Landmark Cases — Subhash Kashinath Mahajan, Prithvi Raj Chauhan

Few statutory provisions have travelled through the Supreme Court as dramatically as the safeguards in the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989. In a span of under two years, a two-Judge Bench read in elaborate pre-arrest safeguards (Subhash Kashinath Mahajan), a three-Judge Bench recalled those directions in review (Union of India v. State of Maharashtra), Parliament inserted a fresh non-obstante provision (Section 18A), and a Constitution-conscious Bench upheld that amendment while preserving a narrow judicial escape hatch (Prithvi Raj Chauhan). This chapter unpacks that trilogy section-by-section, tying each holding to the bare text of Sections 18 and 18A and to the surrounding case law on arrest, anticipatory bail and the ingredients of an atrocity. For the statutory scheme it sits on, read the SC/ST (Prevention of Atrocities) Act hub.

Why this trilogy of cases matters

The 1989 Act is special, deterrent, and victim-protective legislation. To make it bite, Parliament built in two unusual procedural levers: Section 18, which ousts anticipatory bail under Section 438 of the Code of Criminal Procedure, 1973, and (after 2018) Section 18A, which forecloses any preliminary enquiry before registration of the FIR and any prior approval before arrest. Both levers tilt the criminal process sharply in favour of swift action on a complainant's word.

That tilt produced a recurring tension. On one side stands Article 17 (abolition of untouchability) and Article 15(4)'s mandate of protective discrimination; on the other stand Articles 14, 19 and 21, which protect even an accused from arbitrary arrest and false implication. The three decisions studied here are the Supreme Court's attempt to hold that balance — first by judicial reading-in, then by retraction, and finally by a constitutional accommodation. Understanding them requires fluency in the punishment scheme and the specific offences the Act creates.

The constitutional and statutory backdrop

The Act is the legislative expression of Article 17, which abolishes untouchability and forbids its practice in any form. In State of Karnataka v. Appa Balu Ingale, AIR 1993 SC 1126, where high-caste villagers had stopped Dalits from drawing water from a public borewell, the Supreme Court restored convictions under the Protection of Civil Rights Act, 1955 and read Article 17 expansively — describing untouchability as an indirect form of slavery and an extension of the caste system, and holding that mens rea is not an essential ingredient of such social-welfare offences. Justice K. Ramaswamy's concurring opinion treated Article 17 as a charter of human dignity demanding purposive, socially responsive interpretation.

That dignitarian reading frames the entire Atrocities-Act jurisprudence. Section 18 of the 1989 Act provides that "nothing in Section 438 of the Code shall apply in relation to any case involving the arrest of any person on an accusation of having committed an offence under this Act." Read literally, it is an absolute bar on anticipatory bail. The whole trilogy turns on whether that bar is truly absolute, and whether courts may carve out exceptions where the complaint discloses no offence at all. For the conceptual foundation, see Introduction, constitutional background and object.

The constitutionality of Section 18 was not novel even in 2018. In State of M.P. v. Ram Krishna Balothia, (1995) 3 SCC 221, a two-Judge Bench had already upheld Section 18 against an Article 14 and Article 21 challenge, reasoning that offences under the Act form a distinct class — atrocities committed because of caste or tribe identity — and that the exclusion of anticipatory bail bears a rational nexus to the object of protecting a vulnerable community from intimidation and the abuse of the bail process to terrorise victims. Balothia is the doctrinal anchor for treating Atrocities-Act accused as a separate class; both the 2019 review Bench and Prithvi Raj Chauhan drew on the same classification logic when they restored and then sustained the statutory bar.

Subhash Kashinath Mahajan: facts and the problem of misuse

Dr. Subhash Kashinath Mahajan v. State of Maharashtra, (2018) 6 SCC 454 (also reported as AIR 2018 SC 1498), was decided on 20 March 2018 by a two-Judge Bench of Justices Adarsh Kumar Goel and Uday Umesh Lalit. The appellant was the Director of Technical Education, Maharashtra. The trigger was a complaint against two storekeepers under the Atrocities Act for allegedly making adverse entries in an employee's service record; when the complainant sought to prosecute them, sanction was the issue, and the appellant — who had declined sanction in his capacity as the disciplinary authority — found himself named.

The Bench used the appeal as an occasion to address what it perceived as widespread misuse of the Act against public servants and private citizens through frivolous or motivated complaints. It marshalled data on acquittal rates and false cases, and reasoned that protecting the dignity of the SC/ST community could not come at the cost of the liberty of the innocent, invoking Articles 14 and 21.

Two threads of the appellant's grievance deserve emphasis because they shaped the Court's anxiety. First, the complaint reached a person who had only exercised a quasi-judicial discretion on sanction — illustrating how, in the Court's view, the Act could be deployed against officials for bona fide administrative decisions. Second, the bar on anticipatory bail under Section 18 meant that even such an official faced the prospect of arrest before any judicial scrutiny of whether an atrocity was disclosed at all. The Bench treated these as systemic features rather than one-off facts, which is why it framed remedies of general application rather than confining relief to the appellant. That choice — converting a fact-specific grievance into a charter of safeguards — is precisely what the review Bench would later fault as overreach.

The five Mahajan directions

To curb misuse, the Mahajan Bench issued a set of prospective directions that effectively re-engineered the Act's procedure:

(i) There is no absolute bar to anticipatory bail under the Act; Section 18 would not foreclose anticipatory bail where the complaint, on judicial scrutiny, makes out no prima facie case or is found to be mala fide.

(ii) Before registering an FIR, a preliminary enquiry may be conducted by a Deputy Superintendent of Police to verify whether the allegations disclose an offence under the Act and are not frivolous or motivated.

(iii) Arrest of a public servant requires prior written approval of the appointing authority; arrest of a non-public servant requires prior written approval of the Senior Superintendent of Police, with reasons recorded.

(iv) The magistrate must apply mind to those reasons before permitting further detention.

(v) These safeguards were declared applicable retrospectively to pending matters as well. The Court drew heavily on its arrest jurisprudence — Joginder Kumar v. State of U.P., (1994) 4 SCC 260, and Arnesh Kumar v. State of Bihar, (2014) 8 SCC 273 — observing that arrest "brings humiliation, curtails freedom and casts scars forever," and that arrest must satisfy the necessity tests embedded in Section 41 CrPC.

The reasoning — and the criticism it drew

Doctrinally, the directions borrowed the architecture of Lalita Kumari v. Government of U.P., (2014) 2 SCC 1, where a Constitution Bench held FIR registration mandatory for cognizable offences but permitted a preliminary enquiry in narrow categories (matrimonial, commercial, medical-negligence, corruption, and abnormally delayed complaints). Mahajan in effect added Atrocities-Act complaints to that exceptional list and went further by requiring administrative pre-approval before arrest — a safeguard found nowhere in the statute.

The decision was immediately controversial. Critics argued that the Court had legislated, diluting a protective statute by judicial fiat and ignoring that the very purpose of Section 18 was to prevent the intimidation of complainants through bail manoeuvres. The judgment sparked nationwide protests in April 2018 and prompted both a review petition by the Union of India and a swift legislative response. The episode is a textbook illustration of the limits of reading-in safeguards into special penal statutes — a theme that runs through the aggravated forms of atrocity as well.

The 2019 review: Union of India v. State of Maharashtra

In Union of India v. State of Maharashtra, (2020) 4 SCC 761, decided on 1 October 2019, a three-Judge Bench (Justices Arun Mishra, M.R. Shah and B.R. Gavai) allowed the review and recalled the principal directions in Mahajan. The Bench held that the two-Judge Bench had transgressed its province: the directions amounted to judicial legislation, conflicting with the statutory text and with the protective object of the Act under Article 15(4).

The Court reasoned that creating a class of "misuse" exceptions presumed that members of the SC/ST community routinely file false cases — a presumption the Bench refused to endorse, memorably observing that one "cannot treat all of them as a liar" or as a group prone to false implication. It restored the position that an FIR must be registered without a preliminary enquiry and without the requirement of prior administrative approval for arrest, and that Section 18's bar on anticipatory bail operates as the statute provides. Crucially, however, the Bench preserved the narrow proposition that anticipatory bail is not barred where the complaint discloses no prima facie offence under the Act — an exception rooted in judicial review of the FIR itself rather than in a misuse presumption.

The review judgment also corrected a methodological flaw in Mahajan. The earlier Bench had relied on conviction-and-acquittal statistics to infer rampant false implication; the review Bench pointed out that a low conviction rate is a function of investigative shortfalls, hostile witnesses, social pressure and turned-hostile victims, not proof that complaints are fabricated. Treating acquittal as a proxy for falsity, it held, was both empirically unsound and constitutionally suspect, because it stigmatised an entire protected class. The Court added that the safeguard against genuinely false cases already exists in the general law — Section 482 CrPC to quash, and the ordinary power to grant relief where no offence is disclosed — so there was no gap that justified judicial reading-in. This reasoning is what makes the 2019 review more than a correction of result; it is a correction of the interpretive premise. The contrast with Mahajan is a favourite examiner's hook on the limits of using crime statistics in statutory interpretation.

Parliament's reply: the 2018 Amendment and Section 18A

Even before the review was decided, Parliament had acted. The Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Amendment Act, 2018 (Act 27 of 2018) inserted a new Section 18A with effect from 20 August 2018, expressly to nullify the Mahajan directions. Section 18A provides that (1) for the purposes of the Act, (a) preliminary enquiry shall not be required for registration of an FIR against any person; and (b) the investigating officer shall not require approval for the arrest, if necessary, of any person against whom an accusation of an offence under the Act has been made, and no procedure other than that provided under the Act or the Code shall apply.

Sub-section (2) is the sharpest: it states that the provisions of Section 438 of the Code shall not apply to a case under the Act, "notwithstanding any judgment or order or direction of any Court." The non-obstante clause was a direct legislative override of the Mahajan reading that Section 18 left room for anticipatory bail. The amendment thus restored the unconditional statutory bar and abolished both the preliminary enquiry and the pre-arrest approval requirement.

The legislative technique here is worth noting for theory questions. Section 18A is a classic validating-and-overriding provision: rather than purport to set aside the judgment directly (which the legislature cannot do), it removes the legal basis on which the Mahajan directions rested by re-stating the procedure in unambiguous statutory language and by expressly excluding contrary judicial directions. Parliament did not declare Mahajan wrong; it altered the law so that the directions could no longer stand on it. This respects the separation of powers — the legislature changes the law prospectively, the judiciary applies the changed law — and it is the same device the Supreme Court approved when it sustained Section 18A in Prithvi Raj Chauhan. The 2018 protests that preceded the amendment, and the political salience of the Act, form the backdrop but do not bear on its validity; what mattered was legislative competence and conformity with Articles 14, 19 and 21.

Prithvi Raj Chauhan: the constitutional challenge to Section 18A

Section 18A was promptly challenged. In Prithvi Raj Chauhan v. Union of India, (2020) 4 SCC 727 (also reported as 2020 SCC OnLine SC 159), decided on 10 February 2020, petitioners contended that the bar on anticipatory bail and the mandate of arrest without enquiry or approval permitted "automatic arrest" and violated Articles 14, 19 and 21. The argument was that an irrebuttable foreclosure of pre-arrest liberty, with no judicial filter, was manifestly arbitrary.

The three-Judge Bench (Justices Arun Mishra, Vineet Saran and S. Ravindra Bhat) unanimously upheld the constitutional validity of the 2018 Amendment. Justices Mishra and Saran reasoned that much of the challenge had become academic because the directions the amendment targeted had themselves been recalled in the 2019 review. The Bench reaffirmed that the special character of the Act and its Article 15(4) foundation justified the stricter procedure, and that the legislature was within its competence to remove judicially engrafted safeguards.

The narrow carve-out: when anticipatory bail survives

The decisive qualification in Prithvi Raj Chauhan is that the bar in Sections 18 and 18A is not a bar where no offence is made out at all. The Court held that "if the complaint does not make out a prima facie case for applicability of the provisions of the Act of 1989, the bar created by Sections 18 and 18A(i) shall not apply," and the court retains power to grant anticipatory bail in such exceptional cases. Where a plain reading of the FIR shows that no offence under the Act is disclosed, the High Court may exercise its inherent jurisdiction under Section 482 CrPC to quash, and anticipatory bail is not foreclosed.

Justice S. Ravindra Bhat, concurring, located this carve-out in the duty of constitutional courts to prevent a miscarriage of justice in "truly exceptional" cases, balancing the imperative of curbing misuse against the protective object of the statute. The net result: the absolute bar stands as a general rule, but a complaint that discloses no atrocity at all cannot be used to defeat the liberty guarantee of Article 21. This narrow exception is the operative law on anticipatory bail under the Act today.

Reading the offence narrowly: Hitesh Verma and 'within public view'

The same protective-but-disciplined approach surfaces in the substantive offence cases. In Hitesh Verma v. State of Uttarakhand, (2020) 10 SCC 710, decided on 5 November 2020, the Supreme Court interpreted Section 3(1)(r) of the Act — intentional insult or intimidation with intent to humiliate a member of a Scheduled Caste or Scheduled Tribe "in any place within public view." The Court held that abuse uttered within the four walls of a building, in the absence of members of the public, is not an offence under that clause, because the statutory ingredient "within public view" is not satisfied.

Equally important, the Bench held that the offence is not made out merely because the victim belongs to a Scheduled Caste; there must be an intention to humiliate the victim for the reason that he belongs to such caste. Where a property dispute, rather than caste animus, animates the abuse, Section 3(1)(r) is not attracted. Hitesh Verma thus disciplines the substantive reach of the Act much as Prithvi Raj Chauhan disciplines its procedure — both insist on a genuine prima facie atrocity. These ingredient-level requirements connect directly to the definitions of atrocity, SC and ST.

The position today: Kiran v. Rajkumar Jivraj Jain (2025)

The settled state of the law after the trilogy was reaffirmed by a three-Judge Bench in Kiran v. Rajkumar Jivraj Jain, 2025 INSC 1067, decided on 1 September 2025. A Scheduled Caste voter alleged that, after he voted against the accused's preferred candidate in an Assembly election, the accused assaulted him with an iron rod, threatened to burn his house, and abused him and his female relatives with caste slurs in public view. The Bombay High Court had granted anticipatory bail; the Supreme Court set that order aside.

The Court held that Section 18 creates an absolute bar against anticipatory bail once specific accusations disclosing a prima facie offence under Section 3 are made, and that the only exception is the Prithvi Raj Chauhan rule — that the bar does not apply where the FIR, read as a whole, discloses no offence under the Act at all. On the facts, the caste slurs uttered in public view, the assault, and the threats together satisfied the ingredients of Section 3, including the election-retaliation limb in Section 3(1)(o), so the High Court's grant of pre-arrest bail was a "manifest error" and a jurisdictional illegality. Kiran thus confirms three things for the exam: the Balothia–Prithvi Raj Chauhan classification stands; the anticipatory-bail exception is genuinely narrow and FIR-based; and the Hitesh Verma requirements of "public view" and caste-based intent feed directly into whether the bar applies. It is the most current authority to cite on Section 18.

Doctrinal threads tying the trilogy together

Three themes emerge. First, separation of powers: Mahajan shows the danger of a court rewriting procedure under the guise of interpretation, and the 2019 review and Prithvi Raj Chauhan together restore the legislature's primacy over the procedural design of a special statute. Second, protective discrimination: the Article 15(4) and Article 17 foundations of the Act justify a procedure stricter than the ordinary criminal process, and the Court will not presume bad faith on the part of SC/ST complainants as a class.

Third, the irreducible Article 21 minimum: even a non-obstante bar on anticipatory bail cannot be used where the FIR discloses no offence at all. The prima-facie filter — apply your mind to the complaint, and if it is not an atrocity, the bar does not bite — is the constitutional bridge that lets a stringent statute survive Articles 14 and 21. That filter, rather than any administrative pre-approval, is the safeguard the law now recognises.

A fourth thread, often missed, is the relationship between the FIR and the courts' inherent jurisdiction. Because the Act forecloses anticipatory bail and pre-arrest enquiry, the only meaningful check before arrest is a Section 482 CrPC challenge to the FIR itself. Prithvi Raj Chauhan and Kiran both route the accused's relief through that jurisdiction: the High Court reads the complaint, and only if it discloses no atrocity may it quash and, consequentially, entertain anticipatory bail. This places a premium on careful pleading of the Section 3 ingredients in the FIR and on the Hitesh Verma tests — "within public view" and caste-based intent — because those ingredients are now the gateway not just to conviction but to the very availability of pre-arrest liberty. The procedural and substantive limbs of the Act have, in effect, merged at the threshold stage.

Exam takeaways and how to cite the trilogy

For judiciary and CLAT-PG answers, present the trilogy as a sequence. Open with Subhash Kashinath Mahajan v. State of Maharashtra, (2018) 6 SCC 454, and list the five directions (no absolute bar to anticipatory bail; DSP preliminary enquiry; pre-arrest approval for public and non-public servants; magisterial scrutiny). Then state that Union of India v. State of Maharashtra, (2020) 4 SCC 761 (1 October 2019), recalled those directions in review as impermissible judicial legislation contrary to Article 15(4).

Next, note the legislative override: the 2018 Amendment Act (Act 27 of 2018) inserted Section 18A, abolishing the preliminary enquiry and approval requirements and restating the bar on Section 438 "notwithstanding any judgment of any Court." Close with Prithvi Raj Chauhan v. Union of India, (2020) 4 SCC 727, upholding Section 18A while holding that the bar does not apply where the complaint discloses no prima facie offence, with the court retaining power under Section 482 CrPC and to grant anticipatory bail in those exceptional cases. Reinforce the substantive limb with Hitesh Verma v. State of Uttarakhand, (2020) 10 SCC 710, on "within public view" and the requirement of caste-based intent under Section 3.