Every family lawyer knows that the validity of pre-nuptial agreements (pre-nups) is at the mercy of the judge’s discretion, yet a freely entered agreement that is not unfair will be given decisive weight.1 This authority stems from the landmark case of Radmacher v Granatino [2010] UKSC 42, where the Supreme Court ended longstanding ambivalence over the enforcement of such agreements, holding that nuptial agreements are to be given effect by the court out of ‘respect for individual autonomy’.2 This article considers whether the current legal status of pre-nups does indeed promote autonomy. It is argued, first, that the court frequently avoids the question of whether autonomy has been exercised. The second part of this article explores research evidencing why this is problematic. Autonomy is a nebulous concept, and when it is simplified, neutralised, individualised and de-gendered, it is often assumed. The concept of neo-liberal autonomy has been a powerful influence on policies surrounding family breakdown in recent years.3 This version of autonomy focuses on the individual, who is expected to seek the best deal for themselves.4 Yet, if the reality of how individuals make decisions is to be appreciated then it is important to challenge the idea that, in contracts, autonomy simply means a rational and voluntary choice. Particularly in the complex realm of intimate family law agreements, it would be fair to ask whether individuals ever make completely voluntary, rational choices. While neo-liberal notions of autonomy are built on assumptions that decision makers are independent, self-interested and rational actors, relational autonomy asserts that ‘[t]o be autonomous is not to be isolated and free of responsibility, but to be in a network of relationships, with their dependent responsibilities’.5 Our autonomy when making these decisions is inherently impacted by relationships with others, and the relationships with those we are entering into agreements with.6

As a result, this article urges circumspection regarding autonomy in the neo-liberal sense. In the context of nuptial agreements, blind respect for this type of autonomy favours the party with greater bargaining power, often at the expense of the interests of the non-moneyed spouse, because the power struggles in the relationship are not adequately recognised.7 Such defects in the exercise of autonomy are not fully appreciated when agreements are set aside primarily to meet needs.8 Thus, in the final sections, it is suggested that if nuptial agreements are to be made binding in England and Wales, the American Law Institute’s proposals provide an example of how legislation could go some way towards explicitly recognising the issues of power affecting such agreements.

The current legal landscape

Rather curiously, the steps currently required for a nuptial agreement to be given effect by the court could be viewed as simultaneously protecting and overriding autonomy. The first step is obviously tied to respect for individual autonomy, because it requires an agreement to have been freely entered into by the parties.9 But the second step is more complicated, whereby an agreement will not be upheld if the court determines that it would not be fair to do so.10 As a result, while the rationale for enforcing agreements is based on autonomy, an absence of autonomy does not tend to be used by the judiciary to justify an agreement being set aside. Rather, it is an absence of fairness that is important. And, problematically, fairness and autonomy can be treated as conflicting values.

The Supreme Court has described circumstances in which it would be unfair for an agreement to be given effect as follows:

‘The parties are unlikely to have intended that their ante-nuptial agreement should result, in the event of the marriage breaking up, in one partner being left in a predicament of real need, while the other enjoys a sufficiency or more, and such a result is likely to render it unfair to hold the parties to their agreement. Equally if the devotion of one partner to looking after the family and the home has left the other free to accumulate wealth, it is likely to be unfair to hold the parties to an agreement that entitles the latter to retain all that he or she has earned.’11

As a result, the court typically refuses to give effect to an agreement because the judge does not consider sufficient provision to have been made for the parties’ needs (generously interpreted).12 This focus upon needs has prevailed even when the parties have sought to have a pre-nup set aside because their autonomy was defective. In KA v MA (Prenuptial Agreement: Needs) [2018] EWHC 499 (Fam), for example, the wife’s solicitors registered their concern that she had been under pressure to sign the nuptial agreement (the husband allegedly threatened to cancel the wedding unless it was signed), but the court’s rationale for it ultimately being set aside was that it did not provide for her needs.13

This inevitably sets up a tension between needs and autonomy. When giving effect to an agreement is equated to giving effect to individual autonomy, the corollary of this is that varying or disregarding a pre-nup because of needs is viewed as a threat to autonomy. This artificial choice of needs or autonomy means that to avoid accusations of paternalism,14 the scope of needs is in danger of being narrowly constrained by the court. Cummings v Fawn [2023] EWHC 830 (Fam) is one illustration of this:15

‘Imagine that the discretionary range is a line of books on a shelf bracketed left and right by book-ends. The book-ends may be quite far apart. The right book-end represents a comfortable, perhaps even luxurious, life-style. The left book-end represents a spartan lifestyle catering for not much more than essentials. The space in between is the discretionary range. When the Supreme Court says that it may not be fair to uphold an agreement which leaves the applicant in a predicament of real need, it is clearly saying that if the result of the agreement would place the applicant in a standard of living to the left of the left-hand bookend, then that would be unfair. It is also saying that to make the agreement fair it should be augmented by no more than is necessary to move the applicant’s lifestyle just to the right of that left-hand bookend.’16

This case concerned a Xydhias17 agreement rather than a nuptial agreement, but these obiter comments relate to Mostyn J’s assessment of what constitutes a fair agreement more generally. His focus upon basic provision of need in this case – which decontextualised the Supreme Court’s statement concerning one party being left in a predicament of real need – also overshadowed the wife’s contention in this case that there was material non-disclosure; a factor that is of direct relevance to her exercise of autonomy.

In summary, while autonomy is used to justify the current legal status of nuptial agreements more generally, it does not factor much in the way they are adjudicated in practice. Yet if nuptial agreements are to provide a good route to autonomy, then the legal framework must also provide for when autonomy has not been exercised by one of the parties. Research suggests the current legal landscape has produced an overly simplistic and inaccurate picture of what autonomy actually means.18 This is because defects with autonomy are not properly acknowledged when autonomy is presumed. Moreover, it is misleading for fairness and autonomy to be treated as opposing concepts – whereby upholding fairness is viewed as undermining autonomy – since a nuptial agreement may well be unfair because of defective autonomy.

Defects in the exercise of autonomy

Even if we accept that autonomy should be the starting point, in practice this is problematic because the concept of autonomy is often illusory. When the court refuses to ask questions about the exercise of autonomy because there has been adequate disclosure and legal advice, and because the rather high thresholds of duress and undue influence have not been met, it is, in effect, upholding a fictional version of autonomy in many situations. Indeed, practitioners who deal routinely with nuptial agreements will be aware of how changing circumstances, bounded rationality/optimism bias, and unequal bargaining power all affect the autonomy of one or both spouses, while undermining the presumption that giving effect to an agreement necessarily means respecting the autonomy of the parties.

Changing circumstances

Various studies have documented the perennial problem with pre-nups: the circumstances in which the agreement is negotiated before the wedding are likely to be different from the circumstances prevailing when the agreement is brought into effect. This is especially likely in long marriages. In Radmacher, Lady Hale gave three examples to illustrate this:

‘A couple who always thought that one would be the breadwinner and one would be the homemaker may be astonished to find that the homemaker has become a successful businesswoman who is supporting her homemaker husband rather than the other way about.

A couple who assumed that each would run their own independent professional life and keep their finances entirely separate may find this quite impossible when they have children, especially if they have more than one or one of them has special needs.

An older couple who marry a second time round may think it fair at the time to preserve their assets for the sake of the children of their first marriages, but may find that one has to become a carer for the other and will be left homeless and in reduced circumstances if the grown-up children take priority even though they are now well-established in life and have no pressing need of their inheritance.’19

Thus, pre-nups are complex because of the many unforeseeable ways in which the marital relationship may develop over time, and the impact of changing circumstances upon the parties at the time of enforcement is well documented across jurisdictions.20

Bounded rationality/optimism bias

Bounded rationality and optimism bias are further complicating factors connected to the fact that nuptial agreements are created within the circumstances of the parties’ relationship before the wedding. These cognitive limitations affect the negotiation process, since at this time divorce seems to be a distant and unlikely prospect and as Melvin Eisenberg’s work has found, parties to an agreement are likely to be ‘unduly optimistic about the fate of their marriage’.21 Bounded rationality affects the parties’ ability to think clearly about protecting themselves financially on divorce in the early stages of a relationship, which are often marked by altruism and commitment. As Brian Bix has put it: ‘most people are poor at thinking well about events in the distant future, especially if it involves contingencies contrary to our optimistic assumptions’.22 Lynn Baker and Robert Emery’s research has also established this in relation to the closely-related phenomenon of optimism bias.23 When surveying individuals’ ability to assess the likelihood of their relationship breaking down, the median response was that the probability of other couples divorcing was 50%, but the probability that they personally would divorce was 0%. These cognitive limitations were acknowledged by the Law Commission in its 2014 report on marital property agreements:

‘Those who marry or form civil partnerships are adults and can take their own decisions, but it is a matter of experience that most people are willing to agree, when they are in love, to things that they would not otherwise contemplate.’24

Thus, negotiations may be affected if the parties themselves do not believe that the agreement will ever need to be enforced. They may be unable to make decisions when entering agreements that properly represent what is in their best interests. And they are also unlikely to be able to predict the future effect of their agreement, particularly if circumstances change during the marriage, which were not foreseen by the couple.

Unequal bargaining power

A third problem with the current law’s blinkered focus upon needs-based provision is that there is limited scope to appreciate how unequal bargaining power can suppress a party’s exercise of autonomy. Research has shown repeatedly that pre-nups are often one-sided, and that the spouse with more leverage is generally the spouse with property to protect.25 In these cases, the spouse at the short end of the power imbalance must show that she was either subject to duress, undue influence or improper pressure, or that she did not know or could not have understood the impact of what she was doing.26 The Supreme Court did not set a clear threshold for the type of pressure the court will consider, but it did point out that ‘unworthy conduct, such as exploitation of a dominant position to secure an unfair advantage’, could reduce or eliminate the weight to be attached to an agreement.27 While this can provide space for negotiation at FDR hearings, case law suggests that pressure and coercion can be difficult to establish in the courtroom.28 Thus, autonomy is often assumed when the courts are satisfied that the parties are appropriately informed and neither party has unlawfully been pressured into signing the agreement. This risks side-lining contextual factors, including how and why the agreement was made and changes in the power dynamic that occur during the relationship.29 But if autonomy is to be respected, this sort of assessment can help uncover whether both parties have been able to negotiate a nuptial agreement on a level playing field.

Under the current law, rather than focusing upon how the intentions and autonomy of the parties have changed over time, or how optimism bias or unequal bargaining power have rendered the parties’ exercise of autonomy defective, the court is concerned predominantly with whether proper provision has been made for the potentially economically vulnerable spouse. This is also reflected in the Law Commission’s 2014 Nuptial Agreements Bill, which according to Elizabeth Cooke, the Law Commissioner who led the project on marital property agreements, ‘builds on existing law and practice’.30 The Bill stipulates that an agreement could be set aside if the needs of the parties were not provided for, thereby replacing the Supreme Court’s test of attributing decisive weight unless ‘unfair’. But renewed calls to make nuptial agreements binding under legislation present a valuable opportunity to consider a different approach that departs from the current law. This can be found in the American Law Institute’s proposals.

American Law Institute proposals

The American Law Institute (ALI) is an independent organisation that reviews the law of the US and produces proposals recommending reform.31 It is therefore analogous to the Law Commission of England and Wales. When proposing reform of pre-nups, the ALI emphasised respect for the autonomy of the parties, while recognising broader contextual factors influencing the balance of power between the parties.32

One of the most pertinent ways in which it sought to do this was to propose giving the court a ‘second look’33 at a nuptial agreement in limited circumstances. If changes during the marriage mean the agreement ‘would work a substantial injustice’,34 the court would have discretion to set it aside, but in a way that does not disregard the parties’ autonomy.

This ‘substantial injustice’ safeguard ostensibly is not far removed from England and Wales, where an agreement can be varied or set aside if unfair. But there is a crucial distinction between these proposals and the law in this jurisdiction: not all cases would be eligible for consideration. Before the court could consider whether there has been substantial injustice, the party resisting enforcement must show that one or more of the following has occurred since the agreement was created:35

(1) more than a fixed number of years have passed (the ALI leaves this to be determined by the adopting jurisdiction, but gave a period of 10 years as an example, since the rationale is that the agreement’s terms are more likely to become redundant over the course of a longer marriage);

(2) a child was born to, or adopted by, the parties, who at the time of execution had no children in common;

(3) there has been a change in circumstances that has a substantial impact on the parties or their children, but when they executed the agreement, the parties probably did not anticipate either the change, or its impact.

By requiring one of these situations – which are tied to unanticipated change in circumstances – there is recognition that parties’ intentions can evolve and that the autonomy of one of the parties is likely to change with their different circumstances. In other words, these are situations that might have affected the parties’ exercise of autonomy had they known about them when they signed the nuptial agreement. Once one of these situations is proven to have occurred, the judge must consider whether enforcing the agreement in question would lead to substantial injustice.

As ‘substantial injustice’ is a vague term, the ALI set out the following guide to matters to be taken into account, so that the judge’s discretion would be more principled:

(1) the magnitude of the disparity between the outcome under the agreement and the outcome under otherwise prevailing legal principles;

(2) for those marriages of limited duration in which it is practical to ascertain, the difference between the circumstances of the objecting party if the agreement is enforced, and that party’s likely circumstances had the marriage never taken place;

(3) whether the purpose of the agreement was to benefit or protect the interests of third parties (such as children from a prior relationship), whether that purpose is still relevant, and whether the agreement’s terms were reasonably designed to serve it;

(4) the impact of the agreement’s enforcement upon the children of the parties.36

The ALI’s rationale for these provisions is that it recognises the ‘special difficulties’ associated with pre-nups, as well as providing greater scope to consider fairness and changed circumstances according to a standard of ‘substantial injustice’.37 Instead of presuming that autonomy has been exercised by the parties, these proposals provide scope to inspect the purpose and broader context of the agreement, and whether autonomy has been rendered defective.

This does not require the court to examine all nuptial agreements on divorce. Rather, the ALI has stated, it is ‘only a subset in which … difficulties are particularly likely’ that will attract attention.37 As a result, the ALI has asserted that its recommendations retain ‘considerable deference to contractual freedom’.39 Vitiation of a nuptial agreement must be justified by cognitive difficulties or changed circumstances, in other words, occurrences that render autonomy defective. In doing so, the ALI’s proposals are not a panacea for these problems. However, by placing the limitations of autonomy at the heart of its reform considerations,40 these proposals proffer an approach that differs in emphasis from the discretionary approach in England and Wales. In short, instead of being guided by general considerations of needs and fairness, the ALI proposals are linked directly to problems in the exercise of autonomy.

Looking to reform

Misconceptions about pre-nups and autonomy are rife. Research has shown that many potential spouses and civil partners do not understand precisely what a pre-nup is and how it operates.41 Those who do may be affected in other ways, including optimism bias, power imbalances and changes in circumstances, priorities and needs over the course of the marriage. These factors all impact how autonomy is exercised in the pre-nuptial and post-nuptial context.

If there is to be reform making nuptial agreements binding under legislation in England and Wales, now is the time to consider what respect for autonomy really looks like, and how it can best be facilitated in practice.42 While the ALI proposals are not a world away from the current approach in England and Wales (which the Law Commission’s Nuptial Agreements Bill would largely codify), the safeguards of the former focus less on needs-based provision and more on defects in the exercise of autonomy. Therefore, reform influenced by the ALI proposals could circumvent the artificial binary of needs and autonomy currently being grappled with by the court, whereby it risks accusations of paternalism and undermining autonomy when a nuptial agreement is not given effect on divorce.43

There is a further danger with codifying the court’s current approach in line with the Nuptial Agreements Bill. If nuptial agreements are binding unless needs are not provided for, the meaning of needs risks narrow interpretation, as seen in Mostyn’s comments in Cummings v Fawn in the first part of this article.44 It would be wholly undesirable to be left with a public welfare exception to the enforcement of nuptial agreements – one that looks only to absolute need and ignores parts of our system of financial remedies that are vital to ensuring fairness, such as redressing relationship generated disadvantage or recognising the impact of an abusive relationship upon needs.45

While the reform proposed under the Divorce (Financial Provision) Bill would make nuptial agreements more inflexible in the event of divorce,46 this would not promote autonomy. Clause 3 of this Bill provides that a nuptial agreement will be binding provided it complies with tick-box requirements, such as a 21-day cooling off period, proper disclosure, adequate opportunity to receive independent legal advice, and compliance with general contractual rules. These safeguards might help to facilitate the exercise of autonomy in some cases, but the Bill leaves no scope to recognise the defects that can occur with autonomy. If implemented, this Bill would create ironclad nuptial agreements that would uphold only an illusory version of autonomy. This risks ousting the promotion of other values within family law inextricably linked to fairness and equality on divorce.

And so, it is vitally important not to fall into the trap of equating a valid agreement with upholding the individual autonomy of the parties. Reform that takes on board the ALI’s recommendations could be one way of managing this tension that exists in our current law.