, insofar as it would ‘permit parties to move their disputes from public courthouses into the domain of religious or customary sources of law and authority’ (Shachar, Reference Shachar2008, p. 575). I depart from this characterisation insofar as proponents of faith-based arbitration were not actually rejecting or withdrawing from the state's family law regime. Rather, the state's existing regulatory regime of family law already invited and encouraged parties to resolve their family matters outside the courthouse (Bakht, Reference Bakht2006; Farrow, Reference Farrow2006; Shachar, Reference Shachar2008). That regime authorises parties to resolve aspects of their family disputes through private mechanisms as a matter of principle, and encourages them to do so as a pragmatic matter. The process of contracting out of the statutory regime (or ‘privatising family law’) is so pervasive that the under-resourced, over-burdened family court apparatus depends on parties avoiding recourse to the courts by opting for private ordering.

A review of the family law context as it existed when the controversy erupted is necessary to situate the debate. The family law regime governing couples in Ontario is a complicated admixture of federal law (Divorce Act) and provincial law (Family Law Act). For present purposes, suffice to note that default rules for distribution of property, support (spousal and child) and custody following relationship breakdown are set out in statute. Reformist projects over the last thirty years have made considerable progress in addressing the gendered consequences of marriage breakdown. Judicial decisions and legislative changes altered traditional rules of property division to recognise the role that the unpaid labour of wives plays in enabling husbands to maximise their career development and wealth accumulation. Legislative principles guiding provision of spousal support acknowledge that many women forfeit their own careers in the paid labour force to stay home with children, and some may simply never attain full financial independence, much less the financial position they would have attained had they not married (Section 33, Family Law Act 1990Footnote ⁷).

Although these default rules leave considerable room for judicial discretion in recognition of the infinite real-life variation in circumstances, the point of departure is equal division of family property, plus spousal support obligations that recognise the price that a primary caregiver of children (usually a mother) pays in terms of financial independence and earning capacity. At the same time, parties can, to a significant degree, contract out of the process and substance of the statutory family law regime. Rather than litigating property division, custody, child and spousal support in the ordinary courts, parties can resolve these issues by negotiating an agreement on their own (with or without the involvement of lawyers), or with the intercession of a neutral mediator. The resultant agreements were labelled ‘domestic contracts’ under the Ontario Family Law Act (Part IV). They include prenuptial agreements concluded before a couple gets married, which in some cases amount to anticipatory separation agreements (McLeod, Reference Mcleod2004; see MacLeod v. MacLeod 2008;Footnote ⁸Hartshorne v. Hartshorne 2004), as well as actual separation agreements that are devised upon relationship breakdown. In Ontario, judges can intervene and vary domestic contracts if the terms are, in the idiom of contract law, ‘unconscionable’.

Parties who are separating and who do not arrive at a negotiated or mediated settlement may yet avoid litigation by engaging an arbitrator of their choosing to adjudicate the dispute in a less formal process than a courtroom trial before a judge. Prior to 2006, the Arbitration Act dealt with family breakdown in the same way as commercial disputes. However, Canadian law always prohibited parties from arbitrating status, including marriage, divorce or paternity.Footnote ⁹ Negotiated and mediated domestic contracts, along with arbitration, constitute the instruments of private ordering in family law.

Having embarked on negotiation, mediation or (less often) arbitration, the parties may adopt the statutory scheme regarding division of property, support and custody as the normative framework for resolving their dispute. Alternatively, they may operate according to their own preferences, values and priorities. Section 35 of the Arbitration Act 1991 expressly stated that an arbitral tribunal ‘shall apply the rules of law designated by the parties’. The drafters of the legislation probably contemplated that parties who did not elect to be governed by Ontario law would choose the law of another Canadian jurisdiction, an option sometimes exercised in commercial disputes. Nevertheless, nothing in the legislation precluded the choice of religious law over secular law, or the law of Saudi Arabia over the law of Canada, for that matter.

Where parties avoid litigation and resolve their matters privately, a court would likely never see the contract, settlement or award in advance of its implementation. The notable exception involves child support. If a couple wishes to divorce, a judge will require evidence that appropriate child support arrangements have been made prior to issuing a divorce decree.

Parties to a domestic contract or arbitration may choose to file the contract, settlement or award with the ordinary courts and convert the private agreement into an enforceable court order. This means that if one party defaults on the agreement, the other party can seek judicial enforcement. In practice, private agreements tend not to be filed unless and until an aggrieved party believes the other has defaulted. Once again, no judge will review the domestic contract, settlement or award prior to converting it into an enforceable court order. This means, in effect, that the state lends its coercive power to agreements with little or no advance scrutiny of their content.

Prior to the legislative reforms precipitated by the Islamic arbitration debate, parties to arbitration could determine in advance how binding the arbitral award would be. An award could be advisory, meaning that it must be submitted to an ordinary court, where a judge will review it and make the final determination. At the other end of the spectrum, the parties could agree to preclude any appeal of the arbitral award to the courts. This was an important difference between arbitral awards and domestic contracts. Courts always retained authority to set aside domestic contracts if a dissatisfied party persuaded a court that the terms were ‘unconscionable’, but if the parties to an arbitration agreed in advance to preclude appeal to the courts, the arbitral award would be immunised from judicial oversight. Of course, as a practical matter, the courts were and remain relatively inaccessible to those parties (often women), who bear the financial and emotional burden of instigating judicial intervention.

IV Muslim family law arbitration

Marion Boyd's Review of Islamic family law arbitration in Ontario catalysed what could – with only slight exaggeration – be characterised as two opposing answers to Susan Moller Okin's provocative question: ‘Is Multiculturalism Bad for Women?’ (Okin, Reference Okin, Cohen, Howard and Nussbaum1999). At the risk of homogenising internal variations on either side, the following stylised propositions capture the essence of the opposing positions:

versus:

Indeed, Boyd asserts in her executive summary that the ‘fundamental tension that must be addressed is between respect for the minority group and protection of a person's individual rights within that minority’ (Boyd, Reference Boyd2004b, p. 3). Presumably, it is the protection of women's entitlement to choose and the promotion of minority group inclusion within broader society that the report's subtitle ‘Protecting Choice, Promoting Inclusion’, tacitly endorses.

I have analysed the participation of the Muslim women's organisations in this debate through the lens of citizenship (Macklin, Reference Macklin, Benhabib and Resnik2009). For present purposes, I highlight three of Boyd's conclusions. First, Boyd recommended subjecting family law arbitral awards to the same rules as domestic contracts under the Ontario Family Law Act. The main consequence would be that if and when an arbitral award was challenged in court, a judge could set aside the award on the same grounds as domestic contracts, including ‘unconscionability’ (Boyd, Reference Boyd2004a, p. 4). Next, she advocated in favour of supporting civil society efforts to ‘explain rights under Ontario and Canadian law in a way that is likely to be comprehensible to people of diverse backgrounds and culture’ (p. 10).

Finally, while she was mindful of the pressure a woman might experience to engage in faith-based arbitration, Boyd's recommendations rejected the notion that explicit or implicit threats of social, religious, financial or immigration penalties – essentially, non-violent coercion predicated on structural inequality – would or should vitiate consent to arbitrate:

In September 2005, the Ontario government announced that it would not adopt the main recommendations of the Boyd report. Instead, it would amend the law to deny legal recognition to faith-based arbitration of family law disputes under the Arbitration Act, and to bring arbitration of family law disputes under the purview of the Family Law Act. This amendment would mean that parties would no longer have the option of insulating arbitral awards from appeal to the courts by a dissatisfied party. In announcing his decision, Attorney-General Michael Bryant explicitly invoked gender equality and anti-discrimination as the motive for the rejection of the Boyd report's main recommendation in favour of faith-based arbitration:

A few months later, the government enacted the Family Law Statute Amendment Act 2006, which restricted judicially enforceable family law arbitrations exclusively to those applying the default rules contained in Ontario family law legislation, or the law of another Canadian jurisdiction. The Act stated that ‘[i]n a family arbitration, the arbitral tribunal shall apply the substantive law of Ontario, unless the parties expressly designate the substantive law of another Canadian jurisdiction, in which case that substantive law shall be applied’ (Family Law Act 2006). This means that family law arbitration in Ontario privatises the process of dispute resolution, but the substance of the law applied by the arbitrator remains public. In addition, parties to arbitration can no longer shield awards from appeal to the courts.

To the extent that the issue was presented as a staging of Okin's question ‘Is Multiculturalism Bad for Women?’, Boyd gave a qualified ‘no’, while the Ontario government said ‘yes’. Yet, in the end, the decision by the Ontario government to deny legal recognition to faith-based arbitration does not prevent parties from engaging mediators (as opposed to arbitrators) who can guide them toward an agreement using faith-based principles. Nor does it prevent parties from negotiating agreements according to advice from religious authorities about the requirements of religious law. Section 59.2(2) of the amended Family Law Act leaves open the enforceability of domestic contracts negotiated or mediated in accordance with religious norms, including some version of Islamic law:

To reiterate an earlier point, most private dispute resolution in family law occurs by negotiation or mediation, not arbitration. Therefore, the legislated end to the debate about faith-based arbitration did not and will not actually terminate legal enforcement of mediated or negotiated agreements that are based on religious law.

V Another look at the legal landscape

A casual observer of the controversy within and beyond Ontario might plausibly have construed the Muslim arbitration initiative as heralding an unprecedented establishment of a parallel and independent system of religious law and the concomitant retreat of the secular legal system. Part IV of this article demonstrates that this portrayal – one set of rules for Muslims, and another for everyone else – is misleading in respect of Ontario and many other Canadian and common-law jurisdictions. Had Muslim leaders actually mounted a campaign based strictly on multiculturalist arguments to exempt Muslims qua religious minority from the family law regime applicable to everyone else, I believe that they would have encountered implacable and unequivocal opposition. Despite Canada's famous commitment to multiculturalism, instances of autonomous legal regimes outside of federalism hardly exist. Aboriginal peoples continue to struggle to create spheres of territorial and jurisdictional autonomy. Accommodation within existing laws, rather than the delegation of legal authority as such, marks the Canadian practice of legal multiculturalism. The principle that equality and the rule of law require ‘one law for all’ retains considerable durability and resonance in public discourse. A bald demand for a separate and autonomous Muslim legal system would have had no traction.

It follows that I do not share Jean-François Gaudreault-DesBiens's attempt to draw a qualitative distinction between faith-based arbitration, which he considers ‘essentially an identity-based system of parallel justice’, versus ‘the use of non-state norms in the context of particular, individualized settlements arrived at by the parties to family law-related disputes’ (Reference Gaudreault-desbiens2005, p. 166). His distinction is based on two contestable propositions. First, he asserts that the challenges of discerning consent are unique and more vexing in the context of religious communities than in the secular setting. I dispute this proposition in my analysis of Hartshorne, infra. Second, Gaudreault-DesBiens's characterisation of what faith-based arbitration is frequently shades into what he apprehends it may lead to in the future. Calling the Ontario proposal for faith-based arbitration a ‘parallel system of justice’ is not an accurate description; it is rather a prediction based on a slippery slope argument in which Gaudreault-DesBiens asserts that ‘the State recognition of omni-competent [sic] religious arbitral tribunals … is a step closer to a broader recognition of these tribunals’ exclusive jurisdiction and thus sovereignty over a certain community of believers' (p. 164). But arbitral tribunals recognised under the Family Law Act do not and could never claim ‘omni-competence’ in matters of family law. Moreover, tribunal jurisdiction is dictated not by tribunals themselves, but by the parties according to the terms of their agreement to arbitrate. At this point, Gaudreault-DesBiens is no longer addressing the specific impact of family arbitration on individual participants, but rather warning that the recognition of arbitrations conducted within religious communities ‘may lead to the mutation of social minorities into political ones’. He offers no evidence or argument in support of this assertion. In any event, I would suggest that Gaudreault-DesBiens's conception of the character of minority religious communities is overly rigid and monolithic. Moreover, if Gaudreault-DesBiens's fears about the impact of faith-based arbitration are well founded, the fears are as likely to be realised through the ongoing enforcement of faith-based domestic contracts anyway. Negotiated and mediated agreements have historically been, and will continue to be, the dominant mechanism of privatisation in family law.

The twist in the law of arbitration is that the ‘one law for all’ governing families in Ontario was a law that permits parties to fashion their own law: the Family Law Act sets out default rules to be applied by a court in relation to property division, custody, child and spousal support, and then authorises consenting parties to opt out of the process and substance of this public legal regime. Instead of litigating, parties can arbitrate, mediate or negotiate. Instead of applying a set of publicly promulgated rules of general application, parties can reject the provisions in the Family Law Act and choose another set of legal rules off the pluralist legal shelf, or to tailor-make their own terms via contract. Finally, the Family Law Act and the Arbitration Act made the outcomes of private ordering enforceable by a court. Therefore, the ‘one law for all’ turned out to be a law that allows parties to contract out of the process and substance of that law. With the exception of arbitration, this remains the case today.

My central claim, then, is that the debate about faith-based arbitration transpired on a discursive terrain already circumscribed by the logic of privatisation, as promoted by the Arbitration Act and the domestic contract provisions in the Family Law Act. To put it another way, faith-based arbitration and its normative driver, multiculturalism, were already nested within the domain of privatisation and neoliberal ideals of choice, liberty and autonomy. Public facilitation of private ordering in family law paved the way for faith-based arbitration in Ontario. The attempt to create a ‘Shari'a arbitration court’ was not an initiative to privatise diversity. It was an attempt to leverage the privatisation of family law in the name of diversity.

The trend towards privatisation of government functions accelerated in the 1980s and 1990s as neoliberal economic models extolled the virtues of the market and private ordering over state regulation. This disenchantment with the state extended to dispute resolution, especially in the commercial sphere, where sophisticated economic actors set up private arbitration regimes to achieve faster, cheaper and (according to some) better results. Meanwhile, as family law regimes across Canada incorporated greater attention to gender equity in statutory provisions regarding property division and spousal support (partly to prevent divorcing women from becoming impoverished and dependent on state support), they also formalised opportunities for couples to ‘opt out’ of the statutory allocations. Such opportunities came through privately negotiated domestic contracts that regulated custody, access, property division and support according to the parties' own values, priorities and wishes. In all these contexts, privatisation was promoted as autonomy-enhancing, less adversarial and more efficient.

Against these converging patterns of privatisation in the commercial and family law spheres, family law arbitration could best be viewed as a formal articulation of private dispute resolution with the public justice system. By allowing the parties to select their arbitrator, their law and their preferred level of judicial review, family law arbitration played squarely into discourses of individual choice, and of freedom from an interventionist state. At the time of its passage in the early 1990s, feminists critiqued the characterisation of arbitration in family law disputes as autonomy-enhancing when it transpired under conditions of systemic inequality. They cited the potential for abuse of power, sheltered behind the rhetoric of choice and shielded from public oversight by the privacy of the arbitration room. Feminists lost that battle.

This historical context of arbitration opened certain discursive doors for proponents and closed others for opponents of faith-based arbitration. Proponents could correctly assert that they asked for nothing more than equal benefit of the existing regime. The range of rationales – efficiency, cheapness, respect for religious beliefs, user-satisfaction, choice, discretion – operating from diverse normative premises, could each be deployed towards a common position in favour of faith-based arbitration.

The Arbitration Act already permitted arbitration of family law disputes, already permitted parties to choose alternative normative frameworks (including religious law) and already permitted parties to shield outcomes from meaningful judicial oversight. Given these conditions, one could not easily target faith-based arbitration from among alternative normative frameworks without allegedly breaching the constitutionally protected freedom of religion and/or endorsement of multiculturalism. One could not single out Islamic law for exclusion without meeting the accusation of discrimination as between religions. Equality, freedom of religion, and multiculturalism could all be deployed in good faith in the service of faith-based family law arbitration. Supporters could easily accommodate proposals to strengthen procedural protections to ensure free and informed consent. After all, such modifications presupposed that genuine consent could and would result if only participants were free from the threat of physical violence and received enough accurate legal information about the likely consequences of their participation.

Opponents of Islamic arbitration were thus placed in the unenviable position of demanding retraction of some aspect of the status quo. The Canadian Council for Muslim Women, along with the National Association of Women and Law, and the National Organization of Immigrant and Visible Minority Women, adopted a position that corresponded to the margin of overlap between them, namely a shared concern about the exploitation of gendered inequality under a privatised regime of justice. In effect, they tapped into the general case against binding, enforceable arbitration of family law disputes. To the extent that certain renditions of Islamic law were particularly inequitable toward women, and devout Muslim women (especially recent immigrants) were especially vulnerable to forms of coercion not visible as such in law, Islamic family law arbitration could be viewed as an exemplar of the larger problem of private justice in the realm of family law as conceived under the Arbitration Act.

This line of argument was never fully developed by the interlocutors in the debate, but it is worth pursuing the inquiry with explicit attention to the state of Canadian law on domestic contracts that deviate from the statutory default regime. The trajectory of this analysis is not aimed at Islamic law per se. Instead, it focuses on the state's configuration of family law into a regime that permits parties to ‘opt out’ of the public process and the default public rules contained in the statute. This potentially leaves all parties – secular and religious, rich and poor, Canadian-born and immigrant – with qualitatively different outcomes than would obtain under the default public regime.

The mechanism in Ontario for preventing the enforcement of unjust domestic contracts is the option of challenging a domestic contract before a neutral and independent judge. The judge is able to vary or set aside the domestic contract if one party concealed his or her true financial situation, if a party did not understand the nature or consequences of the contract, or if the contract is ‘unconscionable’.Footnote ¹⁰ This combination of private ordering and public oversight is the gold standard against which faith-based arbitration was measured and found wanting. It continues to govern the enforcement of domestic contracts which, to reiterate, may be the outcome of secular or faith-based negotiation or mediation. The lingering question, then, is how well public judicial oversight of private domestic contracts actually ensures that domestic contracts (secular or otherwise) meet a normatively acceptable standard of fairness, gender equality, respect for autonomy and attention to exploitation.

In order to assess the performance of the status quo, I offer the 2004 Supreme Court of Canada judgment in Hartshorne v. Hartshorne (2004), a precedential case involving a secular, professional couple contracting out of the default statutory rules of property distribution in British Columbia. My analysis returns me to the distinction I drew in the ‘Introduction’ between the encultured and the liberal subject. If the oppressed Muslim woman, psychically and socially imprisoned in her cultural context, haunted the ‘Shari'a arbitration’ debate, then it is her antithesis, the culturally unencumbered and rational free agent, who inhabits the majority's mind in Harthshorne. Hartshorne provides a cautionary tale for those who believe that the present regime – with its denial of faith-based arbitration awards but recognition of domestic contracts – eliminates the dangers or reduces the potential harms that opponents associate with faith-based arbitration. The explanation is that putative disadvantage does not derive solely from the religious norms that inform the content of the domestic contract, but also from the principles by which a public, secular court will review and enforce a domestic contract. These include contested conceptions of agency, law's need to hold people to their voluntarily assumed obligations, and related questions about the role of inequality of bargaining power in evaluating the fairness of an agreement.

The case of Hartshorne v. Hartshorne concerns a ‘prenuptial’ agreement presented by Robert to Kathleen less than a month before their wedding, and signed by Kathleen on their wedding day. Robert and Kathleen Hartshorne were both lawyers. They met and commenced a romantic relationship in 1982 when Kathleen was working as a junior lawyer in Robert's law firm. They began cohabiting in 1985. Kathleen bore their son in 1987, after which she left legal practice and remained at home. From this point onwards, Robert ‘gave her a monthly allowance, as well as access to some credit cards for the purpose of charging expenditures for the household’ (Court of Appeal, 2002, p. 10). The couple married in 1989, after she had been out of the paid workforce for two years. She was also (unknown to both at the time) pregnant with their daughter.

This was a second marriage for both parties. Kathleen entered the marriage with significant debt from her first marriage. Robert entered the marriage with significant assets, including his law practice, pension, and vacation properties. In the course of Robert's first divorce, the default regime of property division applied, meaning that he and his first wife (also a practising lawyer) divided their assets equally. Robert Hartshorne testified that ‘he had worked hard to get where he was, he had to divide up his assets following his first marriage breakdown and, in his view, he made it clear to [Kathleen] that he was never going to let that happen again, as he did not intend to share his assets’ (Hartshorne v. Hartshorne, 2002, p. 13).Footnote ¹¹

The marriage agreement stipulated that ‘during their marriage each of them shall be completely independent of the other as regards enjoyment, control, administration and disposal of all property, both real and personal, whether owned at the commencement of the said marriage or acquired thereafter’. The agreement further stated that Robert was the purchaser of the matrimonial home, and that ‘for every year they cohabit, Kathleen, after the marriage, will be entitled to Three (3 percent) per cent of the market value … up to a maximum of Forty-nine (49 percent) per cent of the market value’.

Robert presented Kathleen with the draft agreement two to three weeks before the wedding. Kathleen sought independent legal advice nine days prior to the wedding. The next day, her lawyer advised her that the agreement was grossly unfair compared to her entitlements under the statutory family law regime in British Columbia:

On the day of the wedding, Kathleen Hartshorne made minor amendments to the marriage agreement and signed it. She inserted the following terms at the bottom of the contract:

A majority of the Supreme Court of Canada, reversing two lower court judgments, upheld the terms of the marriage agreement (Hartshorne v. Hartshrone, 2004).

The normative framework governing the marriage agreement might be best described as the abstract principle of formal equality: the property allocation upon divorce should leave each party with whatever property or assets that person brought into the marriage and acquired thereafter in his or own name, including the matrimonial home. In jurisprudential terms, this case represents a reversion to the pre-reform era.

The case that mobilised activists and catalysed reform in Canada was a notorious 1973 Supreme Court of Canada judgment in Murdoch v. Murdoch (1975),Footnote ¹² in which a ranch wife was denied a share of the farm property held in the husband's name, even though she contributed both labour and (arguably) money to its acquisition and development. Within this patriarchal normative framework, property division upon dissolution of marriage tracked the titleholder, typically the husband. The contribution of the stay-at-home wife's labour was not understood as giving rise to an entitlement to assets acquired or improved during the relationship as long as title was held in the husband's name. As Lady Hale recently observed (with refreshing candour), the raison d'etre of the typical prenuptial agreement is to reverse the equalising impact of family law reform on the post-divorce financial situation of husbands and wives in traditional marriages:

The terms of the marriage agreement between Robert and Kathleen Hartshorne roughly instantiate these values, and they predictably operated to the detriment of Kathleen Hartshorne insofar as their marriage conformed to the stereotype of breadwinner husband and stay-at-home wife with no independent income. Kathleen was in no position to acquire property or accrue income of her own.

When the couple separated after nine years of marriage, over twelve years of cohabitation, and two children, Mrs Hartshorne's entitlement under the prenuptial agreement consisted of approximately 20 percent of the family assets. Mr Hartshorne walked away from the marriage with a net worth of $1.2 million and Mrs Hartshorne walked away with a net worth of $280,000. In other terms, the assets available upon dissolution of the marriage were split 80/20 in favour of Mr Hartshorne.Footnote ¹⁴

In comparison to Ontario, British Columbia's default regime is more generous to the economically weaker spouse. In addition, where parties opt out and enter into a domestic contract, a reviewing court in British Columbia may intervene if the outcome of the agreement is ‘unfair’ at the time of distribution, which is less demanding than the standard of ‘unconscionability’ in Ontario (Hartshorne v. Hartshorne. 2002, p. 14).

Kathleen Hartshorne successfully argued at trial that the marriage agreement was unfair, and the British Columbia Court of Appeal upheld the trial judge's reapportionment of assets from 80/20 to a 60/40 split. The Supreme Court of Canada allowed Mr Hartshorne's appeal and rejected the argument that the marriage agreement was unfair. The Supreme Court employed a test developed in an earlier case for assessing whether a domestic contract was fair at the time it was made and, if so, whether it remained fair at the time of implementation. The test can be framed as three questions:

1. 1. Did one party exploit the other's vulnerability at the time of contract formation such that the latter's consent was not voluntary?

2. 2. Did the terms of the agreement ‘fail […] to comply substantially’ with the objectives of the default statutory regime (Hartshorne v. Hartshorne, 2004, p. 41)?

3. 3. Did the actual circumstances of the parties diverge so significantly from their expectations of the future at the time of contract formation as to conclude that implementation of the contract would no longer be fair (Hartshorne v. Hartshorne, 2004, p. 44)?

In answer to question 3, the majority of the Supreme Court of Canada found no significant disparity between the expectations of the parties and what transpired. Life unfolded roughly the way the parties expected: Mr Hartshorne continued his law practice and Ms Hartshorne stayed home and raised their two children, the first of whom had special needs.

Was Kathleen's consent to the agreement voluntary? All judges, at all levels, agreed on the point that Kathleen Hartshorne was not coerced or under duress when she signed the prenuptial agreement on her wedding day. Kathleen's self-serving declaration in the agreement that she was not signing ‘voluntarily’ but rather ‘at the insistence of Robert’ was given no weight. The dissent does acknowledge the imbalance of power between husband and wife before and after the wedding, but concludes that Kathleen's vulnerability, dependence and evident distress about the terms of the prenuptial agreement did not vitiate Kathleen's consent (Hartshorne v. Hartshorne, 2004, p. 90).

Importantly, the inequality of bargaining power did signal to the dissent that a court ‘should be alive to the possibility that the agreement was unfair’ in substance (Hartshorne v. Hartshorne, 2004, p. 90). Conversely, the majority describes the agreement as reflective of the ‘intention of the parties’, and admonishes that ‘if the Respondent truly believed that the Agreement was unacceptable at that time, she should not have signed it’ (Hartshorne v. Hartshorne, 2004, p. 65). In the result, the majority endorses a zero-sum approach to autonomy and consent: if the circumstances do not amount to ‘duress, coercion or undue influence’ in law then the irrebuttable presumption is that both parties acted autonomously, and the imbalance in power is not relevant to any subsequent assessment of the fairness of the content of the agreement.

The requirement that a private agreement substantially comply with the statutory objectives of the default regime provides a constraint on the range of private arrangements that will be deemed fair. Section 65 of the British Columbia Family Relations Act 1996Footnote ¹⁵ states as follows:

In Hartshorne, the majority of the Supreme Court of Canada weighed two factors heavily in arriving at the conclusion that the 80/20 property split was not unfair when viewed against the objectives of the Family Relations Act. First, under Section 65(1)(c), Robert Hartshorne acquired most of his property before the marriage. Second, Katherine Hartshorne's need to regain economic self-sufficiency (Section 65(1)(e)) and the career impediment imposed by her continued responsibilities as custodial parent (Section 65(1)(f)) could be addressed by appropriate provision of spousal and child support. Spousal and child support were not affected by the marriage agreement, but had also not yet been resolved at the time of trial.

Justice Deschamps dissented from the majority's reliance on the parties' expectations as the measure of fairness. She insisted that ‘fairness is a concept that is independent of any agreement’ (para. 78). The fairness of a marriage agreement should be measured by the breadth of its deviation from the outcome a judge could reasonably have ordered under the statutory regime. A ‘great disparity’ should not merit deference by a court.Footnote ¹⁶

For present purposes, I am willing to concede that Hartshorne is a difficult case about which reasonable people can disagree. On the one hand, Kathleen Hartshorne was in a precarious position at the time of the wedding. She had been at home with their child and financially dependent on Robert for two years when they married (Shaffer, Reference Shaffer2004). Regardless of Robert's intention and desire that the couple share their lives but not their property and assets, the social reality of cohabiting in a conjugal relationship over time casts doubt on the possibility and legitimacy of fixing this financial bifurcation permanently and indefinitely. The imposition of the agreement by Robert on Kathleen immediately before the wedding doubtless exacerbated the pressure on Kathleen to sign.Footnote ¹⁷ On the other hand, Kathleen Hartshorne was an educated person, had the benefit of independent legal advice, and understood the nature and consequences of the agreement when she signed it. The embarrassment of cancelling a wedding, and the alternative of returning to the paid workforce as a lawyer and single parent (assuming the couple did not simply continue cohabiting) did not amount to coercion that vitiated her consent.Footnote ¹⁸ She might have made a bad choice in signing the prenuptial agreement, but respect for autonomy requires the law to enforce bad choices as well as good ones, otherwise autonomy itself is made meaningless. Rescuing people from their own poor judgment may seem attractive in the short run, but ultimately patronises and infantilises them (Leckey, Reference Leckey2008). For a variety of reasons, legal doctrine (in family law and elsewhere) cannot easily discard or dilute liberalism's commitment to individual agency and autonomy that ‘contributes to the personalization of politically contoured conflicts and inequalities’ (Brown, Reference Brown2006, p. 17).

Hartshorne is a thoroughly secular product of the privatisation authorised by the current family law regime. My excursion into this case of a middle-class, professional couple is intended to clarify the normative benchmark against which faith-based domestic contracts will be judged. What can it tell us about private resolution of family law issues in the context of religious couples using faith-based principles to determine property, custody and support?

Hartshorne teaches that respect for the autonomy of parties ‘to apply their own values and pursue their own objectives in reaching a settlement’ (Miglin v .Miglin, 2003, p. 55)Footnote ¹⁹ means that domestic contracts can instantiate norms that might otherwise appear retrograde or inegalitarian – like the idea that a stay-at-home wife's domestic, social and emotional labour do not contribute to the acquisition or value of assets held in the husband's name. Much of the debate over Islamic arbitration concerned the alleged rules of Islamic family law, and their unfairness to women as judged by contemporary North American standards. Assume for present purposes that a devout Muslim couple wish to be guided by their imam's understanding of what religious law requires upon divorce, and that the imam's interpretation would in fact result in the wife of nine years and the mother of two children obtaining 20 percent of the family assets and a 27 percent share of the family home. According to Hartshorne, the fact that the wife agreed to be guided by patriarchal inegalitarian norms should not warrant intervention simply because the norms that guided her find their provenance in religious beliefs.

Hartshorne also teaches that the test for coercion is stringent, is not satisfied by proof of unequal bargaining power, and that it will be difficult for a party resisting enforcement to demonstrate lack of consent at the time of contract formation. One can easily imagine that a Muslim woman, like Kathleen Hartshorne, might obtain independent legal advice that fully informs her of the nature and relatively detrimental consequences of a proposed agreement. Yet she signs it anyway. Independent legal advice only ensures that the party is informed, not empowered or, for that matter, wise in her judgment.

Unfairness (much less unconscionability) will be difficult to prove if events unfold more or less as anticipated at the time of the agreement. Similarly, an imbalance of bargaining power between the parties that falls short of vitiating consent will not influence the assessment of whether the terms ‘substantially comply’ with the objectives of the state's default regime from which the parties have contracted out.

Opponents of faith-based arbitration emphasised the many potential sources of vulnerability faced by a Muslim wife to acquiesce to arbitration by a religious authority. Indeed, it is conceivable that lack of education and/or job skills, linguistic abilities, financial desperation, precarious immigration status and, of course, any real or apprehended risk of violence, could lead a court to conclude that a particular Muslim woman did not genuinely consent to the terms of the domestic contract. But these factors do not depend upon religiosity and are not unique to devout Muslims.Footnote ²⁰

This leaves the social pressure that a Muslim woman might experience from within her community to resolve family matters internally and according to religious norms. Recall here that the Boyd report states that ‘more subtle community pressure may not qualify as coercion … whereas threats of violence from a partner or family member almost certainly would, (Boyd, Reference Boyd2004a, p. 136). This acknowledgement seems to correctly state the law. Hartshorne implicitly disregards the pressure Kathleen Hartshorne felt at the prospect of calling off a wedding on the church steps, or the putative stigma attached to being a single working mother. The pressures upon Kathleen seem typical of mainstream North American cultural values regarding women, marriage and motherhood circa 1989. Signs that these values continue to exert considerable force are easy enough to find. But we are less likely to understand these attitudes as immutable cultural commitments; we are more inclined to view them as political commitments that can be scrutinised, revised and rejected. In declaring that Kathleen should not have signed the agreement if she thought it was unfair, the Court ascribes to Kathleen Hartshorne the capacity to withstand the pressure of cultural expectations (as Kathleen perceived them). If this is correct, then the same standard should apply to Muslim women, with the same result.

It is tempting to argue that Hartshorne was rightly decided for a domestic contract between a secular couple but would have been wrongly decided had the agreement's terms disclosed a religious pedigree. In my view, an attempt to draw a distinction along these lines is untenable. Perhaps one could insist that the social and personal cost of ‘exit’ from a religious community is qualitatively different and higher than the costs to Kathleen Hartshorne of rejecting marriage to Robert Hartshorne. In other words, the cultural context within which the encultured woman finds herself is more encompassing and less porous to other currents of thought, and so exerts a distinctively greater coercive force on her. Thus, Kathleen's consent might plausibly be construed as voluntary, while a Muslim woman's is not. However, can we read this conclusion from the fact of the encultured subject's membership in a faith community? A distinction of this sort risks the same patronising stance towards the encultured woman that supporters of the majority in Hartshorne worry about. The encultured subject is regarded as intrinsically impaired in her capacity for consent because of her cultural embeddednesss (and not because of contingent vulnerabilities, like immigration status or poverty). Conversely, her secular counterpart is recognised as an autonomous subject, unencumbered by the cultural influences upon her, or at least able to critically reflect upon and detach from them:Footnote ²¹ Kathleen Hartshorne possesses her culture, but the Muslim woman is possessed by hers.Footnote ²² This stark rendering of the liberal subject on one side, and the encultured subject on the other, means that the politics of subordination and inequality disappear between the cracks.

Of course, the difficulty is that both the encultured and the liberal subject are stylised types that suppress the supplement that constitutes the other. If one applies Hartshorne consistently, religious women should be held to the terms of their faith-based domestic contracts according to the same principles as secular women. This would mean that even if patriarchal religious norms underpin a domestic contract, and the resultant contract disadvantages the woman in comparison with the default regime, the contract should not be unenforceable for those reasons. This is the ironic outcome of liberalism's demand that culture be ‘liberalised’ through ‘privatisation and individualization’ (Brown, Reference Brown2006, p. 21). This is not necessarily a bad outcome, but it should trouble opponents of faith-based arbitration who identify Islam as the villain of their narrative.

The only other alternative for disqualifying faith-based contracts is to regard the encultured subject as intrinsically incapable of genuine consent. This might allow a court to invalidate terms of a religiously inspired domestic contract it would otherwise uphold if they appeared in a non-religious domestic contract. The problems with this approach should be obvious. Indeed, the very activism of the Canadian Council of Muslim Women in opposing Islamic arbitration belies the stereotype of the meek and shrouded Muslim woman. Ironically, the Boyd report invoked the dynamism of the CCMW to refute the claims of vulnerability made by CCMW on behalf of other Muslim women. Boyd was so impressed with the CCMW and other grassroots organisations that she concluded that these organisations could undertake legal literacy programmes within their communities, thereby redressing the ignorance of Muslim women about their rights under Canadian family law. Thus, insulation of arbitration from public scrutiny need not raise undue concern about the welfare of Muslim women because civil society (in the form of the CCMW) would fill the void with education ‘about rights, obligations and options with respect to family law’ (Boyd, Reference Boyd2004a, p. 131). The fact that the CCMW ‘worked hard since its inception within the various Islamic communities to enhance the role of women within the faith and to foster an understanding of the principle of equality so central to Islamic teachings’ (Boyd, Reference Boyd2004a, p. 130) is deployed in part to obviate public responsibility for the protection of vulnerable women in favour of civil society assuming the task. In effect, the Review commends privatised protection (by civil society) as the principal safeguard against the risks created by the privatisation of dispute resolution in family law.

The Boyd report goes on to extol the Canadian Coalition of Jewish Women as ‘a strong role model to other women's groups concerned about potential abuses of religiously based mediation and arbitration’ (Boyd, Reference Boyd2004a, p. 130). The Coalition has indeed produced and disseminated educational materials to observant Jews on this theme. Nevertheless, promoting the Canadian Coalition of Jewish Women as a role model for Muslim women seems ironic, and not for the reasons that one might think. Apart from educating Jewish women about the Jewish divorce process, the notable achievement of the Coalition was to successfully lobby for amendments to the Divorce Act and the Family Law Act to protect Jewish women from intransigent husbands who withheld the get from their wives.Footnote ²³ In other words, the Coalition sought and obtained state intervention precisely because its own educational efforts could publicise but not ameliorate the substantive inequity of Jewish family law.

VI Multiculturalism meets privatisation: options for reform

A central issue for feminists in the Islamic arbitration debates concerned whether and how secular law should protect encultured women from the risk of oppressive intra-communal practices. The public spotlight shone most intensely and harshly on Islamic law, its content, its variability, its manipulation in the service of patriarchy, and the extent to which women would be coerced into submitting to Islamic arbitration. Few proponents of Islamic arbitration actually denied the possibility that certain interpretations of Islamic personal law could be inimical to the welfare of Muslim women. However, the availability of multiple interpretive practices, the ability of Muslim women to assert and defend their own interests, and the competing value to participants of engaging in faith-based dispute resolution were each hotly contested. The discussion then devolved into a set of options as stark as it was false: permit Islamic law arbitration or prohibit it.

In reality, the question was never whether Muslim men and women could lawfully rely on religious authorities to negotiate domestic contracts, to mediate disputes or to arbitrate the consequences of marital dissolution. They have done so in the past, they do so now, and they will continue to do so in the future.Footnote ²⁴ They may participate more or less voluntarily, and abide by more or less fair outcomes. The state possesses neither will nor resources to police whether and how people resolve disputes outside the formal judicial system. In fact, unless one of the parties deliberately engages the formal legal system, the agreement will be insulated de facto from judicial scrutiny. Even if one party litigates, Hartshorne cautions against what one should expect from judicial scrutiny.

This complex question of the interface between the state and private dispute resolution was largely deflected in the discussions regarding Islamic arbitration. For example, Marion Boyd's report made only one relevant recommendation:

Boyd's proposal accepts for the sake of argument what she earlier rejected as unfounded, which is the claim that settlements informed by religious principles are more likely to harm women's interests than settlements generated by secular bargaining practices. Therefore, she proposes consideration of options for calibrating judicial oversight of contracts depending on whether they are the product of religious or secular principles.

I take no position on the empirical validity of Boyd's assumption. However, I reiterate the point that domestic contracts mediated or negotiated with the guidance of a religious advisor remain legally enforceable. Judicial oversight is available to an aggrieved party if and when she has the resources to commence litigation, which may never happen, or may not happen until long after the contract has been implemented. There is one important exception, however: a judge will review a domestic contract to ensure adequate provision of child support prior to granting a divorce.Footnote ²⁵ The rationale for this ex ante review is that there is an overriding public interest in ensuring that the material needs of children are met, and the state does not believe that children can or should bear the burden of litigating to secure adequate support.

This practice of ex ante review of child support is suggestive of a different possible configuration of the private/public interface. I offer it as one alternative to the solution devised by the province of Ontario to the challenges posed by faith-based arbitration. Rather than confining their advance scrutiny of settlements to child support, judges could review the property division and spousal support terms as well. Upon reaching an agreement or arbitral award, the settlement could be filed with a court for automatic paper review by a judge for substantial compliance with the objectives of the applicable statutory regime.

This judicial ‘vetting’ would not be for purposes of guaranteeing future enforcement. Circumstances may change significantly after an agreement is concluded which render one or more of its terms unfair or unconscionable. Rather, settlements that meet the stipulated standards would simply be registered with the Court, and only registered settlements could be relied upon by either party in future litigation. For example, if one party fails to pay child support under the terms of the settlement, the other party can rely upon the existence of the vetted agreement in litigation. Alternatively, if one party seeks an increase in spousal support beyond what the agreement stipulates, the other party can rely on the existence of the vetted agreement in defending against variation. Advance judicial screening functions as a prerequisite to future reliance on the settlement by either party. Otherwise, the agreement does not ‘exist’ for purposes of judicial consideration or enforcement. This consequence of judicial vetting gives both parties an incentive to engage advance review, and to arrive at arrangements that a third-party judge can recognise as substantially compliant with the objectives of the default family law regime.

I do not contend that this option is directly transferable to all components of mediated or negotiated domestic contracts. It is but one model for how a private (and possibly religious) legal order can be brought into constructive dialogue with public legal institutions.

As noted earlier, several proponents of faith-based arbitration who participated in the Review insisted that decisions rendered by religious arbitrators could and would be consistent with the Canadian Constitution and Ontario family law (Boyd, Reference Boyd2004a, pp. 43–45, 63).Footnote ²⁶ Indeed, Boyd observed that ‘again and again, members of the Muslim community assured the Review that Muslims who live in countries not governed by Islamic law are required by their faith to be obedient to the law in place in their country of residence’ (Boyd, Reference Boyd2004a, p. 124) Taking these claims at face value, subjecting faith-based domestic contracts to some form of public judicial oversight should be less threatening to the integrity of faith-based domestic contracts than it might otherwise appear. In any event, it would certainly provide an open opportunity for testing the hypothesis of the compatibility of religious and secular norms.

Having said that, religious adherents might harbour legitimate concerns about judicial sensitivity and respect towards traditions or practices that seem novel and beyond the range of judicial experience. At the same time, the review process creates an opportunity to educate and sensitise judges and religious mediators alike. Moreover, one could institute this relatively intense level of judicial scrutiny for a specified period of time with a view to developing an empirical basis for re-evaluating the desirability and necessity of maintaining it in the future. In other words, one could structure a relationship between the state and faith-based communities that contemplates the possibility of evolution over time.

Any policy option carries implications for utilisation of judicial resources, the interaction of religious and secular norms, the nature of the task performed by mediators and arbitrators, and the allocation of the burden of accessing the courts. And any scheme animated by a concern for vulnerable parties that requires them to initiate the process by commencing litigation is less likely to be effective than one that triggers scrutiny automatically upon conclusion of the agreement or issuance of an arbitral award (see also Shachar, Reference Shachar2008, p. 598).

The option presented above does not exhaust the range of possibilities and, indeed, other commentators have offered creative proposals that resemble elements of my proposal and differ in other respects (Emon, Reference Emon, Korteweg and Selby2012; Shachar, Reference Shachar2008). To the extent that I endorse bringing the secular, public legal order and the full range of private mechanisms amenable to the application of faith-based principles – arbitration, mediation and negotiation – into conversation with one another, my recommendation is broadly compatible with Ayelet Shachar's promotion of ‘transformative accommodation’ as a model for managing relations between the state and minority religious communities (Shachar, Reference Shachar2001). However, any serious attempt to craft a viable policy must delve more deeply into the precise features of legal architecture in the specific jurisdiction, the configuration of religious dispute resolution in that jurisdiction, and the interaction between the two. For example, even though the Canadian common-law legal order closely resembles the UK legal order, the English courts will have regard to domestic agreements, but will not regard them as binding and enforceable contracts (Radmacher, 2010). ‘Shariah Councils’ in England offer Muslim divorces and resolve the ancillary issue of the [re]payment of mahr (dowry). However, John Bowen explains that the ‘councillors do not rule on the issue of child residence or the division of property, knowing that these matters will be determined in civil court if they are not agreed to by the parties’ (Bowen, Reference Bowen2010, p. 419). This illustrates how the functions of the Shariah Council in England take shape against the specific legal landscape of English law.