Birch v Birch

The husband and wife (as it will be convenient to describe them notwithstanding the grant of a divorce) entered into a consent order dated 28 July 2010, by which they compromised their claims against each other for financial orders on the basis of a clean break between them. The wife wished to continue to live with the two children of the family, namely a girl then aged ten and a boy then aged eight, in the matrimonial home, which was held in the joint names of herself and the husband subject to a substantial interest-only mortgage. Part of the order provided, by way of property adjustment, that the husband should transfer to the wife his legal and beneficial interest in the home subject to the mortgage. In return the husband was to be released from his covenants under the mortgage. His release from them was provided in undertakings given to the court by the wife which were recited as a prelude to the order. Thus she undertook at para 4.3 of the recitals to discharge all instalments payable under the mortgage; in any event to indemnify the husband against any liability under it; and to use her best endeavours to obtain the mortgagee’s agreement to release him from his covenants under it. Then, crucially, she undertook at para 4.4 of the recitals that, if the husband had not been released from his mortgage covenants by 30 September 2012, she would secure his release by placing the home on the market for sale and proceeding to sell it. The several other provisions of the order, including for the division between the husband and wife of responsibility for specified debts, are for present purposes irrelevant.

On 18 November 2011 the wife issued an application which she described as being to “vary” her undertaking at para 4.4. In her statement in support of it she explained that she had not been able to secure the husband’s release from his mortgage covenants and would not be able to do so by 30 September 2012. She said that, when she had given the undertaking at para 4.4, she had expected to be able to secure his release either by the provision instead by her brother or sister of guarantees to the mortgagee for the performance of her obligations under it or by her obtaining employment by reference to which she could persuade the mortgagee to accept her as the sole mortgagor; but that in the event neither the brother nor the sister had proved able to provide the guarantees and she had been unable to obtain employment. She explained that the two children were in fine schools in the vicinity of the home and that it would be gravely damaging to their interests for them to have to move home while still at school. In such circumstances she sought a “variation” of the undertaking at para 4.4. so as to postpone her obligation to secure the husband’s release from his covenants under the mortgage by sale of the home from 30 September 2012 until (as she soon made clear) 15 August 2019, being the date of their son’s 18th birthday.

In 2012 the husband countered by contending, through counsel, that the court had no jurisdiction to hear the wife’s application. He requested the court so to rule in the determination of a preliminary issue. In retrospect it is unfortunate that the court acceded to his request. As of now, in 2017, the merits or demerits of the wife’s application have never been ventilated. On any view, albeit subject to the extent of prejudice suffered by the husband as a result of remaining a co-mortgagor of the home, the wife’s application for postponement of the sale for seven years was highly ambitious even if there was jurisdiction to hear it. But, unsatisfactory though it is, she has secured postponement for five of those years solely as a result of the continuing litigation in respect of the preliminary issue. So she is now able to present her request for postponement as being only for the two final years, being (so she says) more important than ever in the interests both of the parties’ daughter who is about to embark on the second and final year of her A-level course and of their son who is about to embark on the first of them.

On 15 January 2014 District Judge Chesterfield, sitting in the Watford County Court, concluded that the court had no jurisdiction to hear the wife’s application and so ordered that it be dismissed. On 12 May 2014 HHJ Waller CBE upheld the district judge’s conclusion and dismissed the wife’s appeal against his order. On 31 July 2015 the Court of Appeal (Gross and Kitchin LJJ, and McCombe LJ who gave the main judgment): [2015] EWCA Civ 833; [2016] 2 FLR 467 held that there was jurisdiction to hear the wife’s application but that it was only a “formal” jurisdiction which existed only “technically”; that scope for its exercise was “extremely limited indeed”; and that, in the light of what the court had been told, there was no basis for its exercise upon the wife’s application. Against the Court of Appeal’s dismissal of her second appeal the wife brings a third appeal to this court.

All three lower courts adopted without demur the wife’s description of her application as being to “vary” her undertaking. But her description betrays a conceptual confusion which it is as well to dispel as this early stage. An undertaking is a solemn promise which a litigant volunteers to the court. A court has no power to impose any variation of the terms of a voluntary promise. A litigant who wishes to cease to be bound by her (or his) undertaking should apply for “release” from it (or “discharge” of it); and often she will accompany her application for release with an offer of a further undertaking in different terms. The court may decide to accept the further undertaking and, in the light of it, to grant the application for release. Equally the court may indicate that it will grant the application for release only on condition that she is willing to give a further undertaking or one in terms different from those of a further undertaking currently on offer. In either event the court’s power is only to grant or refuse the application for release; and, although exercise of its power may result in something which looks like a variation of an undertaking, it is the product of a different process of reasoning. In Cutler v Wandsworth Stadium Ltd [1945] 1 All ER 103 Morton LJ said at 105D-E:

In my opinion these proceedings have been bedevilled by a failure to distinguish between the existence of the court’s jurisdiction to release the wife from her undertaking (conditionally, on any view, upon her offering a further one in different terms) and the exercise of its jurisdiction. The preliminary issue has related only to its existence with the result that factors relevant to its exercise have not been the subject of investigation or argument. Nevertheless, influenced by a decision of the Court of Appeal in Omielan v Omielan [1996] 2 FLR 306 which will receive my careful consideration in paras 21 to 27 below, the lower courts have in my view looked over their shoulders at the ostensibly ambitious nature of the wife’s application (being one of the factors relevant to exercise of the jurisdiction); and they have deployed it as a basis for denying the existence of the jurisdiction or, in the Court of Appeal, as a basis for concluding that the jurisdiction was no more than formal and technical (which, irrespective of what in this context those adjectives precisely mean, seems tantamount to a conclusion that, for practical purposes, the jurisdiction does not exist).

In Russell v Russell [1956] P 283 the husband appealed against a judge’s refusal to release him from an undertaking that, unless he was out of work, he would not apply for a downwards variation of an order for maintenance in favour of the wife. Jenkins LJ said at 294:

In Kensington Housing Trust v Oliver (1997) 30 HLR 608 the Court of Appeal stressed the universality of the jurisdiction to grant release from an undertaking. A tenant had caused flooding of flats underneath her flat. As a result the landlord had obtained an order for possession of it but it had undertaken to the court to offer the tenant specified alternative accommodation at basement or ground floor level before seeking to enforce the order. When, following further flooding in breach of a reciprocal undertaking which the tenant had given to the court, the landlord applied for release from its undertaking, a recorder held that the court lacked jurisdiction to grant release from it. The Court of Appeal held, however, that the recorder did have jurisdiction to grant release and that the court should itself exercise it. Both Butler-Sloss LJ at 612 and Judge LJ at 616 quoted the statement of principle by Jenkins LJ in the Russell case, set out above. Butler-Sloss LJ held at 612 that the principle applied to all civil litigation and at 613 that the fact that the undertaking was recorded as a prelude to a consent order was irrelevant. Judge LJ held at 617 that the principle was not confined to matrimonial proceedings. Thorpe LJ held at 614 that it applied in the fields both of family law and of civil law.

Other than in the decisions of the Court of Appeal in Omielan, which pre-dated the Kensington Housing Trust case, and in the present case, the universality of the jurisdiction to grant release from an undertaking has, it seems, never been doubted. But, outside the realm of undertakings given in proceedings for financial orders, there has been debate about the criteria by which the jurisdiction should be exercised. For, in the Kensington Housing Trust case, the Court of Appeal, applying the reference by Jenkins LJ in the Russell case to the interests of justice, appeared to hold that the sole criterion was whether it would be just to grant release. Giving the leading judgment, with which both Thorpe and Judge LJJ agreed, Butler-Sloss LJ said at 613:

In Mid Suffolk District Council v Clarke [2006] EWCA Civ 71, [2007] 1 WLR 980, the Court of Appeal considered whether it was correct to say that the jurisdiction fell to be exercised solely by reference to what was just. The council had taken proceedings against a farmer whose production of swill, for feeding to pigs, was emitting a smell which local residents found scarcely tolerable. Rather than suffer the making of an injunction against him, the farmer had undertaken not to cause a public nuisance at his premises. The duration of the undertaking was not limited in time. Thereafter the council took committal proceedings in which the farmer admitted that he had broken the undertaking by the continued emission of smells. The judge duly fined him but then decided to release him from part of his undertaking by limiting its further duration to two years. The Court of Appeal allowed the council’s appeal. Lloyd LJ noted at para 20 that, as both parties had accepted, the court had jurisdiction to grant release from an undertaking, in whole or in part, and that the jurisdiction was discretionary. Both he and Buxton LJ then addressed the criteria by which the jurisdiction should be exercised. Having considered the remarks of Butler-Sloss LJ in the Kensington Housing Trust case, both Lloyd LJ at para 17 and Buxton LJ at paras 55 and 56 held that it was no doubt necessary for a grant of release to be just but that it had also to be predicated on a significant change of circumstances, which in the present case did not exist.

It is, I suppose, inconsistent with the admitted existence of a discretionary jurisdiction to say that it can never be exercised unless a particular fact, such as a significant change of circumstances, is established. If a discretionary jurisdiction is shackled in that way, the result is, instead, that the jurisdiction does not even exist unless the fact is established. For all practical purposes, however, the Court of Appeal in the Mid Suffolk case gave valuable guidance. I summarise it as being that, unless there has been a significant change of circumstances since the undertaking was given, grounds for release from it seem hard to conceive.

By reference only to the reasoning in the cases of Russell, Kensington Housing Trust and Mid Suffolk, one would confidently conclude that there was a full jurisdiction to hear the wife’s application for release, albeit that its exercise in her favour would be likely to attract lively debate.

There is, however, a completely different line of reasoning. In my view it neatly leads to the same conclusion although, for reasons which I will explain, it led the courts below to the opposite conclusion. It relates to sections 24A and 31 of the Matrimonial Causes Act 1973 (“the Act”).

Section 24A provides:

Section 31 provides:

Both the district judge and the circuit judge held that the wife’s undertaking in para 4.4 of the recital could instead have been framed as an order for the sale of the home pursuant to subsection (1) of section 24A of the Act, albeit subject to a direction pursuant to subsection (4) that it should not take effect unless by 30 September 2012 the wife had failed to secure the husband’s release from his covenants under the mortgage. In the Court of Appeal neither party argued that the two judges had fallen into error in that respect. Nevertheless the court held that the undertaking could not have been framed as an order under section 24A. In paras 29 and 34 of his judgment McCombe LJ based his reasoning in this respect on the admitted fact that the wife’s undertakings in para 4.3 of the recital, namely to discharge the mortgage instalments etc, could not have been framed as orders. He held that, since para 4.4 was linked to para 4.3, it followed that the undertaking in para 4.4 was likewise not susceptible of being framed as an order. Kitchin LJ agreed with the judgment of McCombe LJ, as did Gross LJ although he expressed doubts on this point. It is important to note, however, that at para 41 McCombe LJ observed that, even if he had held that the undertaking in para 4.4 could have been framed as an order for sale, he would still have concluded that in the circumstances the jurisdiction to release the wife from it was no more than formal and technical.

Counsel for the husband realistically accepts that he cannot defend that part of the Court of Appeal’s reasoning in which it held that the undertaking in para 4.4 could not have been framed as an order for sale. With respect, the fact that the undertakings in para 4.3 could not have been framed as orders in no way precluded the making of a conditional order for sale in the terms set out above.

In circumstances in which an undertaking could have been framed as an order, it would be illogical for answers to questions about the existence and exercise of the jurisdiction to grant release from it to be different from answers to questions about the existence and exercise of the jurisdiction to vary any such order. In L v L [2006] EWHC 996 (Fam), [2008] 1 FLR 26, the husband had accepted an obligation to make periodical payments to the wife but the obligation had been expressed as an undertaking on his part rather than as an order by consent for periodical payments pursuant to section 23(1)(a) of the Act. Such an order would have been variable under section 31(2)(b) of the Act. The husband subsequently sought to be released from his undertaking on the basis that, in return, the court would make an order for periodical payments against him in favour of the wife on different terms. But he cast his application for release as an application for variation under section 31(2)(b). Munby J declined to strike his application out. He held at para 113 that his entitlement to apply for variation “is not in any way affected either by the fact that the order was a consent order or by the fact that the relevant provisions are contained in undertakings rather than in the curial part of the order”.

In the present case, therefore, the equivalence of the wife’s undertaking at para 4.4 with an order for sale under section 24A of the Act, variable under section 31(2)(f) of it, seems clearly to confirm the existence of the court’s jurisdiction to hear her application for release from it.

But this conclusion the husband disputes. He does so by reference to the decision of the Court of Appeal in the Omielan case cited at para 6 above.

In the Omielan case the husband and wife entered into a consent order which related in particular to their home, vested in their joint names, in which the wife wished to continue to reside with the children of the family. Part of the order was a property adjustment order: it was, specifically, a variation of settlement order under which the proportions of the beneficial ownership of the home were recast so as to become 25% for the wife and 75% for the children. But there was also an order for sale of the home under section 24A of the Act. It provided that the home be sold but only on the occurrence of any one of four trigger events, including the event that the wife had cohabited with another man for at least six months. Shortly after the order was made the husband and wife executed a deed of trust under which they declared themselves to be trustees of the home on the above terms. Subsequently, on discovering that the wife had cohabited with another man for at least six months, the husband applied for an order that the sale of the home should take place at once; and the wife countered with an application under section 31(1) and (2)(f) of the Act for the order for sale of the home to be varied so as to postpone it until the youngest child, then aged nine, attained the age of 18.

In the Omielan case the Court of Appeal was clearly correct to allow the husband’s appeal against a judge’s refusal to dismiss the wife’s application for variation. There were patently no grounds for exercising the jurisdiction to vary the order for sale. In a judgment with which Butler-Sloss and Peter Gibson LJJ agreed, Thorpe LJ pointed out at 313 that the vested beneficial interest of the children in reversion had, once the wife had cohabited for six months, become an interest in possession; and that she was seeking to put it back into reversion. But, coupled with that feature, there was (so I would add) the absence of any evidence of a relevant change of circumstances since the order was made: for the fresh circumstance of the wife’s cohabitation was a specified ground for triggering the sale and could hardly be deployed as a ground for further postponing it.

The trouble is, however, that, instead of allowing the husband’s appeal in terms of a refusal to exercise the jurisdiction to vary the order for sale, the Court of Appeal in the Omielan case preferred to hold that the jurisdiction to vary it did not exist. In this regard Thorpe LJ referred at 310A to 311G to the following three authorities as “most in point”.

The first was Dinch v Dinch [1987] 1 WLR 252, in which the House of Lords made no reference to section 24A of the Act.

The second was Thompson v Thompson [1986] Fam 38, in which an order had been made in 1981 for the home not to be sold until the youngest child had attained the age of 17 “or further order”. The Court of Appeal inevitably allowed the appeal of the wife, who was living in the home with the children, against a judge’s determination that he had no jurisdiction to entertain her subsequent application for an order for its sale prior to that child’s 17th birthday; and the court remitted her application for the jurisdiction to be exercised. The court held at 53 that, even though the property adjustment order had been made before section 24A came into force, it provided the vehicle by which the wife could apply for the “further order” which the property adjustment order had envisaged.

The third was Taylor v Taylor [1987] 1 FLR 142, in which an order had been made for the wife to have exclusive occupation of the home and on its sale to receive 40% of the net proceeds. A recorder had acceded to a subsequent application by the husband under section 24A for the immediate sale of the home. The Court of Appeal upheld the wife’s appeal against the order for sale. But it expressly rejected her contention that the recorder had had no jurisdiction to make an order under section 24A. She alleged that it constituted an impermissible variation of such part of the property adjustment order as had conferred on her a right to occupy the home. But the court held that the effect of any order under section 24A on the property adjustment order was relevant to the discretionary exercise of the jurisdiction, which the recorder had not properly conducted, rather than to the existence of the jurisdiction conferred by the section when literally construed: see the judgments of Sir John Arnold P at 144F and 146D-F and of Ralph Gibson LJ at 147B-D.

In the Omielan case Thorpe LJ proceeded to articulate a statement of principle which in my view is hard to square with the decisions in the Thompson case and, in particular, in the Taylor case. For there is an obvious correlation between the jurisdiction to vary an order for sale and the jurisdiction to make such an order in the first place; and his narrow view of the former was predicated on his narrow view of the latter. He said at 312H:

Prior to stating the above Thorpe LJ had stressed the fact that, for strong reasons of public policy, orders for property adjustment and (subject to para 26 below) for payment of a lump sum are not variable under section 31 of the Act. He was understandably concerned that the variation of an order for sale might, by the back door, amount to a variation of the property adjustment order which it had accompanied. It was, however, bold for the Court of Appeal to hold that, when Parliament had provided an ostensibly unrestricted jurisdiction to vary an order for sale, the jurisdiction was nevertheless restricted; and equally bold for it to hold that the jurisdiction was restricted by reference to territories, namely that it was restricted to the “territory” of the order for sale as opposed to the “territory” of the property adjustment order.

It is worthwhile to note that an order for payment of a lump sum is occasionally variable even if, as is likely, the variation will directly prejudice the interests of the payee. Thus section 31(2)(d) of the Act expressly empowers the court to vary an order for payment of a lump sum by instalments. In the words of Bodey J (with whom Schiemann and Sedley LJJ agreed) in Westbury v Sampson [2001] EWCA Civ 407, [2002] 1 FLR 166, at para 18, the subsection “not only empowers the court to re-timetable / adjust the amounts of individual instalments, but also to vary, suspend or discharge the principal sum itself, provided always that this latter power is used particularly sparingly, given the importance of finality in matters of capital provision”. Indeed, although there is no jurisdiction under section 31 to vary an order for a single payment of a lump sum, there is even an inherent jurisdiction in the court to direct a modest extension of the time for its payment provided in the order: Masefield v Alexander (Lump Sum: Extension of Time) [1995] 1 FLR 100.

With respect to the Court of Appeal in the Omielan case, I cannot subscribe to its determination of that appeal by reference to the non-existence of the jurisdiction to vary the order for sale rather than to a refusal to exercise it. Where Parliament has conferred jurisdiction on a court, I see no scope for a court to say that part of it does not exist. Nor in my view can the existence of jurisdiction sensibly be left to a demarcation of territories. I find the concept of different territories hard to apply to the terms of a financial order, which are usually interlinked and which, in the case of an order for sale under section 24A of the Act, can be made only as an accompaniment to an order for property adjustment or lump sum payment. I wonder whether underneath the concept of different territories lies no more than a rough and ready distinction between an apparently minor variation and an apparently major one. At all events the demarcation of territories within the order is no proper criterion for identifying the existence of a jurisdiction.

Understandably the lower courts felt the need to identify the territory of the wife’s application for release; and each concluded that it fell within the territory of the order for transfer to her of the equity in the home. There is no need to seek to determine whether their conclusion was correct in this regard. I would suggest, however, that perhaps it merited greater consideration than they saw fit to give to it. For insight into the territory of the wife’s application might have been gained by seeking to identify the effect on the husband of the proposed postponement of the sale of the home. Since any resulting prejudice to the husband is on any view likely to be a highly relevant circumstance in the future disposal of the wife’s application, it may be helpful to make three points in this regard. The first is obvious: the situation in the present case is unlike the more usual situation in which the respondent to the application for postponement of the sale (or a third party, or third parties as in the Omielan case) has any interest in its proceeds. So the inquiry into prejudice to the husband requires focus elsewhere, in particular into such prejudice as might flow from his remaining liable under his mortgage covenants. Perhaps the most obvious prejudice of that sort would be a call to pay under the mortgage. But, although not noticed in any of the judgments below, the second point is that, with the assistance at times of income support, the wife has discharged all the mortgage instalments in accordance with her undertaking, with the result that the husband has not been, and appears unlikely to be, called upon to make any payment under the mortgage notwithstanding his continued liability under it. That leaves - third - what may or may not prove to be substantial resulting prejudice in the form of a reduced capacity to obtain another mortgage loan for deployment in the purchase of a home for himself. In 2012 he disclosed “mortgage promises” from the Halifax of an advance of £180,000 in the event that he was released from his existing mortgage covenants and of only £117,000 in the event that he was not released from them. It follows that no inquiry has yet been conducted into what, even if those figures are reliable, he might have achieved with the larger advance but could not achieve with the smaller one. Nor has an inquiry yet been conducted into the likelihood that, if released from his covenants, he would have purchased, or would now purchase, a home of his own in the light of his previous cohabitation for about two years with a partner in her own home and of his present cohabitation with a second partner in rented accommodation.

I propose that this court should allow the wife’s appeal; should hold that jurisdiction exists to hear her application; and should remit to HHJ Waller, in the light of his past experience as the Senior District Judge of the Principal Registry, Family Division, the difficult and urgent inquiry into whether the jurisdiction should be exercised. In the light of the equivalence of the wife’s undertaking with an order for sale, his inquiry will be conducted in accordance with section 31(7) of the Act, set out in para 15 above. He will give first consideration to the welfare while minors of the two children; but it is a consideration which may be outweighed by other factors. He will have regard to all relevant circumstances including in particular, so I suggest, whether the wife can establish a significant change of circumstances since her undertaking was given and whether, and if so to what extent, the husband has suffered, and is likely to continue to suffer, prejudice by remaining liable under his mortgage covenants.

The husband, so this court was told, recently made an open offer to the wife that he would accede to the further postponement of the sale of the home until their son’s 18th birthday on condition that she undertook to pay him 30% of the net proceeds of the sale. The wife has rejected it. It is not for this court to judge whether it would be appropriate to attach a condition of that character to the release of the wife from her undertaking. It seems that the parties cannot even agree upon the value of the home and thus upon the figure which 30% of the net proceeds would represent. If, however, the court finds that the husband has suffered prejudice as a result of the delay in selling the home, and/or would be likely to suffer prejudice as a result of any further delay, it is possible that it might favour compensating him by asking the wife to make provision for him out of the ultimate net proceeds as a condition of release.

Since drafting the above, I have had the benefit of reading in draft the judgment of Lord Hughes. My respectful comments on it are as follows:-

I acknowledge the difficulty in some cases of exercising the power under section 31 of the Act to vary an order for sale under section 24A in the light of the absence of any power under it to vary an order either for property adjustment under section 24 of it or (subject to para 26 above) for payment of a lump sum under section 23(1)(c) of it.

The variation of an order for sale, which is no more than an order for the conversion of one form of property into another, can never directly affect the allocation of property between the husband and wife in an order for property adjustment or lump sum.

But it may have an indirect effect on the allocation, in particular in a case (unlike the present) in which a postponement of the date of sale would postpone a party’s receipt into his possession of the capital allocated to him. Another example (also unlike the present) would be an order that the sale should proceed at a specified price.

In determining whether, and if so how, to exercise its jurisdiction to vary an order for sale, a court should place in the balance any indirect effect of the suggested variation on the order for property adjustment or lump sum; and the effect might in some cases precipitate the dismissal of the application. But in my view there is no way in which this proposition can properly be expanded.

In particular I, for my part, cannot subscribe to the “acid test” articulated by Lord Hughes in para 54, namely “whether the application is in substance to vary or alter the final [capital] order or is to support it by working out how it should be carried into effect”. This test is a reiteration in different words of the test of “territories” suggested, albeit as a demarcation of the jurisdiction itself, in the Omielan case.

If, when applying the “acid test”, the court concludes that the application is in substance to vary the final capital order, it must, so it is said, be dismissed without wider inquiry. But this would run counter to Parliament’s instruction in section 31(7) to have regard to all the circumstances. These include the welfare of the two children in the present case, to whom therefore Lord Hughes makes no reference even though their welfare is supposed to be the court’s first consideration.

Furthermore the “acid test” would in my view be difficult to apply in practice and would generate collateral dispute. Take this very application. Lord Hughes has no doubt that is in substance an application by the wife to vary the final capital order. But, in circumstances in which the husband stands to receive nothing upon the sale, I myself do not so regard it.

The wife’s application was indeed at first to postpone the sale by seven years. But five of those years have been spent in addressing the husband’s denial of any jurisdiction to entertain it; and, were this court to determine that there is no need to remit it for hearing because it is now bound to fail, it should confront the reality that the application is now for a postponement of two years.

An informal indication of the likely outcome of the applications for financial orders, given by the district judge prior to hearing evidence, should not carry weight sufficient to figure in this court’s analysis.

The court proceedings to date have never progressed to the point at which the wife has been able to present her case. We cannot assess the strength of it.

It seems odd that this court should pre-empt the conventional inquiry so as to leave the wife obliged to effect the immediate sale of the home in circumstances in which the husband has himself conceded its further postponement, albeit subject to a condition which it may or may not prove to be appropriate to put to her as the price of release.

I respectfully agree with Lord Wilson’s insightful analysis of

I also agree that the present order, whilst framed by way of undertakings, could equally have been structured as an order for sale under section 24A, subject to one or more conditions pursuant to section 24A(4) and/or consequential or supplementary provisions under subsection (2). Exactly the same effect could have been achieved by such route(s).

It follows that the remaining question in this case is: what are the principles for the exercise of the jurisdiction to vary an order for sale under section 24A of the Matrimonial Causes Act, or its equivalent achieved via undertakings? I do not think that it can be right to leave the exercise of that jurisdiction in quite the place that Lord Wilson does. It is one thing to say that the jurisdiction to vary an order for sale under section 24A is given by section 31(2)(f), as clearly it is and always has been. It is another to say that it follows that the exercise of that jurisdiction is open ended. In particular, to say that an order for sale under section 24A (or an undertaking to like effect) can be modified whenever the applicant demonstrates a significant change of circumstances since the order was made is, as it seems to me, too wide a gateway for variation. It would be likely to intrude upon an underlying clean break where, as is very often the case, that is what the order for sale is designed to serve. It may also operate as a regrettable deterrent to the inclusion in court orders of a provision for deferred sale, when that kind of provision is of real practical use to parties and to courts.

It is trite to record that the scheme for financial provision after divorce contained in the 1973 Act seeks to combine two features. First, it aims to give the courts as flexible a set of powers as is practicable to re-organise the financial affairs of parties when their marriage or partnership has collapsed. Such flexibility is necessary in order to achieve justice between the parties in the very wide range of different factual circumstances which may exist. It is needed to protect the more disadvantaged of the parties. It is particularly called for because whilst the marriage was successful it is very likely that many, perhaps most, couples will have treated their finances with great informality, and with much less regard for who owns what and for the source of expenditure than would be the case as between partners engaged in a commercial venture. Second, however, the Act aims to achieve finality in the re-organisation of financial affairs when it can. That means avoiding continuing financial dependence between those who are now not happy together if this is not unavoidable. It also means putting as early an end as one can to litigation between them. This last is a goal of general application to all litigation but it is particularly important in the context of soured relationships, where the tendency to continue to ventilate old disputes may be especially strong and where the ordinary constraints of pragmatism are especially likely to be overcome by emotion. If not thus tempered by finality, the court’s protective flexibility can easily become an unwitting tool for prolonged and painful litigation.

The Act thus deals both with income provision, which is generally of necessity by way of continuing orders, and with capital or property allocation. Orders for continuing provision are necessarily ones which the court needs power to vary if circumstances change. The payer’s income may rise or fall; the recipient’s needs may wax or wane. Orders for capital or property allocation require no such power to vary. For the latter, the aim to bring finality prevails. A “clean break” is to be achieved where it can, although obviously it is not always possible. For this reason, the broad scheme of the Act has been for 40 years that continuing orders are subject to a power to vary, but that orders for capital provision are not. Orders for periodical payments (section 23(1)(a), (b), (d) and (e)) can be varied: see section 31(2)(a) to (c). Property transfer orders (section 24) cannot. Lump sum orders (section 23(c) and (f)) cannot ordinarily be varied, but only if they are given a continuing element by way of direction for payment by instalments (see section 31(2)(d)).

The House of Lords confirmed as long ago as Minton v Minton [1979] AC 593 that this was the scheme of the Act. That case concerned an application for periodical payments made by a spouse after a previous order by consent had settled her capital claims by transfer to her of the former matrimonial home and had provided for nominal periodical payments only until that transfer was accomplished, whereupon they were to “cease”. There was thus no continuing order to be varied, and the House held that there was similarly no jurisdiction to entertain a second application after the first had been dismissed or, as there, ordered by consent to come to an end. At 608E Lord Scarman gave expression to the principle in terms which are now very well known:

Subsequently, this scheme of the Act was reinforced by provisions encouraging clean breaks, where possible, also in relation to periodical payments. Section 25A, inserted by the Matrimonial Proceedings and Property Act 1984, contains three subsections with this design. By subsection (1) the court must consider whether it is appropriate so to use its powers to make financial provision orders in such a way as to terminate the financial obligations of the parties towards each other as soon as just and reasonable. Subsection (2) requires the court to consider whether any periodical payments order ought to have a definite term set to it, to enable the recipient to adjust to termination of dependence on the other party. Subsection (3) spells out the power of the court to dismiss an application for periodical payments and to couple with the dismissal a direction that no further application may be made. It is well-established law that if an application for financial provision is dismissed, no subsequent application can be made for that provision: L v L [1962] P 101 and Minton v Minton. In Miller v Miller [2006] UKHL 24; [2006] 2 AC 618 at para 130 Baroness Hale aptly described these provisions as devised to “encourage and enable a clean break settlement”.

The question raised by the present case, as also by other situations, is in what circumstances is an order to be treated as in substance a continuing one, subject to the fullest power of variation in the interests of flexibility, and when is it to be treated as a clean break order, which should be final? That involves considering the nature and purpose of an order for sale made under section 24A.

The power in section 24A to make an order for sale was introduced into the 1973 Act by the Matrimonial Homes and Property Act 1981 following the recommendation of the Law Commission (Law Com No 99, 13 February 1980) Family Law Orders for sale under the Matrimonial Causes Act 1973. The Commission’s report made clear why this was done. It was that sometimes when operating the 1973 Act courts wished to order the sale of an asset in order to facilitate a financial provision order of a capital nature. There had been debate whether it was or was not necessary to employ the rather cumbersome machinery of a separate application under section 17 of the Married Women’s Property Act 1882 (see for example Ward v Ward and Greene [1980] 1 WLR 4), but in any event that kind of application was probably not available when the asset to be sold was in the sole beneficial ownership of one party, so that some different device would have to be found to insist on a sale. The new provision was designed to put the power beyond doubt and to make it available to the court whether a party asked for it or not: see section 5 of the Commission’s report. At para 9 the Commission gave examples of the kind of situation in which an order for sale is useful. Chiefly, it envisaged it being used when an asset needed to be sold to enable a lump sum order to be satisfied, particularly if the money was to come from realisation of the matrimonial home. Another case was that of a spouse who ought to be given a share of capital assets in a family business but had no claim to run it; in such a situation a transfer of shares would not achieve the aim but an order that some of them be sold and the proceeds paid over would. These examples still hold good, indeed the more so now that the principles of sharing and compensation are recognised (see Miller). The Commission summarised its proposal at para 8:

The key characteristic of the order for sale is thus that it is ancillary to a capital order. It is an aid to carrying such an order out. The description of section 24A by Thorpe LJ in Omielan v Omielan [1996] 2 FLR 306 at 312 as “purely procedural” may possibly involve modest oversimplification. The section gave a power which in some cases went somewhat beyond what might have been achieved by other routes such as the Married Women’s Property Act; that was the whole point. But that description was nevertheless apt and was borrowed from the judgment of Oliver LJ in Thompson v Thompson [1986] Fam 38 at 53B.

In Thompson a property adjustment (variation of settlement) order under section 24 had been made, by consent, before the commencement of section 24A. It had modified the trust for sale of the jointly owned matrimonial home by providing, in the familiar Mesher v Mesher [1980] 1 All ER 126 type form, that the wife and children should occupy it and that it should not be sold until the youngest child reached the age of 17, or completed his education, “or further order”. Quite soon afterwards, and well before the child had reached 17, the wife, who wished to move house, sought an order for earlier sale. It was the foundation of her argument that her application was for an order “working out or giving effect to the original order” but did not seek a variation of its substance (see 40G). The County Court judge had felt he had no jurisdiction to entertain the application because it amounted to varying a section 24 order. The Court of Appeal held that he had, because it did not.

The Court’s primary decision was that independently of the advent of section 24A, the order sought by the wife did not amount to an impermissible variation of the original order and that there was, accordingly, no jurisdictional obstacle to adjudicating upon it. Oliver LJ was at pains to formulate the question in the case as whether an application made under the liberty to apply reserved in a common form Mesher order “is an application to vary the order, or an application for the working out of the order” (46H). He held that whilst such an application might amount to an impermissible variation, it did not necessarily do so. The question was, in each case, which it was. He held, plainly correctly if I may say so, that orders deferring sale of jointly held property, in the common Mesher form, have an obvious need for scope to adjust them to work out the order. Whilst an application further to delay sale would, he held, ordinarily amount to an impermissible variation, an application for an earlier sale need not do so, and often would not. He instanced examples such as the resident spouse going bankrupt, or wishing to emigrate, or one of the residents becoming incapacitated. It would no doubt have been different if it had been the husband who had sought an earlier sale, thus significantly altering the protection given to the wife and children by the original order; that did not arise, but would, as it seems to me, plainly be the kind of application which would be treated under the approach of Oliver LJ as in substance one to vary the final property adjustment order, and as such one which therefore ought not to be entertained.

It was necessary to address the then new section 24A only because it was argued that, notwithstanding the stipulation in the consent order for liberty to apply, the only source of jurisdiction to direct the husband to concur in the sale was found in the non-matrimonial-relief framework of section 30 Law of Property Act 1925, whilst the value of the house was outside the financial limits within which such power was available to the County Court. Concluding that that argument was correct, the court went on to hold that the new section 24A power to order sale was the kind of procedural provision which could be exercised in a case where the substantive order had been made before its commencement, without improper retrospectivity. It was in this context that Oliver LJ observed at 51H-52A that if the original order had been made after the commencement of section 24A there could have been no question of jurisdictional limits (ie the financial limits on the County Court). The jurisdiction being spoken of was a jurisdiction to make a new order for sale; variation of an order for sale which had already been made did not arise. There is accordingly nothing in the judgments in Thompson inconsistent with the approach subsequently adopted in Omielan, providing that the latter must be seen correctly as a statement of the basis on which a jurisdiction is to be exercised, rather than as defining the existence of the jurisdiction. On the contrary, the whole basis of Thompson was to identify the difference between substantive variation of the original order, which is not permitted, and further order to work out or give effect to the original order, which is. What the Court of Appeal subsequently held in Omielan is that exactly the same principles still need to be applied when considering section 31(2)(f).

Thompson was applied in Taylor v Taylor [1987] 1 FLR 142. The original order was again made before the commencement of section 24A. It was not a common-form Mesher order, because it fixed the beneficial interests of the spouses in the former matrimonial home, charged the property with the wife’s two-fifths interest, and gave her the occupation of it, but said nothing about whether or when it could or should be sold and the interests realised. The husband applied, something over 11 years afterwards, for an order for immediate sale. The Recorder had made such an order, ostensibly taking the view that such gave effect to the original order, but he had heard no evidence and given no consideration to any competing argument. The wife’s appeal was thus allowed and the case was remitted for re-hearing. Her additional argument that there was no jurisdiction to make a section 24A order for sale in relation to a pre- commencement original order failed by the application of Thompson. Giving (extempore) judgment in the Court of Appeal, Sir John Arnold P expressly reserved any question of the basis on which the jurisdiction to make a (subsequent) section 24A order ought to be exercised. There was likely to be in that case an open question whether the original order, which Ralph Gibson LJ described as poorly drafted, had been intended to be a Mesher-type order with sale after some deferred period contemplated, or had been meant to give the wife indefinite occupation of the house; thus there was an open question whether, in the terms used in Thompson, the application was to work out and give effect to the original order or was to vary it. Referring back to the approach in Thompson, the President said this of the argument that there should be no order for sale because the application was made not by the wife, who had the right of occupation, but by the husband who sought to bring that right to an end:

Similarly, Ralph Gibson LJ, agreeing, summarised his conclusion as follows at p 147:

As Lord Wilson says at para 24, there is an obvious correlation between the power to make a section 24A order for sale subsequently to the original order and the power subsequently to vary one made in the original order. The same principles ought to apply to the exercise of each of these powers. Both in Thompson and in Taylor the acid test was seen to be whether the power was being invoked to give effect to, and carry out, the original order, or was impermissibly to vary it.

Thompson was also applied by the House of Lords in Dinch v Dinch [1987] 1 WLR 252. That was a case of a common-form Mesher-type order deferring sale of the former matrimonial home during the minority of the children and giving the wife occupation meanwhile. When the stipulated time arrived and the husband sought sale, the wife, who had powerful and legitimate complaints that the husband had failed to comply with periodical payments orders in her favour and had caused her some hardship, responded with counter-applications for further postponement of sale, for a further transfer of property order in relation to the house, and for a lump sum. Those applications, which had been granted in part in the Court of Appeal, failed on the principles explained in Minton and similar cases, as well as on the basis of Thompson. They failed despite very considerable sympathy for the wife, who had a strong case on the general merits. There was no reference to section 24A, perhaps surprisingly, but perhaps because the original order had been pre-commencement and the husband’s entitlement to sale was sufficiently demonstrated by it; he needed no new section 24A order. But the principled approach to applications which are in substance to vary a final capital settlement was plainly stated by the House. Lord Oliver, with whom all their Lordships agreed, said this (at p 263)

These three cases of Thompson, Taylor and Dinch were cited to the Court of Appeal in Omielan. As Thorpe LJ said, giving the judgment of himself and of Peter Gibson and Butler-Sloss LJJ, none of the three was directly in point upon section 31(2)(f). But the principles which they enunciated were relied upon by that court for its decision. For the reasons largely set out above, I think that they were relevant, although what they led to was not the absence of jurisdiction, but clear principle on which the jurisdiction should be exercised.

Given the terms of section 31(2)(f), it is impossible to say that there is no power to vary a section 24A order such as the one in the present case. I agree that in Omielan the court fell into this error. The judgment contains the following (at 312):

But the force of those earlier cases, as also of the Minton line of authority, is not diminished by mischaracterising them as defining the jurisdiction rather than setting out the principles on which it is to be exercised. Thorpe LJ went on, correctly, to say that the principles underlying the earlier cases in relation to the making of (subsequent) section 24A orders must apply equally to applications to vary orders under that section. He said this:

It is of course true that on the particular facts of Omielan, there was the additional factor that the agreed (and ordered) trigger for sale had come about, and there had vested in the children beneficial interests in the former matrimonial home. Those were, in that case, powerful additional reasons why any attempt to vary the original order was doomed to failure. But the principle derived from the earlier lines of authority, and confirmed by the structure of the Act, does not depend on such additional factors. The outcome should, and clearly would, have been the same in the commoner case where the beneficial interests in the property are confined to the husband and wife and are already vested, and it should have been the same if the wife’s application to delay sale had been made before the trigger event rather than after it. In all these instances, the effect of variation would be to undo a final capital order, whether made by consent or not.

Any variation application under section 31 is governed, inter alia, by section 31(7). This requires the court to have regard to all the circumstances of the case, first consideration being given to the welfare while a minor of any child of the family under 18. A change in any of the factors which are made by section 25 material to the making of the original order is stipulated to be among all the circumstances of the case. There is nothing in this which requires the discarding of the principles, derived from Minton, Thompson, Taylor and Dinch, that the power to vary must not be used in a way which amounts to a disguised variation of a final capital order and an evasion of the rule that there can be no second application for capital orders, both of which are inherent, and in some instances explicit, in the terms of the Act itself. In particular, the interests of the children will have been first consideration at the time of the making of the original order: see section 25(1). Any order, whether following a contested hearing or a settlement, must be endorsed by the court, whose approval is no rubber stamp. The compatibility of any final order with the interests of the children can, and indeed must, be assumed. A change in the section 25 circumstances is but one of the factors relevant to an application to vary. It can perfectly properly become relevant only when there is no evasion of the rule against variation of final capital settlements. It is after all well established that, barring the kind of supervening event considered in Barder v Calouri [1988] AC 20, which is not suggested here, there is no power either to entertain a second application for a capital order, or to vary such an order, even if the calculations on which it was based have proved to have been misjudged by one party: see for example, amongst many other cases, the kind of events considered in Myerson v Myerson (No 2) [2009] EWCA Civ 282; [2010] 1 WLR 114.

There may of course be difficult marginal cases where there is scope for real argument whether the new order sought is (permissibly) to work out or enforce the original order or (impermissibly) to vary it. Some cases of deferred sale orders may well give rise to this difficulty. It would be a mistake here to attempt to anticipate the kind of situations which might fall on one side of the line or the other. So long as the principles are kept firmly in mind that (1) the section 24A order has always been devised as ancillary to a capital order and (2) that final capital orders cannot be varied in their substance (whether or not there is a change of circumstance), so that (3) the acid test is whether the application is in substance to vary or alter the final order or is to support it by working out how it should be carried into effect, then cases ought to be capable of resolution and there should be sufficient certainty for those, on both sides, who are subject to the orders.

If, on the other hand, the undoubted power to vary under section 31(2)(f) is taken to authorise a review of “all the circumstances” whenever there is a section 24A order, it is very difficult to see what weight judges who are asked to vary an order ought to accord to the fact that it was made as an ancillary to a final capital settlement. To say that that is one of the circumstances of the case poses, but wholly fails to answer, that question. It seems to me inevitable that if this is the position, the utility of an order for sale will be very much diminished. Orders for sale, and particularly orders for deferred sale, are not a universal panacea for matrimonial financial strife, but they can be very useful and are much resorted to. If a party is being advised in delicate negotiations about the settlement of the matrimonial finances, she will have to be told that if there is an order for sale, she is exposed to an application, at any time before the sale takes place, on the grounds at least of a suggested change of circumstances. Of course, such an application may fail. Sometimes its failure may be attended by adverse orders for costs. But the risk of further litigation, not necessarily confined to a first-instance hearing, will always be there. It is very likely that she will not wish to run it, but instead will stand out for something which does not involve an order for sale, for then the capital order will indeed be final. The extremely lengthy litigation which has ensued in the present case is no doubt exceptional, but all applications take time to be dealt with, and sometimes much time. The same considerations will, perhaps to a lesser extent, be present to the minds of judges considering the best form of order to make. It is not in the general interest of spouse litigants, whether potential occupiers of the property concerned or out-of-occupation spouses awaiting sale, that this constraint should operate on the ancillary orders devised by the Law Commission in 1980 and operated ever since on the basis explained in Thompson, Taylor, Dinch and (although there mischaracterised) in Omielan.

It should be recorded that Mr Hockman QC, for the wife in the present case, accepted the general principle that the power to vary a section 24A order should not be exercised in such a manner as to alter the substance of the original capital order. His case was confined to the argument that the wife’s present application would not do so, because the beneficial interests in the proceeds of sale were not affected, the husband having transferred the whole of his interest to the wife as part of the settlement. In my view, his concession was realistic, although if Lord Wilson is right it went further than was needed, but his argument was not.

Although there may be difficult marginal cases, I cannot for myself see that the present is one of them. This seems to me to be one where there is no doubt that the application is one which amounts to an attempt to vary, not to carry into effect, the originally agreed and court-endorsed order. Other forms of settlement were plainly available, and indeed an indication was given by the District Judge that sale of the house looked inevitable. As it was, by agreement, sale was directed at a fixed date, not on the occurrence of a trigger event, unless in the meantime the husband was released from his liability under the mortgage. In effect he gave up all interest in the house and in return was to be relieved by a specific deadline of a substantial outstanding liability in relation to it. An application to extend the date by a month or two because the finances to release the husband were unexpectedly held up might well have been an application to carry the order into effect. But the application was to extend the period from two years to nine. Even if the wife is able to point to some change in her circumstances, that is no justification for re-opening a final property settlement. Of course, if the correct rule of law is that the husband must face an open-ended application to vary, he will indeed have to address exactly what difference it has made to him, over the past five years, to remain liable on a substantial mortgage. But it takes little imagination to understand that for all except the very wealthy such a liability makes a real impact on one’s credit rating and ability to borrow, for any purpose, not necessarily for housing. It does not seem likely that the continuing liability has made no difference at all to his personal planning. In any event, whatever may have been the exact impact on him, the purpose of the original order was to give certainty and freedom of financial decision, in the face of competing cases about what should happen to the house and whether it was inevitable that the wife and children needed to move. The application has always been such as to remove the certainty which was the aim of the order. It seems to me that the Court of Appeal was right to hold that the application was bound to fail.

For those reasons, I would myself dismiss the appeal and uphold the decision of the Court of Appeal, although on somewhat different grounds from those stated by it.