Anisha Gupta & Shreya Banchhor

Abstract: The article examines how Section 13B(2) of the Hindu Marriage Act, 1955—which permits withdrawing consent before a divorce decree—can be weaponised to trap a spouse in a dead marriage as guidance on balancing opposing spouses’ autonomy remains unaddressed. To resolve this, the author proposes a three-stage test evaluating threshold conditions, motives behind withdrawal, and equitable remedies.

inTRODUCTION

Divorce by mutual consent under section 13B of the Hindu Marriage Act, 1955 (“HMA”) rests on the essential condition that both the spouses must want the marriage to end. They  must want it even when the court passes the decree. The Supreme Court has held that either spouse may withdraw consent at any point before the decree. The requirement exists to guard against hasty decisions. But it also gives each spouse a veto that can be used long after the time for reflection has passed. A spouse can sign a settlement, reaffirm it before a mediator and a High Court, and then simply not appear for the second motion. The only harm to the non-appearing spouse is a comparatively smaller risk of contempt. The other spouse is left with no route out of the marriage except fresh litigation.

In Shilpa Sailesh v Varun Sreenivasan, a Constitution Bench held that the Supreme Court may dissolve a marriage on the ground of irretrievable breakdown under Article 142(1) of the Constitution, even where one spouse opposes it. In 2026, the Supreme Court used this power to enforce a mediated settlement when one spouse withdrew consent after a settlement. But neither the Constitution Bench nor the later benches explained how the Court should weigh the autonomy of the spouse who withholds consent against the autonomy of the spouse who wants to leave. This article uses a live fact pattern and a line of jurisprudence, which has struggled to deal with the same question, to demonstrate why that question matters. Further, it proposes a balancing test to answer it.

Background

Imagine a scenario with the following facts. The parties are referred to as W (the wife) and H (the husband).

W and H lived together for a short span of time during their marriage. W worked from home and H worked on a hybrid schedule. W’s account is that her in-laws regarded her as contributing nothing, although she carried both the household work and her paid job, and that they controlled her salary. The marriage broke down soon after.

The parties signed a memorandum of understanding for divorce by mutual consent. Under it, H would pay W ₹60 lakh at the stage of the second motion. H did not appear for the second motion. W filed a contempt petition. H filed one of his own, alleging that W had spoken ill of him. The parties were then sent to the High Court’s mediation centre, where they reached a fresh settlement. W accepted ₹52.5 lakh, ₹7.5 lakh less than before. The High Court recorded the settlement and passed a consent decree. H again did not appear. The eighteen-month window for the second motion under section 13B(2) lapsed in July 2026.

Two features of these facts matter. First, the spouse withholding consent is also the spouse who must pay. Since payment was due at the second motion, his refusal to appear both keeps the marriage alive and suspends the payment. Second, W has made concessions twice, while H has given consent three times: in the memorandum, at mediation, and before the High Court. Under the present law, each withdrawal costs H little beyond the risk of contempt, while W stays married and unpaid.

The Statutory Position and Its Limits

Section 13B(2) requires the parties to make a second, joint motion not earlier than six months and not later than eighteen months after the first petition, “if the petition is not withdrawn in the meantime”. In Sureshta Devi, the Supreme Court read this to mean that mutual consent must continue until the decree. If one party withdraws, the court has no jurisdiction to pass a decree. In Hitesh Bhatnagar, it added that the eighteen-month period does not limit the right to withdraw; it exists only to ensure quick disposal. The lapse of the window in July 2026 therefore does not turn H’s silence into consent. If anything, it closes the section 13B route altogether.

What remedy does W have? The leading answer comes from the Delhi High Court in Rajat Gupta v Rupali Gupta. Here, the Division Bench held two things. First, a party may withdraw consent at any stage before the decree, even after undertaking to the court to file the second motion. Second, breach of an undertaking or a consent decree can still amount to civil contempt under section 2(b) of the Contempt of Courts Act, 1971. However, the Court also drew a firm limit. No order, even in contempt, may compel a party to consent to divorce. It also asked courts to record such settlements with penalty clauses and stated consequences of breach.

Earlier, in Avneesh Sood, a single judge had gone further. A party who gives an undertaking to obtain benefits under a settlement cannot, the Court said, walk away without facing contempt, save in exceptional circumstances arising after the undertaking. But the more recent decision in Chhavi Agarwal v Anurag Goel narrows even the contempt route. There, a Division Bench set aside a contempt finding against a wife who had resiled from a 56-clause settlement. It held that her failure to appear for the second motion could not ground contempt, because she was exercising a statutory right. It added that forcing her to perform the other terms on pain of contempt would compel her consent indirectly, which the court could not do directly. Additionally, given the quasi-criminal nature of contempt, wilful disobedience must be proved beyond reasonable doubt.

The result is that the law offers W two remedies, and neither are helpful for her. The first is the status quo: the settlement fails and both parties return to where they began, usually to fault-based litigation. The second is contempt, which may punish H but cannot dissolve the marriage. This demonstrates how a protection meant to stop a hurried divorce has become a way to prevent any divorce at all.

Article 142 After Shilpa Sailesh

The Constitution Bench in Shilpa Sailesh set out the limits of Article 142(1). According to the honourable Bench, the power may allow the Court to depart from existing procedure and even from substantive law. However, fundamental considerations of general and specific public policy must be respected and adhered to. The Court read specific public policy narrowly to mean an express, pre-eminent prohibition in a substantive law, instead of the requirements of a particular statutory scheme. On that reading, the second motion is a feature of the section 13B scheme, not a bar on the Court’s power. The Court accordingly held that it can dissolve a marriage on settlement without a second motion, guided by the factors in Amardeep Singh and Amit Kumar.

The Court went further on irretrievable breakdown. It held that it may dissolve a marriage that is “totally unworkable, emotionally dead and beyond salvation”, even over the other spouse’s objection. It called this a discretion, not a right, and listed factors: how long the parties lived together, when they last did so, the nature of the allegations, the proceedings between them, attempts at settlement, the period of separation, and the economic and social position of the parties, including any dependence. As a court of equity, it must also weigh the circumstances and background of the spouse who opposes dissolution. The Court also held that fault and blame do not fetter this power, and read down Hitesh Bhatnagar[1] and similar decisions to that extent.

Thus, only two limits of Article 142 remain. Only the Supreme Court holds the power, and a party cannot invoke it directly through a writ petition under Article 32 or 226.

A Balancing Test

The question is not whether a spouse may withdraw consent. Under section 13B, they may. The question is what weight that withdrawal should carry when the Supreme Court decides whether to use Article 142. I propose that the Court ask three questions in sequence to create a uniform and consolidated test.

Stage one: the threshold. Two conditions should be met before the Court weighs anything. First, the marriage must be dead in fact, judged by the Shilpa Sailesh factors. Second, the withholding spouse must have given consent in a form the law treats seriously: a settlement recorded by a court, or one reached in court-annexed mediation and confirmed by the court, ideally in the form Rajat Gupta requires. This second condition keeps the test narrow by becoming applicable only where the spouse has already told a court that the marriage should end.

Stage two: weighing the withdrawal. The Court should then ask why consent was withdrawn, and what the withdrawal achieves. Some reasons that deserve full weight include force, fraud, coercion or undue influence in obtaining the settlement, which the Court must in any case rule out before acting on a settlement. Failure by the other spouse to perform; and serious events after the settlement that change its basis are also important considerations. The withholding spouse’s economic position also matters. Shilpa Sailesh also asks whether the opposing spouse is dependent, and how the spouse seeking divorce will provide for them. A dependent spouse who withdraws to protect her security stands in a very different place from a spouse who withdraws to avoid paying.

Other features should count against the withdrawal. These include repeated consent before different fora; acceptance of benefits, or inducing the other spouse to perform. Use of the withdrawal to reopen agreed terms; fresh proceedings filed after the withdrawal as leverage; and failure to appear without any stated reason are also important to note. One further question deserves explicit weight: who holds the economic power? Where the withholding spouse is also the paying spouse, the veto does two things at once. It keeps the marriage alive and it postpones the payment. The protective purpose of section 13B(2) is then vitiated.

Stage three: shaping the remedy. If the balance favours dissolution, the decree should secure what the settlement promised. In Ashok Hurra, the Court made its decree conditional on payment of ₹10 lakh to the wife. In Salil Dhawan, it fixed alimony as a full and final sum and closed all pending proceedings. In Rathi, it held that a party who resiles should bear heavy costs. These tools let the Court free the spouse who wants to leave without letting the other spouse escape the financial terms he or she agreed to. If the balance does not favour dissolution, because the withdrawal was in good faith or the withholding spouse is vulnerable, the Court should decline to act and leave the parties to the statute.

Engaging with Counter-arguments

Four objections deserve an answer.

First, it may be said that Article 142 cannot override a statute that makes continuing consent a condition of divorce. Shilpa Sailesh has largely answered this. The Court held that the procedure under section 13B is not the kind of pre-eminent prohibition that limits Article 142, and that fault-based limits do not fetter the power. The test applies only where consent was recorded by a court and the marriage is dead in fact.

Second, it may be said that the test disregards the autonomy of the spouse who withholds consent. But the waiting period has a stated purpose. In Amit Kumar, the Court explained that it gives the parties time to reflect and prevents an impulsive end to the marriage. Where a spouse has reaffirmed consent before a mediator and a High Court, the period has done its work. Section 13B protects the right to change one’s mind. It does not protect a right to keep another person married as leverage. And the autonomy of the spouse who wants to leave is equally at stake.

Third, it may be said that contempt is a sufficient remedy. However, the Court has itself observed in Rajat Gupta and Chhavi Agarwal that contempt cannot compel consent. Contempt also requires proof of wilful disobedience beyond reasonable doubt. At best, contempt punishes the defaulting spouse. It does not release the other spouse from the marriage.

Fourth, it may be said that the permanence of marriage, and with it the right to refuse divorce, has historically protected wives. That concern was real when most wives depended on their husbands and divorce left them without support. The Supreme Court has itself noted the social and financial difficulties women may face after divorce, and has required their financial interests to be secured before dissolution. But the social position has changed, and the facts in this post show the protection working against a wife. The answer is not to remove the protection but to place it where it belongs. The proposed test is neutral as to gender, and it keeps dependence and economic vulnerability as weighty factors at stage two. A spouse who truly needs protection still has it.

One limit should be stated plainly. Only the Supreme Court can apply this test. A spouse in a family court has no access to it without years of further litigation. Here, only the Parliament to act. The Law Commission recommended irretrievable breakdown as a ground for divorce in 1978 and again in 2009, and the Rajya Sabha passed a bill to that effect in 2013, but it was never enacted.

Conclusion

Section 13B was designed to protect spouses from a hasty divorce. In a case like that of W and H, it now protects a spouse who has agreed to divorce three times and has paid nothing. Shilpa Sailesh gave the Supreme Court the power to correct this, and the 2026 decisions show that the Court will use it. What is missing is a stated method.

The three-stage test proposed here asks whether consent was seriously given and the marriage is dead; why consent was withdrawn and who gains from the withdrawal; and how the decree can secure what was promised. It respects the statutory right to change one’s mind, while denying the use of that right to hold another person in a marriage for leverage. Until Parliament acts, the Court should state such a test, so that the outcome in each case turns less on the bench and more on the reasons.

Adv. Anisha Gupta (co-founder, MGLO) is a practising advocate at the Delhi High Court and a certified mediator.

Shreya Banchhor is a third year student at NLSIU.