The Mediation Act 2023 is the statute most often misdescribed in Indian legal writing, and the misdescription is always the same one. It is said, in summaries, examination notes and a good deal of professional commentary, that the Act made mediation compulsory before a civil or commercial suit can be filed. It did not. The Bill that Parliament received proposed exactly that compulsion, the proposal was examined and criticised, and the enacted text withdrew it. What the Act made compulsory is nothing that was not compulsory already: the one mandatory mediation regime in Indian law, the pre-institution mediation required for commercial disputes of specified value under the Commercial Courts Act, sits in a different statute, predates this Act, and is preserved by this Act through a proviso rather than created by it. Everything else in the Act is an offer, not an order. A party may mediate before suing, a court may refer a pending dispute to mediation, and a settlement reached through the process carries the force of a decree. Whether to enter the process at all is, outside the commercial exception, the parties’ own choice.

That single distinction, between a statute that facilitates and a statute that compels, organises everything else in the Act, and it is the distinction this article keeps in view throughout. The Act is India’s first dedicated central statute on mediation. It defines the process, states which disputes it may and may not touch, regulates mediators and the institutions that supply them, prescribes a timetable, protects the confidentiality of what is said, gives the resulting settlement the enforceability of a judgment, creates a regulator in the Mediation Council of India, recognises community mediation and online mediation, and amends eight other statutes so that the mediation references scattered through Indian law point in one direction. It also leaves gaps that an honest account has to record: large parts of it await commencement, its regulator was provided for long before it was established, and the international instrument that would carry an Indian mediated settlement abroad has been signed by India and not ratified. Each of those points is taken in turn below.

The Mediation Act 2023 in outline

The Mediation Act, 2023 is Act 32 of 2023. It received the assent of the President on 14 September 2023 and was published in the Gazette of India on 15 September 2023. Its long title states its purposes in full: to promote and facilitate mediation, especially institutional mediation, for the resolution of disputes, commercial or otherwise; to enforce mediated settlement agreements; to provide for a body for the registration of mediators; to encourage community mediation; and to make online mediation an acceptable and cost effective process. The Act extends to the whole of India and contains sixty five sections arranged in eleven chapters, followed by ten schedules.

The architecture is worth setting out because later argument returns to it repeatedly. Chapter I contains the short title, extent and commencement provision and the definitions. Chapter II states where the Act applies. Chapter III, headed Mediation, contains the mediation agreement in Section 4, pre-litigation mediation in Section 5, the disputes not fit for mediation in Section 6, and the power of a court or tribunal to refer parties to mediation in Section 7. Chapter IV deals with mediators: appointment, the preference of the parties, conflict of interest and disclosure, termination of a mediator’s mandate, and replacement. Chapter V governs the proceedings themselves, from territorial jurisdiction and commencement through the role of the mediator, the time limit, the settlement agreement and its optional registration, confidentiality, termination and costs. Chapter VI, containing Sections 27 to 29, is the enforcement chapter: enforcement of the settlement, the narrow challenge to it, and the exclusion of the mediation period from limitation. Chapter VII provides for online mediation, Chapter VIII establishes the Mediation Council of India, Chapter IX deals with mediation service providers and mediation institutes, Chapter X provides for community mediation, and Chapter XI collects the miscellaneous provisions, including the amendments to other Acts in Sections 58 to 65.

Two features of this outline matter for everything that follows. The first is that the Act is a framework statute rather than a code of procedure. It fixes the points on which uniformity was thought necessary, namely who may mediate, how long the process may run, what a settlement is worth, and who regulates the field, and it leaves the conduct of individual sessions largely to the mediator and the parties. The second is that the Act absorbs conciliation into its own vocabulary. The definition of mediation in Section 3 provides that mediation includes a process referred to by the expressions pre-litigation mediation, online mediation, community mediation, conciliation, or expressions of similar import, in which parties attempt to reach an amicable settlement of their dispute with the assistance of a mediator who has no authority to impose a settlement on them. That definitional move, reinforced by the amendments to the Arbitration and Conciliation Act discussed below, is the Act’s quiet answer to a question that had dogged Indian alternative dispute resolution for decades: whether mediation and conciliation were two processes or one. The statute treats them as one process under one name.

Commencement: an Act in force in part

Section 1(3) follows the familiar pattern of allowing the Central Government to appoint different dates for different provisions, and the commencement that followed the assent was partial in a way that any accurate account of the Act must state plainly. By notification S.O. 4384(E), the Central Government brought a specified set of provisions into force on 9 October 2023: Sections 1 and 3, Section 26, Sections 31 to 38, Sections 45 to 47, Sections 50 to 54, and Sections 56 and 57. The pattern of that selection is instructive. What commenced at once was the definitional apparatus, the saving for Lok Adalat and Permanent Lok Adalat proceedings in Section 26, the whole of the Mediation Council chapter, the Mediation Fund and the rule making and allied machinery, and the transitory provisions. What did not commence in that notification was the operative core of the Act: the application provision in Section 2, the mediation agreement, pre-litigation mediation itself in Section 5, the exclusion provision, the referral power, the chapters on mediators and proceedings, the enforcement chapter, online and community mediation, the service provider provisions, and the amendments to other Acts in Sections 58 to 65, all of which stand enacted and await appointed days under Section 1(3).

This staged commencement is not a defect in the drafting, and it should not be reported as one. A statute that creates a regulator, a registration system for mediators and a recognition system for institutions cannot sensibly switch on its operative provisions before the machinery exists to administer them; commencing the Council chapter and the fund and rule making provisions first is the sequence the design requires. But the consequence for readers and advisers is real, and it conditions several statements made elsewhere in this article. When this article describes what Section 5 provides or what Section 27 enforces, it describes the enacted law of India as Parliament passed it, in a statute whose central chapters await their appointed days. The distinction between an enacted provision and a commenced provision is elementary, and much of the confused commentary on this Act consists of ignoring it in one direction or the other, either by treating the whole Act as operative from October 2023 or by treating the whole Act as somehow not yet law. Neither is correct. The Act is law; part of it is in force; the rest awaits notification.

The commencement question also explains the position of the regulator. Chapter VIII, containing Sections 31 to 38, has been in force from 9 October 2023, and under Section 31(1) the Central Government establishes the Mediation Council of India by notification. The Council was established by notification under that provision, with its head office at Delhi. The establishment of the body and the commencement of the registration regime it will administer are separate steps, and the second depends on the Council’s own regulations and on the commencement of the chapters that give the registration system its subject matter. Section 39, on monitoring and reporting, was not among the provisions commenced in October 2023. An account of the Council that describes a functioning registration system without attending to those steps repeats the commonest error in writing on this Act, and this article does not make it: the Council chapter is in force, the Council is established, and the operative registration framework operates through the machinery the Act prescribes as that machinery is brought into use.

Where the Act applies, and what mediation means under it

Section 2 confines the Act, in the first place, territorially. It applies where mediation is conducted in India, and then only in the cases the section enumerates: where all or both parties habitually reside in or are incorporated in or have their place of business in India; where the mediation agreement provides that the dispute shall be resolved in accordance with this Act; where the mediation is an international mediation, a term the definitions supply for a mediation conducted in India in which at least one party is a foreign national, a body incorporated abroad, or a similar external party; and, in the government case, where one of the parties is the Central Government or a State Government or an agency, public body, corporation or local authority controlled by them and the dispute is a commercial dispute, together with any other dispute or class of dispute that the Central Government or a State Government notifies for the purpose. The territorial anchor is the constant. A mediation conducted outside India is outside the Act, whatever the nationality of the parties, and that single feature drives the international gap examined later in this article.

Section 49 adds a condition to the government case that advisers overlook at their clients’ cost. Where the Central Government, a State Government or an entity controlled by them is a party, a mediated settlement agreement requires the prior written approval of the competent authority empowered for the purpose before it binds the government party. The provision reflects a settled administrative anxiety, that a settlement signed at a table by an officer without authority could commit public funds or public rights, and it means that a mediation involving a government party has two stages rather than one: the negotiation, and the approval without which the negotiation’s product is incomplete. Section 2’s inclusion of government commercial disputes within the Act and Section 49’s approval condition have to be read together, and the second qualifies the first in every case to which it applies.

The definitions in Section 3 do more work than definitions usually do. Mediation is defined by the absence of decisional power: the mediator assists, facilitates and communicates, and does not impose. The definition also settles the relationship with conciliation, as noted above, and it gives statutory meanings to the institutions on which the Act depends, among them the mediation service provider, the mediation institute, the Mediation Council, and the Legal Services Authorities whose court annexed and community roles run through the Act. The definition of a commercial dispute is borrowed from the Commercial Courts Act, 2015, which is the borrowing that makes the proviso to Section 5 work: when the Mediation Act speaks of commercial disputes of specified value, it is speaking the Commercial Courts Act’s language, and the threshold and the mandatory regime that attach to that language come from that Act. Section 55 then gives the Mediation Act overriding effect over mediation and conciliation provisions in other laws, subject to the savings the Act itself contains, and the Second Schedule lists the enactments for that purpose. The design, in short, is of a hub statute: definitions at the centre, eight amended statutes at the rim, and one pre existing mandatory regime expressly preserved rather than replaced.

The provision the statute was watched for: the Bill’s compulsion and its removal

The Mediation Bill, 2021 was introduced in the Rajya Sabha on 20 December 2021 by the Minister of Law and Justice, Kiren Rijiju. Its pre-litigation provision, as introduced, required a party to take steps to settle a dispute by pre-litigation mediation before instituting a suit, whether or not a mediation agreement existed and whether or not the other party consented. Compulsion was not an implication of the Bill; it was its headline. The Bill also proposed a longer timetable than the Act eventually adopted, and the combination of a mandatory first step and a generous period for completing it drew the criticism on which the Bill’s subsequent history turned.

The criticism deserves to be stated at full strength, because the enacted Act is unintelligible without it. The first objection was conceptual. Mediation, on the account its own practitioners give, works because the parties are present by choice and may leave; a settlement produced by a process both sides entered willingly carries a legitimacy and a durability that an imposed outcome does not. Compel attendance, the critics argued, and what is compelled is attendance only: an unwilling party can sit through the sessions, concede nothing, and collect the certificate of failure that unlocks the court. The process then performs the opposite of its function, adding a procedural toll booth on the road to court and taxing precisely the litigants, the small claimant and the individual defendant, least able to pay in time. The second objection was constitutional in flavour. A right of access to the courts is not a courtesy the state extends; conditioning its exercise on the exhaustion of a prior process, the critics said, is defensible only where the process is genuinely capable of disposing of the dispute, and a process defined by the absence of decisional power cannot guarantee that. The third objection was practical and comparative: jurisdictions that had mandated mediation had generally mandated a single introductory session or an information meeting, not a full process, and had done so with exemptions and with infrastructure already in place.

The case for compulsion was also a serious one, and the stance this article takes requires it to be stated seriously too. Its proponents pointed to the behaviour of litigants and their advisers: left to choice, parties do not choose mediation, because proposing it is read as weakness, because advisers are trained and paid for litigation, and because the party who expects to benefit from delay has no reason to volunteer for a process designed to end the dispute. A nudge that is not compulsory, on this account, is not a nudge at all. The cost of the compelled attempt was said to be small and bounded: the timetable was fixed, the period spent in mediation was to be excluded from limitation, and the empirical claim was made that settlement rates in systems with mandatory referral had risen rather than fallen, because parties who would never have chosen the room settled once they were in it. On this view the voluntary character of mediation resides in the settlement, which no one can be forced to sign, and not in the attendance, which the state may require as it requires a dozen other procedural steps before a hearing.

The Bill was referred to the Parliamentary Standing Committee on Personnel, Public Grievances, Law and Justice, and the Committee’s examination became the hinge of the story. The Committee recommended against the mandatory design of pre-litigation mediation, taking the view that making the process compulsory sat uneasily with the voluntary character that gives mediation its value, and it recommended that pre-litigation mediation be made voluntary. When the Bill returned to Parliament in 2023, the government accepted the direction of that recommendation. The Bill was passed by the Rajya Sabha on 1 August 2023 and by the Lok Sabha on 7 August 2023 with the pre-litigation provision recast: compulsion was deleted from the general law of civil and commercial suits, and the one mandatory regime that already existed, for commercial disputes of specified value, was expressly preserved. The assent followed on 14 September 2023.

The verdict on that history is the namable claim of this article, and it is stated here so that the sections that follow can be read as its proof. The Act made mediation voluntary and left compulsion exactly where it found it. The Bill proposed to change the compulsory scope of mediation in India by extending it to every civil and commercial suit; Parliament changed that scope by nothing at all. A reader who understands that single fact understands the Act’s most litigated question, and a reader who misses it will misread Section 5, the proviso, and the enforcement chapter in succession.

Is pre-litigation mediation under the Act a duty or a choice?

It is a choice. Section 5 provides that parties may, voluntarily and with mutual consent, take steps to settle their dispute by pre-litigation mediation before filing a suit, whether or not a mediation agreement exists. The exception lies elsewhere: commercial disputes of specified value remain under the Commercial Courts Act’s mandatory regime, preserved by the proviso.

Mediation before the Act: the scattered law the statute gathered

A dedicated mediation statute arrived in India late, and the lateness is part of its meaning. Consensual settlement of disputes is older than the court system that came to dominate the field: village and trade panchayats composed differences for centuries, and the colonial and post independence state repeatedly reached for settlement mechanisms when its courts fell behind. What the twentieth century produced, however, was not a mediation law but a series of doors into settlement built into other structures, each with its own key. The Legal Services Authorities Act, 1987 created the Lok Adalat, whose award is deemed a decree and against which no appeal lies, and built a national apparatus of legal services authorities whose panels and camps settled disputes in volume. The Code of Civil Procedure acquired, through the amendments that took effect in 2002, Section 89, empowering a court that saw elements of settlement to formulate terms and refer the parties to arbitration, conciliation, judicial settlement including settlement through Lok Adalat, or mediation. The Arbitration and Conciliation Act, 1996 gave conciliation a statutory scheme of its own in Part III and gave the conciliation settlement the status of an arbitral award on agreed terms. Subject statutes added their own provisions: the Companies Act, 2013 empowered the Central Government to maintain a panel of mediators for disputes before the company tribunals; the Micro, Small and Medium Enterprises Development Act, 2006 routed delayed payment disputes through facilitation councils applying the conciliation provisions of the 1996 Act; the Consumer Protection Act, 2019 built a mediation chapter with cells attached to the consumer commissions; and the Commercial Courts Act, as amended in 2018, inserted the Section 12A regime that made pre-institution mediation compulsory for commercial disputes of specified value.

The Supreme Court spent the same decades explaining how these doors related to one another. In Salem Advocate Bar Association v Union of India, the Court upheld the Section 89 scheme and set in train the framing of mediation rules by the High Courts. In Afcons Infrastructure Limited v Cherian Varkey Construction Company, decided in 2010, the Court sorted the Section 89 categories, explained which classes of dispute were ordinarily unsuitable for reference, among them representative suits, election disputes, cases involving prosecution for criminal offences and claims against minors, and directed courts to consider settlement at the first hearing as a matter of course rather than of exception. The kinship between that judicial list of unsuitable subjects and the First Schedule of the present Act is not accidental: the Schedule is, in substantial part, Afcons written into a statute, enlarged by the regulatory and fiscal exclusions a legislature adds and a judgment does not. Court annexed mediation centres grew up alongside the case law, attached to the Supreme Court, the High Courts and the district courts, and a professional community of trained mediators formed within them years before Parliament legislated for the profession.

This history explains both the Act’s design and its necessity. By 2021 mediation in India was simultaneously everywhere and nowhere in the law: everywhere as a practice and a judicial policy, nowhere as a single framework that answered the elementary questions a stranger asks. What status did a settlement have if the process that produced it was not a Lok Adalat, a Section 12A mediation or a Part III conciliation? Who was a mediator, and who answered for the quality of the name on a panel? How long could the process run, and what happened to limitation while it did? Each door answered those questions differently, and a settlement’s worth depended on which door the parties had entered by. The Act’s answer is the hub and rim design described above: one definition, one timetable, one settlement status, one regulator, and amendments aligning the older doors with the new hall. Whether the hub displaces the doors or merely organises them is a question the older institutions’ survival answers in part: Section 26 expressly preserves Lok Adalat and Permanent Lok Adalat proceedings, the consumer chapter continues in its own statute as amended, and Section 12A continues in the Commercial Courts Act. The Act gathers; it does not abolish.

The mediation agreement: Section 4 and the consent the Act is built on

Section 4 is the Act’s foundation stone, and it is deliberately unglamorous. A mediation agreement must be in writing. It may be a clause in a larger contract, by which the parties agree in advance that disputes arising from the contract will be submitted to mediation, or a separate agreement concluded after a dispute has arisen, and it may cover all disputes between the parties or only some of them. Writing is not a formality in this setting: it is the evidence that consent, the substance on which the whole Act rests, was given, and it is the document from which the mediator’s authority, the service provider’s engagement and the settlement’s statutory pedigree all descend. An agreement recorded in electronic form stands on the same footing, consistently with the Act’s recognition of online process.

Three consequences of the writing requirement deserve emphasis. The first is that a mediation clause, unlike an arbitration clause, does not oust anything. A party bound by an arbitration agreement can be stayed out of court at the instance of the other; a party who promised to mediate and then sues has broken a promise, but the promise was to attempt a process, and the Act does not make the attempt a condition of the suit outside the Section 12A field. The second is that the clause and Section 5 operate independently: Section 5 empowers parties with no agreement at all to take the pre-litigation step by consent at the moment of dispute, so the absence of a clause in an old contract is no bar to mediating under the Act, and the presence of a clause in a new one adds a contractual obligation to a statutory option. The third is that consent given in advance is consent to a process and not to an outcome. Nothing in a mediation agreement, however drafted, can oblige a party to settle, and a clause that purported to bind a party to accept a mediator’s proposal would contradict the definition in Section 3, under which the mediator has no authority to impose a settlement at all.

The agreement also fixes the boundaries of the process in ways that return at the enforcement stage. It identifies the disputes submitted, and a settlement that ranges beyond them is safer the more clearly the agreement, or a supplementary consent recorded in the process, supports the wider composition Section 19 permits. It identifies the parties, which matters when Section 27 binds persons claiming under them. And it is the natural place for the parties’ choices about the mediator, the institution, the place and the language of the process, choices Section 8 then honours by making the parties’ agreed appointment procedure the first rule of appointment. Advisers drafting dispute resolution clauses after this Act will write them in tiers, negotiation, then mediation under the Act, then arbitration or litigation, and the tiers will work only if each is drafted as what it is: the mediation tier a genuine consent to a bounded process, not a toll booth dressed as one. The courts’ treatment of a colourable urgency claim under Section 12A, insisting that a formula inserted to evade a statutory step will be looked through, is the standing warning against drafting the tier as a formality.

Section 5: pre-litigation mediation, voluntary and with mutual consent

Section 5 is short, and every word of its central choice matters. It provides, in substance, that whether any mediation agreement exists or not, the parties, before filing any suit or proceeding of a civil or commercial nature in any court, may voluntarily and with mutual consent take steps to settle their disputes by pre-litigation mediation conducted in accordance with the Act. Three features of that formulation repay attention. First, the absence of a mediation agreement is no obstacle: the section supplies a statutory route into mediation for parties whose contract contains no mediation clause and who made no arrangement before the dispute arose, which is the position of most disputants in India. Second, the operative words are permissive and bilateral: the parties may take the step, and they take it voluntarily and with mutual consent. One party cannot drag the other into a pre-litigation mediation under this section, and a court cannot refuse to receive a suit on the ground that the plaintiff did not attempt it. Third, the section addresses the moment before filing. It is not a provision about pending suits, which are the province of Section 7, and it is not a provision about agreements to mediate, which are the province of Section 4. It occupies the interval in which a dispute exists and no proceeding does, and it fills that interval with an option.

The proviso then performs the Act’s most consequential act of preservation. It provides that pre-litigation mediation in matters of commercial disputes of specified value shall be undertaken in accordance with the provisions of the Commercial Courts Act, 2015. The reference is to Section 12A of that Act, inserted in 2018, which provides that a suit in a commercial dispute shall not be instituted unless the plaintiff exhausts the remedy of pre-institution mediation, except where the suit contemplates urgent interim relief, and which routes that mediation through the authorities constituted under the Legal Services Authorities Act, 1987 and the rules made for the purpose. The specified value for a commercial dispute stands at three lakh rupees. The two regimes are therefore distinct in source, in trigger and in consequence, and keeping them distinct is the whole of the law on this point: the Commercial Courts Act commentary in this series examines Section 12A in its own setting, and the present article takes from it only what the proviso imports. Under the Mediation Act, mediation before suit is voluntary. Under the Commercial Courts Act, for a commercial dispute of specified value in which no urgent interim relief is contemplated, mediation before suit is a condition of instituting the suit at all.

That the Commercial Courts regime is mandatory, and mandatory with teeth, is not a matter of inference. In Patil Automation Private Limited v Rakheja Engineers Private Limited, decided on 17 August 2022, the Supreme Court held that Section 12A is mandatory in character, that exhausting pre-institution mediation is not a mere procedural formality, and that a plaint instituted in breach of it is liable to be rejected under Order VII Rule 11 of the Code of Civil Procedure, 1908. The Court softened the landing by prospective application: the rejection consequence was made to operate for suits instituted from 20 August 2022, so that plaintiffs who had filed during the period when the mediation infrastructure was incomplete were not destroyed by a rule whose machinery had not existed. Later decisions have policed the boundary of the exception rather than the existence of the rule. In Yamini Manohar v T.K.D. Keerthi the Court held that whether a suit contemplates urgent interim relief is to be judged from the plaint and the plaintiff’s standpoint, and that a court may look behind a colourable claim of urgency; in Dhanbad Fuels Private Limited v Union of India the mandatory reading was reaffirmed and settlements under Section 12A were connected to the enforcement provisions of the Mediation Act. The line of authority matters to this Act because the proviso makes it part of this Act’s working content: the only mediation an Indian litigant can be compelled to attempt before suing is the mediation that body of case law construes.

Two practical consequences follow, and advisers who collapse them cause the errors the proviso was drafted to prevent. A plaintiff in an ordinary civil dispute, a partition claim, a contract claim below the commercial threshold, a dispute with a neighbour, who files without mediating has filed perfectly regularly, and a defendant who objects that Section 5 was not complied with objects to a duty that does not exist. A plaintiff in a commercial dispute of specified value who files without exhausting Section 12A mediation, and whose plaint contemplates no urgent interim relief, has filed a plaint liable to rejection, and no amount of willingness to mediate after the event cures the defect in the manner the statute requires. The Mediation Act is responsible for neither outcome. It created the option in the first case and preserved the condition in the second, and an account that attributes either the option or the condition to the wrong statute misleads the reader on the only question most readers bring to it.

The decision table: voluntary, mandatory, excluded

The brief for this article named its findable artifact, and it is the table that follows: a routing of the disputes a reader is likely to hold through the questions the Act asks about them, ending in the only three answers the Act gives, that mediation must be attempted, may be attempted, or cannot be undertaken under the Act at all.

Starting point	The routing question	Position under the Act	Provision
A civil or commercial dispute, no suit filed, both parties willing	Is the dispute within the First Schedule?	Mediation may be attempted before suit, by consent	Section 5
A commercial dispute of specified value, no urgent interim relief contemplated	Has pre-institution mediation been exhausted?	Mediation must be attempted before the suit can be instituted	Section 12A, Commercial Courts Act, preserved by the proviso to Section 5
A commercial dispute of specified value, urgent interim relief contemplated	Does the plaint genuinely contemplate urgent relief?	The suit may be instituted without exhausting pre-institution mediation	Section 12A exception, as read in Patil Automation
A dispute listed in the First Schedule	Is it within any notified exception or a referable aspect?	Mediation cannot be undertaken under the Act for that dispute	Section 6 and the First Schedule
A dispute already pending before a court or tribunal	Does the court see a possibility of settlement?	The court may refer the parties to mediation at any stage	Section 7
A consumer dispute	Which statute supplies the mediation route?	Mediation may be undertaken under the consumer statute’s own chapter, alongside this Act’s framework	Consumer Protection Act, 2019
A dispute where one party refuses all participation	Is there a mediation agreement or a mandatory regime?	No mediation under the Act; the suit proceeds in the ordinary way	Section 5, consent requirement

The table also discloses the Act’s shape at a glance. Compulsion appears in exactly one row, and it is the row whose provision column names another statute. Exclusion appears in one row, drawn from a schedule. Every other row is permissive. That is the statute Parliament enacted, and it is a different statute from the one the Bill proposed, in which the second row’s logic would have occupied the whole table.

Referral by a court, and interim relief

Section 7 addresses the pending dispute. A court or tribunal before which a proceeding is pending may, at any stage, refer the parties to mediation where it appears that elements of a settlement exist which may be acceptable to them, and the parties on a joint application may equally invite such a reference. The referral power is the Act’s bridge to the older practice under Section 89 of the Code of Civil Procedure, under which courts had long directed parties toward mediation, conciliation, arbitration and Lok Adalat, and it places that practice on the footing of the new statute’s machinery: the mediator appointed on a referral is appointed under Chapter IV, the proceedings run under Chapter V, and a settlement that emerges carries the enforcement of Chapter VI. The referral is of the dispute, not of the parties’ rights in it; a party referred to mediation who settles nothing returns to the proceeding with the claim intact, and the Act’s confidentiality provisions, examined below, are designed to ensure that the return is not prejudiced by anything said in the room.

Interim relief is the necessary qualification on everything said so far about sequence, and the Act preserves it expressly. A party is not required to watch a wrong continue, or assets disappear, while a mediation is arranged, begun or completed. The court seised of a dispute retains the power, in exceptional circumstances, to grant interim relief before the commencement of a mediation, during its currency, or after its termination, and the existence of that power is what makes the mandatory commercial regime workable as well as the voluntary one tolerable. Its interaction with Section 12A is the point on which the case law has concentrated: urgency is the door out of compulsory pre-institution mediation, and the courts, from Patil Automation through Yamini Manohar, have insisted that the door is real but narrow, opened by a genuine contemplation of urgent relief disclosed by the plaint and not by a formulaic prayer inserted to evade the mediation the statute requires. For the voluntary regime of Section 5 the point is simpler and is often missed in practice: choosing to mediate under this Act surrenders no right to ask a court for urgent protection, and a party who needs an injunction this week need not choose between the injunction and the mediation.

How does a mediation begin, and how does it end?

A mediation begins by agreement, by a Section 5 step before suit, or by a court or tribunal’s referral under Section 7, with proceedings commencing under Section 14. It ends under Section 24: on a signed and authenticated settlement, on the mediator’s declaration that further effort is unjustified, on a party’s withdrawal, or on expiry of the Section 18 period.

The disputes the Act will not touch: Section 6 and the First Schedule

Section 6 provides that mediation shall not be conducted for the disputes or matters specified in the First Schedule, and the Schedule is the Act’s considered list of subjects on which a private, consensual, confidential process is the wrong instrument. Its entries fall into intelligible groups. The first group protects persons who cannot protect themselves in a bargain: disputes relating to claims against minors, deities, and persons with intellectual disabilities, and claims against persons of unsound mind, where the law has traditionally insisted that any compromise be supervised by a court rather than authenticated by a mediator. The second group reserves public wrongs to public process: matters relating to the prosecution of criminal offences, and disputes involving allegations of serious and specific fraud, fabrication of documents, forgery, impersonation or coercion, where the allegation itself demands findings a mediator has no power to make. The third group protects the revenue and the regulatory state: disputes relating to the levy, collection and penalties in relation to direct and indirect taxes and to refunds, and proceedings before the specialist regulators and tribunals the Schedule names, including matters under the Competition Act, 2002, proceedings connected with the Telecom Regulatory Authority of India and the Telecom Disputes Settlement and Appellate Tribunal, proceedings under the Electricity Act, 2003, the Petroleum and Natural Gas Regulatory Board Act, 2006 and the Securities and Exchange Board of India Act, 1992, and proceedings over which the National Green Tribunal has jurisdiction. The fourth group removes subjects whose outcomes bind strangers: disputes that affect the rights of third parties who are not participating in the mediation, with a careful exception for matrimonial disputes in which a child’s interest is involved, and disputes relating to the determination of compensation for land acquired under the land acquisition laws, where the award speaks to a whole class of claimants. The Schedule closes with complaints and proceedings before statutory authorities concerning the registration, discipline or misconduct of professionals, matters relating to the probate of a will, and a residuary power in the Central Government to notify further entries.

Two observations complete the account of the Schedule. Its function is not merely negative. By listing what mediation cannot do, it defines by contrast the field in which a mediated settlement will be safe from the fourth ground of challenge under Section 28, which strikes at a settlement reached in a dispute that was not fit for mediation; a mediator who checks the Schedule at intake is protecting the enforceability of everything the process later produces. And the Schedule is not the whole of Section 6: the section and the court’s referral power together preserve a judicial discretion to refer a matter that appears suitable for mediation, so that the boundary the Schedule draws is administered by courts in particular cases rather than applied as a mechanical bar in every matter that touches a listed subject. The reader who wants the contrast with a neighbouring institution’s exclusions will find the comparison drawn in this series’ account of the Legal Services Authorities framework and the Lok Adalat, whose proceedings Section 26 of this Act expressly leaves unaffected.

Mediators: appointment, qualification and the discipline of disclosure

Chapter IV builds the human machinery. Section 8 leaves the appointment of the mediator where the philosophy of the Act says it belongs, with the parties, who may agree on a mediator and on the procedure for appointment; failing agreement, a party applies to a mediation service provider, which appoints from its panel, and Section 9 requires the provider, in making the appointment, to consider the suitability of the person and the preference of the parties. The nationality of the mediator is the parties’ affair unless the agreement provides otherwise. For pre-litigation mediation, the Act ties the appointment to the regulated end of the profession: the mediator is to be a person registered with the Mediation Council, or empanelled with a court annexed mediation centre, or empanelled with an authority constituted under the Legal Services Authorities Act, 1987, or with a mediation service provider. The design is a ladder of trust for the case in which the parties arrive without a chosen mediator and without a relationship of confidence with any institution: the statute points them to panels maintained by bodies answerable for the quality of the names on them.

Sections 10 to 12 supply the discipline. A person approached to act as mediator must disclose in writing any circumstance, personal, professional, financial or otherwise, that may raise a justifiable doubt about independence or impartiality, and the duty continues through the proceedings. The mandate of a mediator may be terminated by the service provider on the application of a party, on information disclosing a conflict, or on the mediator’s own withdrawal, and Section 12 provides for replacement so that the process survives the person. Termination for undisclosed conflict is the provision with practical bite: it converts the disclosure duty from an ethical exhortation into a condition of remaining in the appointment, and it gives a party who discovers a connection late a remedy that does not require unmaking anything the process has produced, since a process that has produced no settlement has produced nothing that needs unmaking, and a settlement produced by a mediator labouring under an undisclosed interest is exposed to the challenge chapter on precisely the grounds that chapter enumerates.

What may a mediator do, and what is a mediator forbidden to do?

A mediator may assist the parties independently, neutrally and impartially, facilitate communication, identify issues and explore compromises, as Section 16 provides. A mediator may not impose a settlement, act as arbitrator, counsel or representative in later proceedings on the same dispute, as Section 17 provides, or breach the confidentiality of Sections 22 and 23.

The proceedings: place, timetable, confidentiality and cost

Chapter V is the Act’s procedural core, and its provisions are best understood as answers to the objections that had been made to mediation as an unregulated practice. Section 13 fixes the territorial setting: mediation is to be undertaken within the jurisdiction of the court or tribunal competent to decide the dispute, unless the parties agree on another place or on conducting it online, a flexibility that Section 30 then takes up directly. Section 14 fixes commencement, which matters because commencement starts both the timetable and the limitation exclusion. Section 15 frees the conduct of the process from the codes: the mediator is not bound by the Code of Civil Procedure or the evidence law, and may run joint sessions and separate meetings with each party as the dispute requires. Section 16 has been noticed above. Section 18 then imposes the discipline that distinguishes a statutory process from a conversation: mediation proceedings are to be completed within one hundred and twenty days from the date fixed for the first appearance before the mediator, and the period may be extended by the consent of the parties by a further period not exceeding sixty days. One hundred and eighty days is therefore the outer limit of a mediation under the Act, and the figure is the enacted answer to the Bill’s longer timetable and to the criticism that a mandatory process of indefinite length would tax access to court. In a voluntary process the same limit does different work: it tells a party asked to consent that the exposure is bounded, and it tells a mediator that drift is a breach of the statute and not a style of practice.

Confidentiality is the load bearing wall of the structure, and Sections 22 and 23 build it in two courses. Section 22 obliges the mediator, the service provider, the parties and every participant to keep confidential all matters relating to the proceedings, extends the obligation to the settlement agreement itself except where its disclosure is necessary for implementation or enforcement, and prohibits audio and video recording of the sessions. Section 23 completes the protection from the other side: mediation communications are privileged against disclosure, and a court or tribunal shall not compel the mediator or the participants to disclose what was said, subject to exceptions the section recognises for threats of criminal conduct, domestic violence, child abuse and imminent threats to public safety. Without those two sections, nothing else in the Act would function, because no party concedes anything in a room whose contents can be read back in court; with them, the non-settlement report that Section 21 requires, a bare statement that the mediation did not produce a settlement, without reasons and without attribution of blame, becomes possible, and the returning litigant’s position is preserved exactly as the referral promise requires.

The remaining provisions of the chapter are short and practical. Section 24 states the events on which mediation terminates: the signing and authentication of the settlement, the mediator’s written declaration that further mediation is not justified, a party’s written decision to withdraw, and the expiry of the Section 18 period. Section 25 divides the cost: unless the parties agree otherwise, the fees and costs of the mediation are shared equally, community mediation apart, a default that expresses the Act’s premise that the process belongs to both parties equally and that neither is the applicant for a favour. Parties who mediate under the Act will keep their papers in ordinary office formats, from the mediation agreement to the authenticated settlement and the registration record, and a browser based office file viewer is the kind of tool by which such documents are opened and checked when the file changes hands between a party, an adviser and an institution. Section 29, though it sits in the enforcement chapter, belongs in substance to the proceedings: in computing limitation for any proceeding on a dispute that has been mediated under the Act, the period from the commencement of the mediation to the submission of the non-settlement report or the termination of the mediation is excluded. That exclusion is the statute’s answer to the limitation objection to any pre-suit process, voluntary or mandatory: time spent at the table is not time lost from the limitation clock, and a party who mediates in good faith and fails is restored, in limitation terms, to the position occupied when the process began.

Why does registration not decide whether a settlement can be enforced?

Because the Act separates the record from the right. Section 20 lets the parties register the settlement with the specified authority within one hundred and eighty days of receiving the authenticated copy, late registration being possible for a fee. Section 27 makes enforcement turn on writing, signature and authentication, not registration, which is an evidentiary facility only.

What happens to the limitation clock while a mediation runs?

It stops, in the litigant’s favour. Section 29 excludes, from the computation of limitation for any proceeding on a mediated dispute, the period from the commencement of the mediation under Section 14 until the report under Section 21 or the termination under Section 24, notwithstanding anything in the Limitation Act, 1963.

The settlement: writing, authentication, and the force of a decree

Section 19 defines the product. A mediated settlement agreement is an agreement in writing between some or all of the parties resulting from the mediation, settling some or all of their disputes, and authenticated by the mediator. The informality of the process ends at this document: it must be reduced to writing, it must be signed by the parties, and the mediator’s authentication is what connects it to the statutory process that produced it and distinguishes it from a private compromise note. The parties receive copies of the authenticated agreement, and in institutional mediation the agreement travels to the service provider with the mediator’s covering record. The settlement may settle part of the dispute, leaving the remainder to the proceeding from which the dispute came, and it may reach beyond the referred dispute to compose the parties’ wider differences, a breadth the definition permits and practitioners value, since commercial and family disputes alike rarely confine themselves to the four corners of the original claim.

Section 27 then gives the document its worth, and it is the provision on which the Act will stand or fall. A mediated settlement agreement signed by the parties and authenticated by the mediator is final and binding on the parties and on persons claiming under them, and, subject to the challenge provision, it is enforceable in accordance with the Code of Civil Procedure in the same manner as if it were a judgment or decree passed by a court. It may accordingly be relied on by any party, or by persons claiming through them, by way of defence, set off or otherwise in any legal proceeding. Every element of that formulation is deliberate. Enforcement as a decree means execution, not a fresh suit: the party holding the settlement does not have to sue on it, prove it as a contract and await a judgment that converts it into something executable, which was the position of a mediated compromise outside the Section 12A regime before this Act and which remains the position of an informal settlement that never enters the Act’s process. Reliance by way of defence and set off means the settlement works defensively as well: a party sued in defiance of a settlement pleads the settlement, and the plea is a plea of a decree and not merely of a bargain. Finality means the merits are closed between the parties, as the merits are closed by any decree, subject only to the challenge the Act itself allows.

Section 28 states that challenge, and its narrowness is the counterpart of Section 27’s breadth. A mediated settlement agreement may be challenged only on one or more of four grounds: fraud; corruption; impersonation; and that the mediation was conducted in a dispute or matter not fit for mediation under Section 6. The application lies to the court or tribunal of competent jurisdiction, and it must be made within ninety days from the date on which the applicant received the copy of the settlement agreement under Section 19, with a single condonation: if the court or tribunal is satisfied that the applicant was prevented by sufficient cause from applying within those ninety days, it may entertain the application within a further period of ninety days. There is no ground of error, no ground of unfairness, no ground that the bargain was improvident or the mediator unskilled, and no appeal on the merits dressed as a challenge. The design consciously mirrors the settlement end of arbitration law, where an award on agreed terms is disturbed only for vitiating factors, and it reflects a judgment about what finality requires: a settlement that could be reopened for any reason a decree could be appealed would be worth little more than the paper of the agreement the parties could have signed without the Act, and the decree status of Section 27 would be a courtesy title. The price of that finality is borne by the party who signed a bad bargain freely, and the Act accepts the price, reserving relief for the cases, fraud, corruption, impersonation and a forbidden subject, in which consent itself was not what it appeared to be.

The position on challenge is, as the brief for this article directs, to be described as developing. The enforcement chapter awaits its appointed day, and the body of decisions construing Sections 27 and 28 on facts, the meaning of fraud in a process without pleadings, the treatment of a settlement said to have been procured by a mediator’s undisclosed interest, the interaction with the challenge regime that Section 12A settlements carried under the Legal Services Authorities framework, will accumulate as cases arise. What can be said on the enacted text is structural: the grounds are exhaustive in form, the period is short and extendable once, and the forum is the court or tribunal of competent jurisdiction. Advisers should also note the asymmetry the two sections create at the drafting stage of a settlement. Because challenge is confined to vitiating factors, the protection of a party lies in the process that precedes signature: independent advice before signing, a mediator whose disclosures under Section 10 are on record, and a subject matter checked against the First Schedule. The Act front loads its safeguards into the process and refuses to supply them afterwards, which is a coherent choice, and a demanding one.

Online mediation and community mediation

Section 30 recognises online mediation in terms: mediation, including pre-litigation mediation, may be conducted at any stage by the use of electronic forms or computer networks, electronic mail services, video conferencing or audio conferencing, with the written consent of the parties, and subject to the requirement that the integrity and confidentiality of the proceedings be maintained. The provision’s importance is less technological than jurisdictional and economic. Section 13’s default territorial rule ties a mediation to the place of the competent court; the online provision loosens that tie by consent, so that parties in different cities, and the small claimant for whom travel is the largest cost of any process, can mediate without relocating the dispute. The consent requirement in writing, and the express subjection of the online process to the same confidentiality discipline as a physical one, answer the obvious objection that a screen mediated process leaks: the recording prohibition of Section 22 and the privilege of Section 23 apply to the online room as they apply to the physical one.

Community mediation, in Chapter X, is the Act’s most distinctive social provision. Section 43 provides that a dispute likely to affect the peace, harmony and tranquillity among the residents or families of an area or locality may be settled through community mediation with the prior mutual consent of the parties, and Section 44 supplies the procedure: the Authority under the Legal Services Authorities Act, or the District Magistrate or Sub Divisional Magistrate for the area, maintains and notifies a permanent panel of community mediators, and a panel of three community mediators conducts the process. The subjects are the disputes of proximity: boundaries, noise, water, animals, access, the small abrasions of shared space that litigation magnifies and that a decree, however correct, leaves festering between neighbours who must continue to live adjacently. The community settlement is recorded in writing and signed, and its character differs from the Chapter VI settlement in a way the Act intends: the community process rests on the standing of the panel and the consent that brought the parties to it, and its product binds by the adherence of the neighbourhood rather than by execution as a decree. To present community mediation as a lesser mediation is to misread it. It is a different instrument for a different class of dispute, one in which the enforcing authority is social and the statute’s contribution is to give the panel, the consent requirement and the written record a legal frame.

The profession the Act creates: training, grading and the register

Every framework statute creates a profession in its image, and the Mediation Act is candid about the one it intends. Before the Act, entry into mediation practice in India ran through three informal gates. A lawyer, a retired judge or a professional of standing could be trained by a court annexed centre or a private institution, empanelled by a court or an authority, and thereafter practise on reputation. The training itself was real but unstandardised: courses of forty hours became a conventional benchmark in court annexed practice, yet no public authority prescribed a curriculum, graded the trained, disciplined the errant or answered for the difference between a certificate of attendance and a qualification. The Act replaces that informality with the apparatus of a regulated occupation, distributed among the Council, the service providers and the institutes. The Council registers mediators and may renew, suspend or cancel registration; it recognises mediation institutes, which conduct training and examinations and grade mediators; it recognises service providers, which maintain the panels from which appointments are made; and it lays down the standards, curricula and codes of conduct to which all three tiers are answerable. Section 38’s functions, read with Sections 40 to 42, are therefore the constitution of a profession, and the questions they leave to regulation, the qualifications for registration, the place of experience against examination, the treatment of the large existing body of court trained mediators, are the questions on which the profession’s character will be settled.

Two features of the design are worth a reader’s notice. The first is that registration is not made the universal condition of acting. The parties’ freedom under Section 8 to agree on a mediator of their choice is preserved, and the regulated panels are the default for the cases, pre-litigation mediation and applications to a provider among them, in which the parties arrive without a chosen name. The statute thus runs a dual system: a free market of agreed appointments beside a regulated market of default appointments, with the register disciplining the second and informing the first. The second feature is the depository: the Council’s duty to maintain an electronic record of mediated settlement agreements converts a private process into a measurable system. Mediation’s advocates had argued for decades from anecdote, because a confidential process generates no public data; the depository, fed by the registrations Section 20 facilitates, is the mechanism by which settlement rates, subject matter patterns and time to resolution can enter public argument on evidence. A reader assessing the Act’s success in a decade should ask for the depository’s numbers before reaching for any other measure.

The discipline side of professionalisation should not be understated. Registration that can be suspended or cancelled, a code of conduct laid down by the regulator, the Section 10 disclosure duty enforced through the termination provisions of Chapter IV, and the exposure of a settlement to challenge where the process miscarried, together give a disappointed party somewhere to go that is not a courtroom: a complaint to the registering and recognising authority about the mediator, distinct from the challenge to the settlement itself. That separation, between the fate of the document and the fate of the practitioner, is a mark of a maturing framework, and it mirrors the structure the series has traced in the regulated legal profession itself, where discipline of the advocate and validity of the proceedings run on separate tracks.

Access and cost: who pays for the room, and who is let in free

A voluntary process that costs money will be used by those who have it, and the access question is therefore not decoration on this Act but a test of its premise. The Act’s own provision is the default rule of Section 25, equal sharing of fees and costs unless otherwise agreed, supplemented by the fee structures of the institutions through which mediation is administered. Court annexed mediation centres and the legal services authorities have historically supplied mediation at nominal or no charge in referred matters, funded as part of the legal services mission, and Section 26’s preservation of the Legal Services Authorities framework keeps that channel intact. The Section 12A regime likewise runs through the authorities under that Act, so that the one compulsory mediation in the system is also one in which the state’s apparatus, rather than the parties’ pockets, carries the institution. For the voluntary field, the economics that matter are comparative and the Act’s framers knew it: a mediation’s fee, shared, is set against the court fee, the cost of years of interlocutory practice and the opportunity cost of a dispute unresolved, and on that comparison the process is cheap for commercial parties and dear only for the smallest claims, where the legal services channel is the designed answer.

Access has a second dimension that fee talk misses, which is knowledge. A process parties do not know exists is not available to them in any sense that matters, and the pre-Act decades showed a consistent pattern: mediation was used heavily where courts referred parties to it and lightly where parties had to find it themselves. The Act addresses the information failure obliquely, through the referral power of Section 7, through the Council’s promotional function in Section 38, and through the sheer publicity of a dedicated statute, and the honest assessment is that these are beginnings rather than solutions. The community mediation chapter is the access provision with the deepest reach, because it takes the process to the disputes of people who do not instruct lawyers at all, under panels notified by the district administration and the legal services authority of the area. If the Act changes the distribution of dispute resolution in India, rather than merely its organisation for those already served, the change will be visible there first.

There is also a cost the Act deliberately refuses to impose, and it completes the access account. A party who declines mediation, or who enters it and leaves without a settlement, faces no penalty in the Act itself: no adverse order is prescribed for the refusal, no costs consequence is written into Section 25 for the party who would not settle, and the non-settlement report’s silence prevents the attempt from being characterised against either side. Costs discipline for unreasonable refusal remains where English practice also located it for most of its history, in the discretion of the court seised of the eventual suit, exercised on the whole conduct of the litigation and not by statutory command. The choice is consistent with the voluntariness principle in its strictest form: the Act will not punish a citizen for insisting on the adjudication the Constitution promises, and it relies for its uptake on the attractiveness of the alternative rather than on the unattractiveness of refusing it. Whether that purity survives contact with docket pressure, in the form of judicial costs orders that harden into a practice, is one of the developments the referral case law of the coming years will disclose.

The settlement among its neighbours: decree, award and compromise

Section 27’s choice of the decree as the settlement’s legal character is best understood against the alternatives Indian law already possessed, because each alternative had a constituency and the Act’s choice among them was a judgment about what a mediated settlement is. The oldest neighbour is the compromise decree under the Code of Civil Procedure: where a pending suit is adjusted by an agreement, the court records the agreement and passes a decree in its terms, and the product is executable and largely unappealable. Its limit is structural: it presupposes a suit. Parties who settle before suing, which is the very case Section 5 addresses, had no access to it, and their settlement was a contract, enforceable by suing on it. The second neighbour is the Lok Adalat award, deemed a decree of a civil court under the Legal Services Authorities Act, final and binding with no appeal. Its strength is the same decree status the Act adopts; its setting is the legal services apparatus, episodic sittings and the particular classes of matter that apparatus handles. The third neighbour is the settlement under Part III of the Arbitration and Conciliation Act, 1996, which Section 74 of that Act clothes with the status and effect of an arbitral award on agreed terms under Section 30. That route gave a conciliated settlement real force, and the force was borrowed: the settlement was executable because the statute pretended it was an award.

Against that background, Section 27 does something cleaner than any of the three. It gives the mediated settlement decree status directly, in its own statute, without a pending suit, without the Lok Adalat setting and without the fiction of an award. The borrowing that remains is of machinery rather than of character: enforcement follows the Code of Civil Procedure as a decree is enforced, and challenge follows the vitiating factors model that arbitration law developed for awards on agreed terms. The harmonising amendments then generalise the choice, connecting the settlements produced under the older routes, including the Section 12A route, to Sections 27 and 28, so that the legal character of a settlement ceases to depend on the door by which the parties arrived at it. That uniformity is the enforcement chapter’s real contribution, and it is the answer to the practitioner who asks what the Act adds to a law that already had Lok Adalats and conciliation: it adds a settlement whose status does not have to be explained by reference to something else.

How a mediation is conducted, and the state as a disputant

The Act regulates the frame of a mediation and leaves the room itself to practice, and a short account of that practice belongs in a statute pillar, because the frame only makes sense against it. A mediation ordinarily opens with the mediator’s statement of the process, its confidentiality and the mediator’s lack of decisional power, followed by each party’s account of the dispute. Thereafter the mediator moves between joint session and caucus, the private meeting with one party in which positions are tested, interests beneath positions are explored and offers are carried, improved or abandoned without the loss of face that a rejected public offer inflicts. The mediator’s instruments are questions, summaries, the identification of the costs of continued conflict, and the disciplined communication of each side’s reality to the other. Section 15’s release of the mediator from the codes of procedure and evidence, Section 16’s description of the role as facilitative and impartial, and the confidentiality of Sections 22 and 23 are the statutory conditions of that practice: caucus works only because what is said in it cannot be carried into court, and the mediator can test a party’s case frankly only because the testing leaves no record that can be used. A party may be represented and assisted, and the Act contemplates parties appearing with their advisers, whose role changes character in the room: the adviser who in court persuades a judge must in mediation persuade an opponent, and the skills are related rather than identical.

The state as a disputant raises the frame’s hardest practical problem, and the Act meets it with the two provisions already noticed, Section 2’s inclusion of government commercial disputes and Section 49’s approval condition. Government litigation is the largest single stream in the Indian courts, and its settlement behaviour is governed by an anxiety no private party shares: the officer who settles must answer for the settlement to audit, to vigilance and to successors, while the officer who litigates and loses answers to no one. The predictable result, documented by committees over decades, is that government parties appeal by default and settle by exception. Section 49 does not remove the anxiety; it organises it, by requiring that the settlement of a dispute to which the government or a controlled entity is a party receive the prior written approval of the competent authority, so that the officer at the table negotiates within a mandate and the approval, rather than the officer’s personal exposure, is the settlement’s foundation. Whether that machinery produces settlements in volume depends on the authorities’ willingness to issue mandates before the process rather than verdicts after it, a matter of administration on which the statute can only insist on the form. The Act’s inclusion of government commercial disputes, and its separate treatment of non commercial disputes by or against the government, which the First Schedule and the application provision keep outside the ordinary run of the Act, mark the boundary of the experiment: the state is brought into mediation where its disputes resemble those of other commercial parties, and the approval condition guards the point at which public rights and public money are compromised.

The Mediation Council of India, service providers and institutes

Chapter VIII creates the regulator. Section 31 establishes the Mediation Council of India as a body corporate with perpetual succession and a common seal, with its head office at Delhi and power to establish offices elsewhere in consultation with the Central Government. Section 32 composes it of a Chairperson, full time and part time members drawn from persons of ability, integrity and standing with experience in mediation, alternative dispute resolution, law and allied fields, together with the ex officio presence of the government’s law and expenditure administration and a chief executive for its secretariat under Section 37. Sections 33 to 36 deal with vacancies, resignation, removal, and the appointment of experts and committees. Section 38 then enumerates the functions on which the profession’s future organisation depends: promoting mediation, including institutional mediation, in India; laying down guidelines and codes of conduct for mediators; registering mediators and renewing, suspending or cancelling registration; recognising mediation service providers and mediation institutes, and renewing, suspending or cancelling that recognition; laying down standards, curricula and training for mediators; and maintaining an electronic depository of mediated settlement agreements, the last a function whose value to researchers and to the credibility of the system will grow with the register it creates. Section 39 adds monitoring and reporting duties.

The Council’s significance is best seen against what preceded it. Before this Act, a mediator in India was made by reputation, by a court’s panel, by an institution’s roster or by a training certificate whose weight varied with its issuer, and no public register answered the question a stranger to the field must ask: is this person a mediator in any sense the law recognises? Chapter VIII, with Chapter IX, is the Act’s answer. Section 40 defines and Section 41 empowers the mediation service provider, the institution that conducts mediation under its own rules, maintains panels, appoints mediators on application and administers the process, and Section 42 provides for mediation institutes, the bodies that train and grade mediators and conduct examinations and research. The recognition of both classes of institution by the Council converts a market of names into a regulated field, and it does so in the manner the Act prefers: the state does not conduct mediation, and the Council does not mediate; the Council registers, recognises and sets standards, and the work is done by providers and institutes answerable to it. The proper caution, recorded in the commencement discussion above, is about sequence rather than design: the Council chapter is in force and the Council is established, and the registration and recognition regime operates as the Council’s regulations and the commenced provisions supply it with subject matter. Descriptions of the Council as a fully operational registry, like descriptions of the Act as a fully commenced code, run ahead of the notifications.

The consequential amendments: one process across eight statutes

Sections 58 to 65, with the Third to the Tenth Schedules, amend eight enactments: the Indian Contract Act, 1872, the Code of Civil Procedure, 1908, the Legal Services Authorities Act, 1987, the Arbitration and Conciliation Act, 1996, the Micro, Small and Medium Enterprises Development Act, 2006, the Companies Act, 2013, the Commercial Courts Act, 2015 and the Consumer Protection Act, 2019. The purpose of the set is harmonisation. Mediation had entered Indian law statute by statute over four decades, in the Legal Services Authorities framework, in Section 89 of the Code, in company and micro enterprise dispute provisions, in the consumer statute’s own mediation chapter and in the Commercial Courts regime, and each entry had its own vocabulary, its own settlement status and its own enforcing authority. The amendments align that scattered material with the hub statute: references are conformed, the settlement produced under the amended provisions is connected to the enforcement and challenge machinery of Sections 27 and 28, and the Commercial Courts amendment is the clearest example, since a settlement reached in pre-institution mediation under Section 12A of that Act is, under the amended text, to be dealt with in accordance with Sections 27 and 28 of this Act, so that the mandatory regime and the voluntary regime produce settlements of the same legal character.

The Arbitration and Conciliation Act amendment requires separate statement, because it touches the conceptual boundary this series has patrolled in its arbitration articles. Part III of the 1996 Act had provided for conciliation as a distinct statutory process, with its own settlement agreement and its own status for that agreement as an arbitral award on agreed terms. The Mediation Act’s definition had already absorbed conciliation into mediation as a matter of vocabulary; the amendment of the 1996 Act carries the alignment into that statute, so that the conciliation provisions are read within the unified field the new Act creates rather than as a rival code. The reader comparing the processes should hold the distinctions that survive the unification. Arbitration ends in an award imposed by a tribunal the parties chose but whose decision they do not control; mediation ends, if it ends in agreement, in a settlement the parties made and a mediator who decided nothing; conciliation, as Indian practice used the term, occupied the ground between, with the conciliator permitted a more active, proposing role. The arbitration pillar of this series and its companion on the amendments trace how the 1996 Act’s scheme was built and rebuilt; the present Act leaves arbitration’s adjudicatory core untouched and reorganises only the consensual territory beside it. The consumer amendment performs the same alignment for the mediation chapter of the Consumer Protection Act, 2019, under which consumer commissions refer disputes to mediation conducted through the cells attached to them, a route the consumer pillar of this series describes in its own setting; the statutory mediation of a consumer dispute and the general mediation of this Act are designed to produce settlements of the same enforceable character, and neither is a detour from the other.

Section 89 after the Act: the referral culture consolidated

The relationship between the Act and Section 89 of the Code of Civil Procedure is the point at which the new framework meets the daily practice of the trial courts, and it should be stated with care. Section 89 empowered the court, where it appeared that elements of a settlement existed which might be acceptable to the parties, to formulate the terms of a possible settlement and refer the dispute to arbitration, conciliation, judicial settlement including settlement through Lok Adalat, or mediation. The provision had known defects, catalogued in Afcons: its drafting conflated the reference stage with the settlement stage, it spoke of the court formulating terms in a manner better suited to some of its processes than to mediation, and the rules framed under it varied across High Courts. Yet for two decades it was the working charter of court referred mediation in India, and the centres, panels and habits of the system were built under it. The Mediation Act does not repeal Section 89, and the amendment of the Code in the Act’s schedules conforms the older provision to the new framework rather than abolishing it. The practical position is one of succession in function: the reference continues to be made from within the suit, under the Code and the High Court rules that govern the referring court, while the process referred to is the Act’s process, conducted by the Act’s mediators under the Act’s timetable and confidentiality, and producing the Act’s settlement with the Act’s decree status.

That succession matters to three recurring questions in the trial courts. The first is the stage of reference: Section 7’s at any stage formulation confirms what the better Section 89 practice had already established, that a reference is not confined to the first hearing, and that a dispute may become mediable after issues are framed, after evidence has clarified the parties’ real positions, or even in appeal, where the costs of the contest are no longer estimates but memories. The second is the effect of failure: the non-settlement report’s silence, examined above, is the Act’s guarantee that the reference was a detour and not a reconnaissance, and a court that receives such a report resumes the suit exactly where it left it, with the mediation’s contents sealed from the record. The third is the status of the settlement reached in a referred case: recorded under the Act and enforceable under Section 27, it disposes of the suit as a compromise decree would, and the referring court closes the proceeding on its strength. The consolidation is therefore complete in design: one door out of the suit, one process beyond it, one instrument back. Courts and counsel who continue to speak of Section 89 mediation and Act mediation as different species are describing a distinction the amendments were made to retire.

The international gap: signature without ratification

The Act is territorial, and the international position completes the account of its limits. The United Nations Convention on International Settlement Agreements Resulting from Mediation, known by its place of signature as the Singapore Convention, was opened for signature at Singapore on 7 August 2019, and India signed it on that date. India has not ratified it. The consequence is exact and should be stated exactly: the Convention’s regime, under which a mediated settlement of an international commercial dispute can be enforced directly in the courts of a state party without a fresh action on the contract, does not run between India and the other parties. A settlement mediated in India under this Act is enforceable in India as a decree; carried abroad, it is a contract to be sued on under the law of the place of enforcement. A settlement mediated abroad and brought to India is in the same position in reverse, and Section 2’s territorial anchor means the Act does not supply the defect, since it governs mediations conducted in India. The Act’s own category of international mediation, a mediation conducted in India involving a foreign party, is therefore a domestic category with an international ingredient, not a bridge to the Convention.

The Convention did not arrive alone. It was accompanied by amendments to the UNCITRAL Model Law on International Commercial Mediation, the soft law template from which many national mediation statutes, in their international chapters, are drawn, and the pairing states the international system’s design: a model law for the domestic framework and a convention for the cross border enforcement of its products. India’s domestic framework, this Act, was built without adopting the Model Law’s architecture in terms, and its enforcement chapter stops at the border for the reason already given, that the Act governs mediations conducted in India and speaks to Indian courts. The two gaps, the unadopted template and the unratified convention, are therefore one gap seen from two sides: the Indian mediated settlement is a finished domestic instrument and an unfinished international one, complete with decree status at home and contract status abroad, and the asymmetry will persist until the ratification decision is taken one way or the other.

The gap is a policy choice awaiting a decision rather than an oversight, and both sides of it are intelligible. Ratification would give Indian mediated settlements a passport, would align the treatment of mediated settlements with the treatment long given to arbitral awards under the New York Convention, and would answer the objection that India asks parties to mediate while withholding the cross border enforceability that makes mediation attractive to the commercial parties most likely to use it. The caution against ratification, as it was expressed during the Convention’s negotiation and India’s consideration of it, concerned the Convention’s application to settlements involving government parties and the breadth of its refusal grounds. This article takes no side in that pending choice; it records the position, because the position is the one most often misstated after the compulsion error itself. India has signed the Singapore Convention. India has not ratified it. A settlement does not cross a border on the strength of the signature.

The timetable in practice, and the report that says nothing

The figures of Section 18, one hundred and twenty days from the date fixed for the first appearance, extendable by consent by no more than sixty, repay a practical reading. The clock starts not from the agreement to mediate, nor from the mediator’s appointment, but from the first appearance date, which disciplines the front of the process: a party cannot consume the period in preliminary correspondence, because the period has not begun, and cannot extend it by absence, because the appearance date is fixed by the process and not by attendance. Within the period, the Act prescribes no number of sessions and no intervals; the mediator manages the calendar, and the parties’ consent is required only at the single point of extension. The extension’s design is a small essay in the Act’s philosophy: the state fixes the outer bound, the parties own everything inside it, and even the extension is theirs to give or withhold, so that a party who believes the process has run its course can end it by refusing consent and letting the clock expire. Termination on expiry under Section 24 is therefore not a failure of administration but one of the Act’s four designed endings, and the non-settlement report that follows under Section 21 is its designed record.

The non-settlement report deserves its reputation as the Act’s most elegant provision. It states that the mediation did not result in a settlement, and nothing else: no reasons, no findings, no apportionment of blame, no summary of offers. The elegance is functional. A report with reasons would be a document the parties would litigate about, draft against and fear; it would follow the returning litigant into the resumed proceeding and colour it; and it would convert the mediator, in the parties’ eyes, into a witness in waiting, destroying the candour on which caucus depends. By forbidding content, the section protects the confidentiality scheme of Sections 22 and 23 at its most vulnerable point, the moment of failure, and it protects the mediator’s future usefulness between the same parties. Its counterpart under the Section 12A regime, the report of non starter or failure issued by the authority there, performs the same office for the mandatory process and unlocks the plaint; the two reports are the twin hinges on which the voluntary and mandatory systems swing back into court.

The federal frame: High Court rules, state authorities and the court annexed inheritance

The Act is a central statute in a field the Constitution distributes. Civil procedure, including mediation as a mode of its conduct, falls within the Concurrent List, and administration of justice and the constitution of courts below the High Courts are state subjects in their daily operation; the machinery that delivers mediation on the ground, the district courts, the legal services authorities in their state and district formations, the court annexed centres, is state machinery working to state framed rules under High Court supervision. The Act does not disturb that distribution and could not have been drafted as if it did not exist. Its technique is the one the design predicts: standards, registration and recognition are centralised in the Council, while delivery runs through the inherited institutions, the court annexed mediation centres whose mediators the appointment provisions expressly recognise, the authorities under the Legal Services Authorities Act whose panels serve both the community chapter and the Section 12A regime, and the private providers and institutes the recognition system licenses. The High Courts’ mediation rules, framed under Section 89 of the Code and the rule making powers the new Act preserves, continue to govern the referred case in the court that referred it, and Section 7’s referral power is exercised within that local framework rather than over it.

The federal reading also explains a feature of the application provision that puzzles a first reader: the power of a State Government, alongside the Central Government, to notify disputes for the Act’s coverage. Mediation’s subjects are not uniform across the country; land, tenancy and local revenue disputes dominate one state’s trial courts and barely figure in another’s, and the notification power lets a state extend the statutory framework to the classes its own docket recommends, just as the First Schedule’s residuary power lets the Centre extend the exclusions. The Act is therefore less a uniform code than a national frame with state operated joints, and assessments of its working will have to be state by state for as long as the delivery machinery remains, as it is, a state inheritance.

Failure modes: why mediations fail, and what the Act does about each

A statute should be read against the ways its process fails in practice, and mediation’s failure modes are known well enough to be listed. The first is the absent party: a defendant who calculates that delay is profit and treats the invitation as correspondence. In the voluntary field the Act’s answer is structural rather than coercive: absence ends the attempt, the limitation exclusion protects the claimant’s interval, and the costs and referral powers of the court in the ensuing suit remain available to mark the conduct; in the mandatory field, Section 12A answers absence directly, since a plaintiff who has exhausted the remedy, including against a non appearing defendant, has satisfied the condition and may file. The second failure mode is the unauthorised party: the representative who attends without power to settle, so that every tentative agreement awaits a ratification that becomes a renegotiation. Section 49’s approval architecture for government parties is the Act’s express treatment of the public version of this failure, and the authentication and registration discipline is its treatment of the private version, since a settlement signed by a person without authority is the natural prey of the impersonation and fraud grounds. The third is the empty room of the other kind, the party present in body whose mandate is to concede nothing; no statute can draft against it, and the Act wisely does not try, relying instead on the mediator’s Section 24 declaration to end a process that has become a performance. The fourth is the settlement that unravels, performed in part and repudiated in part, and against it the Act deploys its heaviest machinery: decree status, execution, the bar of the four exclusive grounds and the ninety day window. The fifth is the process abused as reconnaissance, a party mediating to learn the opponent’s case and settlement range before trial; Sections 22 and 23 are the whole answer, and their exceptions are drawn narrowly enough that the reconnaissance yields nothing admissible. A framework that answers five known failures with five distinct devices, none of them a general power to compel, has a claim to be considered well designed on its own terms, whatever view is taken of the compulsion it withheld.

Mediation clauses after the Act: drafting the tier, and the med-arb warning

The Act will be encountered by most commercial parties first as a drafting question, and the drafting deserves a section of its own. The standard Indian dispute resolution clause before the Act ran from negotiation to arbitration, sometimes with a conciliation reference inherited from precedent books. The clause the Act invites is a genuine mediation tier: a defined trigger, a defined institution or appointment route, a defined period tied to the Act’s own timetable, and an express statement that the tier is a process obligation and not a settlement obligation. Three cautions from the Act’s scheme should govern the drafting. The first is to align the tier with the mandatory law rather than against it: in a commercial dispute of specified value, Section 12A will operate whatever the clause says, and a clause that purports to substitute a private timetable for the statutory pre-institution process invites the Patil Automation consequence at the door of the court. The second is to keep the settlement’s pedigree clean: the clause should route the mediation through the Act’s framework, a registered or empanelled mediator, an agreement in writing under Section 4, authentication under Section 19, so that the product carries Section 27 force without argument about which regime produced it. The third is the med-arb warning. Clauses that appoint the same neutral first to mediate and then to arbitrate are efficient in appearance and corrosive in fact: a mediator who has received confidences in caucus cannot thereafter decide the dispute without the parties fearing, and a court suspecting, that the decision rests on material never proved; the Act’s own separation of roles in Section 17, which forbids the mediator to act as arbitrator in the same dispute, states the principle the clause should not attempt to contract around. Parties who want both processes should sequence them under different neutrals, mediation first under this Act and arbitration thereafter under the 1996 Act, and the tiered clause that says so in terms is the clause the two statutes, read together, recommend.

When a settlement is attacked: the four grounds in practice

Because Section 28 will be litigated on facts rather than on text, each of its four grounds deserves a paragraph of anticipation, offered as analysis of the enacted words and not as a report of decisions that, at the time of writing, lie in the future of the provision’s commencement. Fraud, the first ground, reaches the settlement procured by deception: the account overstated, the document suppressed, the asset concealed, where the deception induced the bargain. Its difficulty in a mediation is evidentiary: the process that would prove the fraud is confidential and privileged under Sections 22 and 23, and a court hearing a fraud challenge will have to steer between the privilege that makes mediation possible and the inquiry that makes the challenge real, a steering the Act leaves to the courts by excepting from the privilege nothing in terms and by committing the challenge to the court of competent jurisdiction. Corruption, the second ground, is the rarest and gravest: it alleges that the process itself was bought, and its presence in the list states the obvious truth that a decree born of a corrupted process deserves no protection from the system that lent it force. Impersonation, the third, is the most mechanical and the most purely protective: a person who never attended, never consented and never signed is not to be bound because another person sat in the chair bearing the name, and the ground exists for the case of the substituted party, the forged authority and the agent who exceeded a mandate. The fourth ground, that the mediation was conducted in a dispute not fit for mediation under Section 6, is jurisdictional in character: it says the process was never authorised for this subject, and it makes the First Schedule a live document at the enforcement stage and not merely a checklist at intake.

The interplay of the four grounds with the ninety day period, extendable once by ninety days for sufficient cause, completes the design. Finality is not merely the absence of grounds; it is the brevity of the window. A settlement that survives ninety days unchallenged is, for practical purposes, a decree with the ordinary incidents of a decree and no special vulnerability, and execution may proceed on it accordingly. Advisers on the receiving end of a settlement should therefore treat the window as the settlement’s probation: preserve the Section 10 disclosures, the attendance record and the authentication, because those are the documents that answer a challenge, and diarise the period, because its expiry is the moment the settlement hardens. Advisers on the attacking end should note the converse: the grounds are exhaustive in form, and a grievance that is in substance that the bargain was bad will not be received in the dress of any of the four.

The docket argument, audited

No account of this Act is complete without examining the justification offered for it most often, which is the burden on the courts. The argument runs: the Indian courts carry a pendency counted in crores of cases; a large share of the civil docket consists of disputes the parties could settle if a structured opportunity were forced or offered early; therefore a mediation statute is a docket measure. The premise about pendency is beyond argument. The inference deserves the audit this series applies to any justification that proves too much. A docket argument justifies a settlement process only for the class of cases that would otherwise occupy the court, and it says nothing about the cases mediation draws from the far larger reservoir of disputes that were never going to be filed at all, where mediation’s value is not the court’s saved hour but the parties’ saved years. Conversely, the docket argument, taken alone, argues for compulsion, since a voluntary process relieves the docket only by the cases its volunteers bring; it was the docket argument in its strongest form that powered the Bill, and Parliament’s rejection of compulsion was, among other things, a judgment that the docket is not the only value in the system. Access to a court, the willingness that makes a settlement durable, and the quality of the outcome for the party who would have been steamrollered at a compulsory table all count against raw disposal numbers, and a mediation programme that clears dockets by exhausting litigants into bad bargains would show in the challenge statistics and in the return rate to court.

The audit therefore ends in a division of the claim. As a justification for building the institution, the docket argument is sound: a court system in India’s condition cannot afford to leave its most promising settlement mechanism unbuilt, and the Act’s institution building, the Council, the providers, the register, the timetable, is properly defended on docket grounds. As a justification for compulsion, the argument is incomplete, because it measures only the cases diverted and not the terms on which they were diverted, and the enacted Act is best read as accepting the first proposition and declining the second. The measure of the Act in time will accordingly not be the pendency figure alone. It will be the settlement rate within the process, the challenge rate against settlements, the proportion of settlements performed without execution, and the return of mediated parties to court on the same dispute, figures the Council’s depository and reporting functions exist to supply.

A short guide for the three readers of this Act

The brief for this article named three payloads, and they are collected here without ornament. For the party to a dispute and the adviser, the Act’s message is a sequence of questions: is the dispute within the First Schedule, in which case the Act is not the route; is it a commercial dispute of specified value, in which case Section 12A governs the start and the plaint’s fate depends on exhausting it or genuinely contemplating urgent relief; if neither, do both parties consent, in which case Section 5 opens a bounded, confidential process whose timetable is one hundred and twenty days plus sixty, whose cost is shared by default, whose period is excluded from limitation, and whose product, if signed and authenticated, is a decree. For the mediator and the service provider, the message is professional: registration, recognition, disclosure and grading are the new conditions of practice at the regulated end of the field, and the settlement’s safety depends on the intake discipline, the Schedule checked, the disclosures recorded, the authentication exact, that the challenge chapter presupposes. For the aspirant and the researcher, the message is the distinction with which this article began and on which it ends: the Act is the clearest recent instance of a Bill whose most discussed provision was reversed in passage, and any account that reports the Bill’s compulsion as the Act’s law repeats the error the statute itself was enacted, in part, to correct. Examination answers, professional summaries and editorial commentary on this Act should be graded, in the first instance, on whether they get that single point right.

What the Act does not do

A statute pillar owes the reader the negative account as plainly as the positive, and this Act’s limits are as instructive as its provisions. It does not apply to a mediation conducted outside India, however Indian the parties: the territorial anchor in Section 2 is unqualified, and the enforcement chapter’s decree is an Indian decree for an Indian process. It does not disturb proceedings already pending at its commencement in the manner of a retrospective code: Section 56 keeps the Act away from pending court and tribunal proceedings in the sense the section prescribes, and Section 57 supplies the transitory rule for mediations already on foot, so that processes begun under the older scattered law finish under the understandings on which they began. It does not nationalise the practice of mediation: private mediation by agreement, conducted outside the Act’s framework by persons the parties trust, is not prohibited, and what such a process cannot claim is the Act’s settlement status, which is the statute’s inducement to come within the frame rather than a penalty on staying outside it. It does not make the mediator a judge in any particular: no power to compel evidence, to summon witnesses, to order interim protection or to decide any question is conferred anywhere in its sixty five sections, and a party who wants any of those things must go to the court whose interim powers the Act preserves. And it does not, as the commencement account showed, operate all at once: the reader who acts on a provision should first ask the commencement question the Act’s own design makes necessary, a discipline this article has applied to itself throughout.

Compulsion elsewhere: what the comparisons actually show

Because both sides of the Indian debate argued from foreign experience, the comparisons deserve a paragraph of audit rather than a paragraph of rhetoric. Italy is the strongest case for the proponents: its legislative decree of 2010 made an attempt at mediation a condition of suit for defined categories of civil and commercial dispute, the requirement survived constitutional challenge in modified form and was re enacted, and Italian experience is regularly cited for the proposition that a large share of compelled mediations settle once the parties are in the room, including many in which one party attended under protest. The detail that travels badly is the design of the compulsion: the Italian requirement is satisfied by an initial meeting at modest cost, from which a party may withdraw, so that what is compelled is an encounter with the process and an informed opportunity to use it, not the exhaustion of a full procedure. England supplies the counterweight from a common law system protective of access to court. For decades its courts encouraged alternative dispute resolution by costs consequences rather than by condition, penalising unreasonable refusal in the costs order at the end; and in Churchill v Merthyr Tydfil County Borough Council, decided in 2023, the Court of Appeal held that a court may lawfully stay proceedings and order the parties to engage in a non court based dispute resolution process, provided the order does not impair the essence of the right to a judicial hearing and is proportionate to settling the dispute fairly, quickly and at reasonable cost. The English position after Churchill is therefore not that compulsion is impermissible, but that it is judicial, calibrated and subsequent to the filing of the claim, a matter of case management rather than a statutory toll at the door.

Read together, the two comparisons locate the Indian compromise with some precision. The Bill proposed statutory, pre suit, full process compulsion across civil and commercial litigation: stronger than Italy’s first meeting model in its demand, and stronger than England’s judicial model in its source and its timing. The Act as enacted compels nothing at the door, leaves courts their referral power and their costs jurisdiction, and preserves the one statutory compulsion, Section 12A, whose field is commercial, whose value threshold excludes the smallest claims, and whose operation the Supreme Court disciplined with prospective effect and a policed urgency exception. A reformer who wants more compulsion in India has therefore to argue for the Italian design, a compulsory first session within a voluntary statute, and not for the Bill’s design revived; and a critic who wants none has to explain why the commercial exception, in operation for several years and judicially supervised throughout, should be repealed rather than contained. Neither argument was the one Parliament heard in 2021, and the distance between the debate that occurred and the debate the evidence supports is the last illustration of this article’s thesis: on this statute, the enacted text and the public description of it have rarely been the same document.

The rim in detail: mediation inside the subject statutes

The hub and rim metaphor earns its keep in a short tour of the rim, because each subject statute had solved the settlement problem in its own idiom before the hub existed, and the amendments make sense only against the idioms. The oldest idiom is the labour one. The Industrial Disputes Act, 1947 built conciliation into the state itself: conciliation officers appointed by government, charged with mediating in industrial disputes and inducing settlements, with boards of conciliation beside them and a statutory duty to report success or failure to the government that appointed them. That model, conciliation as a public office rather than a private engagement, formed the template Indian administrators reached for whenever a new field needed a settlement mechanism, and its fingerprints are on the facilitation councils of the micro and small enterprises statute, which apply the conciliation provisions of the Arbitration and Conciliation Act to delayed payment disputes and decide the claim themselves if conciliation fails, a hybrid of mediator and tribunal that the pure model of the present Act, with its strict separation of facilitation from decision, deliberately is not.

The Companies Act, 2013 took a different route, empowering the Central Government to maintain a panel of experts as mediators or conciliators for disputes pending before the company law tribunals, a roster model in which the state supplies names and the tribunal supplies the reference. The Consumer Protection Act, 2019 built the most elaborate of the subject schemes: a mediation chapter of its own, mediation cells attached to the district, state and national commissions, referral by the commission at the stage of admission or at any later stage with the parties’ consent, and a settlement recorded and given effect within the consumer proceeding. The Real Estate (Regulation and Development) Act, 2016 empowered its authorities to facilitate amicable resolution through their own machinery, and the family courts legislation had, a generation earlier, made it the duty of the Family Court to endeavour a settlement in every matter before it, assisted by counsellors attached to the court. Each of these schemes answered, for its field, the questions the hub answers generally: who convenes, who facilitates, what the settlement is worth, and what happens on failure. Their coexistence with the general Act is not redundancy. A consumer commission’s cell, a facilitation council and a family court’s counsellor are institutions with premises, staff and a clientele formed over years, and the Act’s amendments align the status of their settlements with Sections 27 and 28 while leaving the institutions standing, which is the federal and administrative realism the design section of this article described. The practitioner consequence is a choice of rooms in the fields the subject statutes cover, and the choice should be made consciously, because the room determines the panel, the cost and the route by which the settlement hardens into an executable instrument, even where the hardening standard is aligned.

Rules, regulations and the unfinished commencement

The last piece of the framework is the subordinate legislation, and its state is the practical corollary of the commencement account given earlier. The Act distributes rule making in the pattern of Indian framework statutes: the Central Government makes rules under Section 51 on the matters the Act commits to it, the Council makes regulations under Section 52 on the professional and institutional matters within its charge, and both species of instrument are laid before Parliament under Section 53. Sections 51 to 54 were among the provisions commenced on 9 October 2023, so the power to build the machinery has been exercisable from the first appointed day, and the sequence the Act contemplates is legible in its own commencement pattern: machinery provisions first, the Council established under them, regulations for registration and recognition as the Council’s first work, and the operative chapters’ appointed days as the framework becomes capable of receiving the work the chapters assign to it. The Mediation Fund under Section 45, fed by grants, fees and sums the Act directs into it and applied to the promotion of mediation and the Council’s purposes, and the accounts and audit provisions of Section 46, complete a design in which the institution is financed and audited as a public body while the service it regulates is delivered by others.

For the reader, the consequence is a habit of reading rather than a body of doctrine. Any statement about the Act’s operation in a particular field should be tested against three questions: has the provision been commenced, have the rules or regulations it presupposes been made, and is the institution it names in a position to act. The Act’s text answers none of the three by itself, and the confusion in the commentary on this statute is, in large part, the product of answering all three from the text alone. This article has applied the test to its own statements throughout, describing the enacted scheme in full because a statute pillar exists to expound the enacted scheme, and marking the commencement position at each point where operation, rather than design, is the subject. The distinction will narrow as notifications accumulate; the habit will remain the correct one for every framework statute that follows this one’s pattern, which is to say, on the evidence of recent Indian legislation, for a good proportion of the statutes this series has still to expound.

The Act’s wager

Every statute embodies a wager about behaviour, and this Act’s wager is unusually visible because the alternative was drafted, debated and withdrawn in public. The Bill wagered that Indian litigants would not mediate unless made to, and that the system’s need for settlements justified making them. The Act wagers the opposite sequence: that the obstacles to mediation were never principally the litigants’ unwillingness, but the absence of the things the Act supplies, a process whose product is worth having, a profession whose members can be identified and trusted, a timetable that bounds the risk of trying, a confidentiality that makes trying safe, and institutions that make the process findable. On that diagnosis, compulsion treats the wrong variable: parties did not refuse mediation so much as decline a process whose settlement was a mere contract, whose mediator was a stranger of unverified standing, and whose failure cost limitation time and tactical exposure. Remove those defects, the wager runs, and the demand the docket argument predicts will present itself voluntarily, drawn first from the commercial field where the costs of litigation are best understood, and spreading as the register, the depository’s published record and professional habit do their work.

The wager is falsifiable, which is a virtue in a statute as in an argument. If, with the framework commenced and the institutions operating, voluntary uptake remains negligible outside the compulsory commercial field and the court referred stream, the Bill’s diagnosis will have been vindicated after the event and the case for an Italian style compulsory first session will return with evidence behind it. If uptake grows as the defects are removed, the enacted design will stand as the rare legislative instance in which a Parliament, told by its own committee that the premise of a Bill was mistaken, rewrote the premise rather than merely amending the detail. Either outcome is knowable from the Council’s monitoring, and the existence of a statutory mechanism for knowing it, the Section 39 reporting duty and the Section 38 depository, is itself part of the design’s intelligence. A legislature that builds the instrument by which its own wager will be judged has done something the series recording Indian legislation has had too few occasions to praise.

The compulsion question, judged

It remains to return to the debate the article opened with, and to judge it as the stance charter of this series requires. The case for the Bill’s compulsion, restated shortly, is that voluntary mediation is unused mediation: the nudge must be a requirement, the cost of the requirement is a bounded period excluded from limitation, and attendance can be compelled without compromising the voluntariness of any settlement, which remains unsigned unless both parties choose it. The case against, restated shortly, is that a compelled mediation with an unwilling party is a formality that taxes access to court, that settlement rates among the compelled measure the cases that would have settled anyway, and that the state’s proper instrument for its docket problem is the supply of courts and mediators rather than the conscription of litigants. The Standing Committee’s recommendation and Parliament’s acceptance of it resolved the question for the general law, and the resolution this article defends is that Parliament was right, for a reason narrower than either side’s rhetoric. Compulsion is defensible where the parties are commercial actors of comparable sophistication, the dispute is valued and documented, and the process is administered by an authority with panels and timelines in being; those are the conditions of the Section 12A regime, and they are the conditions the Supreme Court’s prospective application in Patil Automation itself acknowledged when it refused to destroy plaints filed before the infrastructure existed. Compulsion across the whole field of civil disputes, where one party is routinely an individual and the other an institution, and where the mediation infrastructure outside the commercial centres is thin, would have reproduced the worst feature of Indian procedure, a mandatory step whose function is performed by adjournment. The enacted compromise, voluntary generally and mandatory for commercial disputes of specified value, is therefore not a failure of nerve. It is a judgment about where compulsion earns its keep, and the judgment is defensible on the evidence the passage of the Bill itself supplied.

Within the courts’ own landscape, the Act takes its place as the consensual member of a family of processes the judicial system survey in this series maps as a whole: adjudication that decides, arbitration that decides by consent to be bound, Lok Adalats that settle with the authority of the legal services institutions, and mediation that settles by consent alone and borrows its force, after the event, from the decree status Section 27 confers. This article is general legal information on the statute and not advice on any dispute. The Act’s contribution to that landscape is real but should be measured in its own terms. It did not conscript the Indian litigant into mediation. It built the room, regulated the person in the middle of it, fixed the clock on the wall, sealed the door, and made the paper signed inside it worth a decree. Whether litigants enter the room in numbers is a question the statute, having been made voluntary, deliberately leaves to them.

A final word on reading the Act itself is owed to the student who meets it first in a bare text. The Act rewards being read in the order of its own logic rather than the order of its numbering: Section 3 for the definition that fixes what the process is, Section 5 with its proviso for the question of compulsion, Section 6 with the First Schedule for the boundary, Sections 18 to 20 for the discipline of the process and its product, Sections 27 and 28 for what the product is worth and how it may be undone, and only then the institutional chapters for the machinery that sustains the whole. Read in that order, the Act’s argument is unmistakable, and it is the argument this article has followed: consent is the principle, the decree is the inducement, the Schedule is the boundary, the Council is the guarantor, and compulsion is the exception that another statute owns. A reader who carries those five propositions away has the statute, whatever detail later commencement notifications and the Council’s regulations add to its working life.

Frequently Asked Questions

Q: What does the Mediation Act 2023 provide?

The Act, Act 32 of 2023, is India’s first dedicated central statute on mediation. It defines mediation to include conciliation and pre-litigation, online and community mediation; states where it applies; provides for mediation agreements and voluntary pre-litigation mediation; lists disputes not fit for mediation; empowers courts to refer pending disputes; regulates the appointment and conduct of mediators; fixes a timetable of one hundred and twenty days extendable by sixty; protects confidentiality; makes a mediated settlement agreement enforceable as a decree; creates the Mediation Council of India; and amends eight other statutes to align their mediation provisions with it.

Q: Is pre-litigation mediation compulsory?

No, not under this Act. Section 5 provides that parties may, voluntarily and with mutual consent, take steps to settle a dispute by pre-litigation mediation before filing a civil or commercial suit, whether or not a mediation agreement exists. The Bill as introduced in 2021 proposed compulsion; the Standing Committee recommended voluntariness, and Parliament enacted the voluntary form. The one compulsory regime is separate and older: Section 12A of the Commercial Courts Act, 2015 requires pre-institution mediation for commercial disputes of specified value where no urgent interim relief is contemplated, and the proviso to Section 5 preserves that regime rather than creating it.

Q: Which disputes cannot be mediated?

Section 6 bars mediation for the disputes listed in the First Schedule. They include claims against minors, deities and persons with intellectual disabilities; matters relating to the prosecution of criminal offences; disputes involving allegations of serious and specific fraud, fabrication of documents, forgery, impersonation or coercion; disputes about the levy, collection, penalties or refunds in relation to direct and indirect taxes; proceedings before named regulators and tribunals, including competition, telecom, electricity, securities and environmental forums; land acquisition compensation disputes; professional discipline proceedings; probate matters; and disputes affecting the rights of third parties who are not participating, with an exception for matrimonial disputes involving a child’s interest.

Q: How long does a mediation have to run?

Section 18 fixes the timetable. Mediation proceedings are to be completed within one hundred and twenty days from the date fixed for the first appearance before the mediator. The parties may by consent extend that period, but the extension cannot exceed sixty days, so one hundred and eighty days is the outer limit of a mediation under the Act. If the period expires without a settlement, the mediation terminates and the mediator submits a non-settlement report under Section 21, which records the failure without reasons and without attributing blame to either party.

Q: How is a mediated settlement agreement enforced?

A mediated settlement agreement must be reduced to writing, signed by the parties and authenticated by the mediator, as Section 19 requires. Under Section 27 it is then final and binding on the parties and on persons claiming under them, and it is enforced in accordance with the Code of Civil Procedure, 1908 in the same manner as if it were a judgment or decree passed by a court. No fresh suit on the settlement is needed: the holder proceeds by execution, as with any decree, and any party may also rely on the settlement by way of defence or set off in other legal proceedings.

Q: On what grounds can a mediated settlement be challenged?

Only on the four grounds Section 28 exhaustively states: fraud, corruption, impersonation, and that the mediation was conducted in a dispute or matter not fit for mediation under Section 6. The challenge is made by application to the court or tribunal of competent jurisdiction within ninety days of the applicant receiving the copy of the settlement agreement under Section 19. If the court or tribunal is satisfied that sufficient cause prevented an earlier application, it may entertain the application within a further ninety days. There is no challenge for error, unfairness or an improvident bargain, which is the price of the settlement’s finality.

Q: What does the Mediation Council of India regulate?

The Council, established under Chapter VIII with its head office at Delhi, is the Act’s regulator for the profession. Under Section 38 it promotes mediation in India, lays down guidelines and codes of conduct for mediators, registers mediators and renews, suspends or cancels their registration, recognises mediation service providers and mediation institutes and supervises that recognition, lays down standards, curricula and training requirements, and maintains an electronic depository of mediated settlement agreements. It does not itself conduct mediations. Sections 31 to 38 were among the provisions brought into force on 9 October 2023, and the Council was established by notification under Section 31.

Q: What are community mediation and online mediation?

Community mediation, under Chapter X, settles disputes likely to affect the peace, harmony and tranquillity among residents or families of an area or locality, with the prior mutual consent of the parties, through a panel of three community mediators maintained by the Legal Services Authority or the District Magistrate or Sub Divisional Magistrate. Online mediation, under Section 30, is mediation conducted at any stage through electronic forms, computer networks, electronic mail, video conferencing or audio conferencing, with the written consent of the parties and subject to the same confidentiality discipline as a mediation conducted in person.

Q: When did the Mediation Act 2023 come into force?

The Act received the assent of the President on 14 September 2023 and was published in the Gazette on 15 September 2023, but its commencement is staged under Section 1(3). By notification S.O. 4384(E), a specified set of provisions was brought into force on 9 October 2023: Sections 1 and 3, Section 26, Sections 31 to 38, which contain the Mediation Council chapter, Sections 45 to 47, Sections 50 to 54 and Sections 56 and 57. The operative chapters, including Sections 2, 4 to 25 and 27 to 30 and the amendments in Sections 58 to 65, stand enacted and await appointed days.

Q: What is a mediation agreement under the Act?

Section 4 provides for a mediation agreement in writing, by which the parties agree to submit to mediation all or certain disputes that have arisen or may arise between them. It may take the form of a mediation clause in a contract or of a separate agreement, and it may be concluded before a dispute arises or after. A mediation agreement does not itself compel a settlement, and a party cannot be forced to sign one; it records consent to the process. Section 5 operates independently of such an agreement, so parties with no mediation clause may still choose pre-litigation mediation by mutual consent.

Q: Who can be appointed as a mediator under the Act?

Section 8 leaves the choice of mediator to the parties, who may agree on the person and the appointment procedure; failing agreement, a mediation service provider appoints from its panel, considering suitability and the parties’ preference under Section 9. For pre-litigation mediation the mediator must come from the regulated end of the profession: a person registered with the Mediation Council of India, or empanelled with a court annexed mediation centre, or with an authority under the Legal Services Authorities Act, 1987, or with a mediation service provider. A person approached must disclose any circumstance raising a justifiable doubt about independence or impartiality, as Section 10 requires.

Q: Can a court refer a pending dispute to mediation?

Yes. Under Section 7, a court or tribunal before which a proceeding is pending may refer the parties to mediation at any stage where it appears that elements of a settlement exist which may be acceptable to them, and the parties may jointly invite such a reference. Referral does not transfer the dispute: if the mediation produces no settlement, the proceeding resumes with the parties’ positions protected by the confidentiality provisions of Sections 22 and 23. If it produces a settlement, the agreement is written, signed and authenticated under Section 19 and carries the enforcement of Section 27.

Q: Is a mediation under the Act confidential?

Yes, in two reinforcing ways. Section 22 obliges the mediator, the mediation service provider, the parties and all participants to keep confidential every matter relating to the proceedings, extends that duty to the settlement agreement except where disclosure is necessary for its implementation or enforcement, and prohibits any audio or video recording of the sessions. Section 23 makes mediation communications privileged: a court or tribunal cannot compel disclosure of what was said or produced in the mediation, subject to exceptions for threats of criminal conduct, domestic violence, child abuse and imminent threats to public safety. The non-settlement report gives no reasons, so nothing from the room follows the parties back into court.

Q: What happens if a mediation ends without a settlement?

The mediation terminates under Section 24, on the mediator’s written declaration that further mediation is not justified, on a party’s written withdrawal, or on the expiry of the Section 18 period. The mediator then submits a non-settlement report under Section 21, which states only that no settlement was reached, without reasons and without blame. The parties return to the position they occupied before the process: a suit not yet filed may be filed, and a referred proceeding resumes. Section 29 protects the interval, excluding the mediation period from the computation of limitation for the dispute.

Q: Does the time spent in mediation count toward limitation?

No. Section 29 provides that, notwithstanding the Limitation Act, 1963 or any other law, in computing the period of limitation for any proceeding relating to a dispute in respect of which mediation has been undertaken under the Act, the period from the commencement of the mediation under Section 14 until the submission of the report under Section 21 or the termination of the mediation under Section 24 is excluded. A party who spends four months at the table and leaves without a settlement therefore loses no limitation time by having tried, which removes the standard objection to attempting a pre-suit process.

Q: Who bears the cost of a mediation under the Act?

Section 25 supplies a default rule: unless the parties agree otherwise, the fees of the mediator and the costs of the mediation are to be borne equally by the parties. The equal division reflects the Act’s premise that the process belongs to both sides rather than to an applicant seeking a favour, and it gives each party the same stake in the timetable the Act imposes. Parties remain free to agree a different division, and settlements frequently allocate the costs as one term of the bargain recorded in the mediated settlement agreement. Community mediation stands outside this default sharing rule.

Q: What is institutional mediation under the Act?

Institutional mediation is mediation conducted with the administrative support of a mediation service provider, the body Chapter IX of the Act recognises. Under Sections 40 and 41, a service provider maintains panels of mediators, appoints a mediator when the parties apply and cannot agree on one, provides facilities and secretarial support, and conducts the process under its own rules consistently with the Act. Mediation institutes, recognised under Section 42, train and grade mediators and conduct examinations and research. The Mediation Council recognises both classes of body and may renew, suspend or cancel recognition, which is how the Act turns a scattered practice into a regulated field.

Q: Does the Mediation Act apply to a dispute involving the government?

It can. Section 2 extends the Act to a mediation conducted in India where one of the parties is the Central Government or a State Government, or an agency, public body, corporation or local authority controlled by them, and the dispute is a commercial dispute, as well as to any other dispute the government notifies for the purpose. A qualification follows in Section 49: where a government or a controlled entity is a party, the mediated settlement agreement requires the prior written approval of the competent authority before it binds the government party, so a mediation involving the state has an approval stage after the negotiation stage.

Q: Has India ratified the Singapore Convention on Mediation?

No. India signed the United Nations Convention on International Settlement Agreements Resulting from Mediation, known as the Singapore Convention, at Singapore on 7 August 2019, and has not ratified it. Signature expresses an intention to consider the instrument; ratification is what would bind India to it. Until ratification, the Convention’s direct enforcement regime does not run for India: a settlement mediated in India and taken abroad must be enforced as a contract under the law of the place of enforcement, and a settlement mediated abroad gains no direct enforceability in India from the Convention. The Mediation Act, which governs mediations conducted in India, does not fill that gap.

Q: How does the Mediation Act change conciliation under the Arbitration and Conciliation Act 1996?

The Act unifies the vocabulary and then amends the statute. Section 3 defines mediation to include a process referred to as conciliation or by an expression of similar import, and Sections 58 to 65, with their Schedules, amend the Arbitration and Conciliation Act, 1996 along with seven other enactments, aligning the conciliation provisions of Part III of that Act with the unified field. The adjudicatory core of the 1996 Act is untouched: arbitration still ends in an award a tribunal imposes, while mediation and conciliation produce only the settlement the parties themselves sign, enforceable under Section 27 of this Act.