Discrimination Ends; Segregation Remains: Habitual Offenders in Indian Prisons

By Nandinii Tandon, Rajiv Gandhi National University of Law, BA LLB (Business Law Hons.) Class of 2027 and Mehul Sharma, Rajiv Gandhi National University of Law, BA LLB (Hons.) Class of 2028.

Introduction

When the helm of Bentham’s Panopticon rests in the hands of prison officials, the opacity of power ensures that the realities of prison life remain trapped within four walls. Inside these walls, the Indian Constitution’s protections often fall short, and discrimination behind bars takes its place. It is precisely this hidden architecture of power that the Indian Supreme Court (SC) sought to confront in Sukanya Shantha v. Union of India (Sukanya Shantha), where it struck down provisions in several state prison manuals that sanctioned caste-based segregation. The historic judgment was delivered on October 3, 2024 by a Bench comprising of Chief Justice D.Y. Chandrachud, Justice J.B. Pardiwala, and Justice Manoj Misra. In the wake of such reforms as recommended by the SC and the Ministry of Home Affairs (MHA), state governments, such as Tamil Nadu and Jammu & Kashmir, have begun amending their respective prison manuals to abolish caste-based segregation. In this article, we examine how segregation has been justified in prisons, the continued risks posed by habitual offender laws, and why a stricter constitutional standard is needed to prevent such practices from resurfacing. We further argue that Indian courts must adopt a stricter standard of review, akin to the strict scrutiny test, to address disguised discrimination in prisons.

Understanding Segregation

The most fundamental aspect of prisons is the enforcement of institutional rules, with security and discipline serving as the cornerstone of these systems. They function not only as mechanisms of control but also as tools to maintain order and safety within the prison walls, generally imposing consequences beyond incarceration. However, security and disciplinary measures are frequently implemented based on discretionary and ambiguous criteria rather than definitive evidence of their requirement. Before examining segregation as a tool of prison administration, it is necessary to understand what segregation entails in the Indian context. Segregation in prisons refers to the systematic separation of inmates into different physical spaces, such as barracks, enclosures, or cell blocks, based on specific classifications. While such classifications are often justified on administrative or security grounds, in practice, they have frequently mirrored entrenched social hierarchies, particularly caste.

The Indian Supreme Court in Sukanya Shantha recognized and struck down the persistent caste-based practices within Indian prisons, as reflected in several State prison manuals and longstanding institutional arrangements. In particular, it noted that segregation on the basis of caste has been affected through the allocation of separate barracks or wards for inmates belonging to different caste groups. This separation is not merely spatial. It extends to the distribution of labor, where prisoners from marginalized castes are disproportionately assigned tasks considered “menial” or “polluting,” while others are exempted from such work. In this way, caste-based ideas of purity and pollution continue to operate within prisons.

Segregation has been an apparatus used by prisons across jurisdictions to make prisons disciplined and perform other security-related actions. When examining various countries, the underlying reasons for segregation executed by the respective prison officials have been diverse and multifaceted. These reasons include, for example, the imposition of administrative segregation on an inmate as a result of his “suspected” involvement with a Black prison gang in the United States (Richardson v. Runnels), the maintenance of good order and discipline with a pending investigation into a serious assault to one other inmate in the United Kingdom (R (Bourgass) v. Secretary of State for Justice), because of incident reports indicating personal drug use and stand-over tactics against other inmates in New Zealand (Bennett v. Superintendent of Rimutaka Prison), and because the offender cannot be permitted to be in contact with mainstream inmates in Australia (R v. Mohammed Fahda).

At this juncture, it becomes pertinent to address whether segregation within prisons is a legitimate exercise of authority. To examine this, the case of Washington v. Lee (Lee) assumes relevance. The case arose from the race-based segregation practices in the jails, juvenile jails, and prisons of the state of Alabama in the United States, wherein white prisoners were housed separately from African American prisoners. In 1966, a suit was filed against Alabama prison officials, seeking that the prisons be desegregated. The U.S. District Court for the Middle District of Alabama held that this complete and permanent racial segregation within penal facilities, under the garb of prison security or discipline, was unconstitutional. The District Court explicitly recognized, however, that “in some isolated instances prison security and discipline necessitates segregation of the races for a limited period.” It is pertinent to note that any statutes or practices which require or permit prison officials to segregate the races arbitrarily will collapse under the weight of American constitutional principles. The District Court further ordered Alabama to desegregate all prisons and jails.

Alabama subsequently appealed to the U.S. Supreme Court, where the apex court affirmed the lower court’s ruling prohibiting racial segregation in public prisons. Following a brief per curiam decision, Justices Black, Harlan, and Stewart issued a concurring opinion. They observed that “prison authorities have the right, acting in good faith and in particularized circumstances, to take into account racial tensions in maintaining security, discipline, and good order in prisons and jails.” Particularly, this approach maintains a balance by recognizing the legitimacy of limited, temporary measures only where genuinely necessary for maintaining prison security and order, while rejecting systemic racial segregation and preventing its misuse under the guise of discipline.

Habitual Offenders: Law, Label, and Prejudice

India’s social structure has historically been shaped by the caste system, a hereditary and hierarchical form of social stratification. Broadly, caste divides society into four primary groups, namely, the Brahmins (priests and scholars), Kshatriyas (warriors), Vaishyas (traders), and Shudras (laborers), with communities outside this framework, often referred to as Dalits, subjected to the most severe forms of exclusion. Caste status is determined by birth and has traditionally governed access to resources, occupations, and social mobility. Communities at the lower end of this hierarchy have faced systemic discrimination, including social segregation, economic marginalization, and criminal stereotyping. It was within this context that colonial laws such as the Criminal Tribes Act, 1871 (CTA) was enacted. The Act institutionalized pre-existing caste-based prejudices by converting them into state policy, enabling the British to categorize certain marginalized communities as “hereditary criminals” and to subject them to increased surveillance and control. Post-independence, the CTA was repealed and the “criminal tribes” were officially “denotified,” signifying that the state could no longer brand them as “hereditary criminals.” In a regressive move, many states enacted their respective Habitual Offender Acts in 1952. Although these laws did not explicitly target the “criminal tribes,” they “only re-stigmatised the marginalised tribes,” as the old prejudices were effectively transferred onto the same populations once again. This continuity is not merely structural but also linguistic: the very label “habitual offender” carries forward colonial associations of criminality tied to identity, rather than conduct, and should be replaced with terminology that focuses on individual actions. In fact, it was recognized by the court in Sukanya Shantha that these acts have been used and continue to be used to target members of the denotified tribes, especially in prisons.

India does not have a single, central legislative regime governing the administration and management of prisons. Under List II, Entry 4 of the Seventh Schedule to the Constitution, prisons are a state subject, which means their regulation is carried out through state-specific prison laws and manuals. Most of these, however, continue to trace their foundations to the colonial-era Prisons Act, 1894, which is to be replaced by the Model Prisons and Correctional Services Act, 2023 (the Model Act). In 2016, the MHA also issued the Model Prisons Manual, 2016 (the Model Manual) aimed at bringing uniformity in the laws governing prisons all over the country. Yet, despite these attempts at uniformity, prison governance in India has often reflected deep-seated social prejudices. To illustrate, Rule 411(iv) of the Madhya Pradesh Jail Manual, 1987 states that any member of a denotified tribe can fall under the meaning of habitual criminals, subject to the discretion of the state government. Such misleading and overarching definitions of habitual offenders have been declared unconstitutional in Sukanya Shantha. Following this, the MHA amended the Model Manual and the Model Act in December 2024 to standardize the definition of a “habitual offender” as a person convicted and sentenced to imprisonment more than twice within a continuous five-year period. This move is welcome, as it removes the ambiguities and inconsistencies that had persisted across various state versions and had often been used to disproportionately target members of denotified communities. On the basis of this improved and amended definition, segregation of “repeat offenders,” i.e., habitual offenders, is legitimized, as the law now places emphasis on the criminal conduct by individuals rather than stigmatizing entire communities. However, this segregation cannot become a pretext for arbitrary action: its legitimacy depends upon a clear rational nexus between the classification and its objective, namely, the demonstrable pattern of repeated criminal conduct by the same individual, established through prior convictions within a defined period. It is this shift from identity-based suspicion to conduct-based classification that sustains the constitutional validity of such segregation; absent this, the amendment’s corrective intent would stand diluted.

Segregation, though widely debated across jurisdictions, has also found its way into Indian prisons, with prison authorities similarly invoking it as a means of maintaining order, often in ways that reflect entrenched social hierarchies. To illustrate, in 2014, the Madras High Court in C. Arul v. The Secretary to Government (C. Arul) upheld a caste-based segregation of barracks in order to prevent caste rivalries. The reasoning given by the Superintendent of Prisons, Palayamkottai Central Prison, TN was that “in order to avoid any untoward incident and put an end to such rivalry, the Prison Authority is compelled to house the inmates of different communities in different blocks, which is not actually on caste basis.” This reasoning bears similarity to the defense in Lee, where segregation was also sought to be justified on grounds of prison security and discipline. The SC in Sukanya Shantha refuted the reasoning in C. Arul, holding that the prison administration is responsible for maintaining discipline and order without resorting to such extreme actions that perpetuate caste-based segregation. The Court further noted that, “The prison authorities ought to be able to tackle perceived threats to discipline by means that are not rights-effacing and inherently discriminatory.”

Following the judgment in Sukanya Shantha, caste-based discrimination and segregation has been explicitly and completely prohibited by the amendments made to the Model Manual and the Model Act. Further, the definition of “habitual offenders” has also been amended to ensure that the members of denotified communities are not unjustly targeted. It is important to note that the segregation of habitual offenders, in accordance with the amended definition, does not stand prohibited. As seen above, it is clear that segregation cannot be justified under the mere pretext of security and discipline and is permissible only in limited situations where prison security genuinely and temporarily requires it, provided such action is non-arbitrary and has a rational nexus to the objective, as advanced in Lee.

However, a precarious situation may arise where such segregation may continue to be used to perpetuate arbitrary and discriminatory segregation against denotified communities. Section 67 of the Model Act provides immunity from legal proceedings for the Government or any functionary regarding actions taken in “good faith” while executing their duties under the Model Act. Since good faith is not defined in the statute (and is inherently difficult to define exhaustively) this vague requirement creates scope for potential misuse by prison authorities. To illustrate, consider a situation where a prison official segregates fifty inmates on the ground that they are “habitual offenders,” but forty-five of them are found to belong to a particular denotified tribe. This concern finds resonance in the doctrine of disparate impact, developed in jurisdictions such as the United States, where facially neutral policies may be challenged if they disproportionately affect a particular group. However, as recognized in cases such as Washington v. Davis, disproportionate impact alone is insufficient to establish a constitutional violation; discriminatory intent must also be demonstrated. This makes such claims inherently complex, as intent is often difficult to prove or disprove, particularly where decision-making is shielded by broad statutory defenses such as “good faith.” In this light, allowing prison officials to invoke Section 67 without rigorous scrutiny risks shielding potential patterns of discrimination from judicial review, even where the outcomes strongly suggest otherwise. So, where does the solution lie?

In 2005, the U.S. Supreme Court in the case of Johnson v. California (Johnson) announced a strict scrutiny standard for racial classifications in prisons, requiring the government to show that such measures are narrowly tailored to serve a compelling governmental interest. The Court dismissed the contention that “equal segregation” of all prisoners regardless of their race rendered the policy immune from strict scrutiny. In other words, racial classifications attract strict scrutiny even if they appear to impact all races equally. Alternatively, in the Indian context, such a strict-scrutiny standard should be applied to ensure that denotified tribes (or any community) are not segregated under the guise of being habitual offenders. If such segregation is undertaken, prison authorities or the government must demonstrate both the existence of a compelling governmental interest (as per the strict-scrutiny test) and that the action was taken in good faith (as per Section 67 of the Model Act). Notably, the SC in Subhash Chandra v. DSSSB recognized that strict scrutiny may apply where a statute or executive action is patently arbitrary, violates core rights, or curtails life and liberty. Segregation, based on community identity, would squarely fall within these conditions, being manifestly arbitrary and directly implicating Articles 14 (equality before law) and 21 (protection of life and personal liberty) of the Indian Constitution.

Conclusion

The trajectory from Lee to Johnson in the United States, and more recently to Sukanya Shantha in India, makes one thing clear: segregation, when grounded in prejudice and disguised as discipline, violates the very foundations of constitutional principles such as equal protection and due process. In the context of prisons, classification can only survive if it is functional, i.e., temporary, narrowly tailored, and demonstrably necessary. As Sukanya Shantha reminds us, constitutional guarantees do not bend at the prison gates. But this promise requires Indian courts to adopt a stricter standard, such as strict scrutiny, when reviewing prison classifications. The amendments to the Model Act and Model Manual are applauded, but their promise will remain fragile unless they are accompanied by linguistic and structural reforms. Nomenclature matters: to describe someone as a habitual offender is to summon a history of colonial stigma against denotified tribes. By contrast, using the term repeat offender could be a better alternative as it shifts the focus to conduct rather than identity, caste, or community. As argued earlier, this term risks tying criminality to identity instead of actions. If prisons are allowed to legitimize prejudice under the guise of security, discipline, and good faith, they hand prison authorities the gavel to halt time itself, just as the clock was beginning to strike justice after decades of silence.