The Paradox of Protection: Dr. B. R. Ambedkar, Gandhi, and the Limits of Safeguards

Share

By Guneet Talwar —

 August, 2023: a padlocked house with blood splattered across the porch, bloody handprints staining the walls, and the raw, palpable stench of blood still lingering. That was the fate of Chinnadurai, a student in class 11 who had been assaulted. He was a straight-A student who, even at the time of his assault, was carrying test papers with perfect scores in his school bag. That was found unacceptable by his dominant-caste classmates, as Chinnadurai belonged to a marginalised community. He was attacked on the basis of caste stigmatisation.1

I contend that the constitutional and legal safeguards that are meant to protect against possible dangers, such as the Poona Pact and the SC/ST Act, have invariably failed to close the gap between constitutional protection and lived reality for the marginalised. The same gap is now repeating itself in the form of digital caste-based hate crimes, confirming Dr Ambedkar's original scepticism of safeguards that were granted rather than being managed by the oppressed.
First and foremost, this article will examine the history of the safeguards debate between Ambedkar and Gandhi, then it will analyse the records of legal enforcement, before exposing the same patterns or systematic failures that have invariably spread into digital spaces.

Historical debate

The safeguard debate started in 1930, at a Round Table Conference in London which consisted of 90 delegates. Ambedkar was invited to represent Dalits, so he did. Then in 1931, the second Round Table Conference took place. Gandhi attended the second conference as the sole representative of the Congress political party. He argued that Dalits who were represented by Dr Ambedkar, a Dalit himself, came under the fold of Hindus, which Gandhi and Congress represented. However, Ambedkar disagreed, saying Dalits were “a part apart from Hindus”, hence why they needed safeguards. Following this, the Ramsay MacDonald Award was announced. It gave separate electoral rights to Dalits. After the award, Gandhi went on a hunger strike. Ambedkar knew that if Gandhi died due to the hunger strike, the hatred and mistreatment against the Dalits would increase exponentially, resulting in killings and riots. Therefore, he complied with Gandhi and signed the Poona Pact, which gave Dalits reserved seats and formal safeguards.[2] Gandhi's supporters argued that the pact was a compromise that sought to preserve national unity, while still expanding Dalit representation through reserved seats, as they believed separate electorates would divide Hindus and Dalits politically. A week after, Gandhi formed an organisation which came to be known as Harijan Sevak Sangh; it was made to focus on gaining civil rights for the Dalits by the Dalits. However, it became a paternalistic organisation run by the upper-caste Hindus working for the "upliftment" of Dalits that flowed from Gandhi's theory of caste, which portrayed caste as a sin of Hinduism. Ambedkar heavily condemned the Sangh, as he stated that the work of the Sangh was the most inconsequential kind. It neglected the most important purposes for which Dalits needed help and assistance. For them, Dalits were no more than beggars, mere recipients of charity; Ambedkar firmly stood for the annihilation of caste as he believed that untouchability was a fundamental result of the caste system. He believed no alleviation, no uplift and no relief from untouchability without the abolition of caste; hence, even after signing the Poona Pact, he wrote that what Gandhi and the Congress did after Dalits got separate electorates was a "foul and filthy act". The fast was not for the benefit of Dalits but the worst form of coercion against a marginalised community to give up their safeguards.[2]

Legal enforcement

After the 1932 Poona Pact, another safeguard for marginalised communities was added in 1989: the SC/ST Act. In other words, it is a federal law enacted to protect individuals belonging to the Dalit and Scheduled Tribes communities from discrimination, humiliation, and hate crimes. It also grants the marginalised communities a set of protections such as no pre-arrest bail to prevent pressure on victims and a fast-tracked justice system, which mandates the creation of special courts to handle these cases quickly.[3] It has strengthened legal protections for the marginalised by criminalising caste-based atrocities, establishing special courts, and encouraging greater reporting rates of offences, giving many people better paths of justice. However, the law is far from perfect, and it requires strict reforms. India fails to protect its marginalised communities. When looking at conviction rates, the typical conviction rate for ordinary crime law is about 40 percent, while under the SC/ST Act it falls to 25-30 percent. The root cause of this is poor evidence collection, not using the right parts of the Act, and, in certain cases, watering down charges.[4] All of these make cases weaker, and the first step of filing an FIR (First Information Report) is extremely tedious, as victims from marginalised groups have to deal with threats or discouragement. Many state governments have also failed to set up any special courts, but instead they have given judges who are already busy with other criminal cases the job of handling atrocity cases as well.[4]

Furthermore, this law has had major inconsistencies. For example, in 1995, State of M.P. v. Ram Krishna Balothia, the Supreme Court affirmed the denial of anticipatory bail since communities from the SC/ST category are very vulnerable. However, in 2018, Subash Kashinath Mahajan v. State of Maharashtra, the courts established prerequisites for preliminary inquiry before FIR registration, expressing concerns over potential abuse. Later in the same year, a quick change was made by Parliament, which was later again reaffirmed in 2019 to bring back the Act's original provisions, showing that it was meant to protect people.[4]

Another major problem is the pattern of enforcement gaps; the Chinnadurai case mentioned in the introduction is an example of this. Chinnadurai, a Dalit student in class 11 in a government school, was attacked and maimed for calling out caste oppression by dominant caste schoolmates. Although disciplinary actions were taken against three upper-caste students that day, the same night, seven upper-caste classmates, three minors and four adults went to Chinnadurai's house and assaulted him.1 Chinnadurai's case is not the only story. In just one of the 28 states in India, Tamil Nadu, alone reported 65 caste-based honour killings in the span of eight years, with only eight getting favourable verdicts.[1]

This is how caste discrimination still prevails, although in books it is taught that untouchability is a sin and something of the past; many never see a teacher explain what it implies in real life, what it means in the right social and political context and what safeguards are there to protect these marginalised groups.¹ This is why the rights given to the marginalised are viewed negatively; reservation, a right given to marginalised communities after the Poona Pact, is viewed as a privilege which society believes that Dalits and scheduled tribes do not deserve since they have lower merit requirements.[1] This is viewed as unfair by upper caste groups, as it invokes a fear of equal equity and social inclusion for each citizen. However, reservation of marginalised communities has still not fulfilled its original purpose of giving equal equity to Dalits and scheduled tribes. These marginalised communities make up around 25% of India's population, yet they only make up roughly 19 percent of judges in district and subordinate courts, while representation in High Courts remains below 5 percent.8 In medical education, SC enrolment across medical courses stands at 13.42 percent and ST enrolment at 4.95 percent — both falling short of their 15 and 7.5 percent reserved quotas.9 In the civil services, SC and ST officers accounted for roughly 8 percent and 4 percent, respectively, of direct recruits to the IAS between 2020 and 2024.10 And across the wider central government workforce, SC and ST employees make up 16.84 and 8.7 percent of the total — figures that fall even further, to roughly 14.2 and 6.5 percent, once only senior posts are counted.11 However, on the other side of the spectrum, SC/ST communities are funneled into manual labour. Roughly 97 percent of India's identified manual scavengers who unsafely and manually handle or dispose of human excreta with basic tools like brooms and buckets12 and 70 percent of waste pickers13 belong to the SC/ST communities.

Digital spaces

Those same paternalistic roots of casteism have seeped into social media since India had approximately 398 million social media users aged 18 and over as of early 2023, equivalent to 40.2% of the country's population in that age group.[14] A majority of users are from dominant caste groups; therefore, social media has become a breeding ground for humiliating Dalits. New methods continue to be created to ridicule Dalits; a recent study showed that thirteen percent of hate posts on Facebook in India were related to caste-based hate speech.5 These hate posts are propagated by exclusive dominant-caste groups that exist to circulate hate speech. They serve their interests of reinforcing casteism in digital and social spheres. Frequently being associated with some of these groups gives advantages in offline spaces such as housing and jobs. A popular form of hate speech in these groups is using casteist slurs or memes to mock Dalits.[5] Despite all of this happening in India, it took Facebook almost 15 years to address caste-based hate speech in its community speech guidelines. Twitter took 13 years to include caste and 11 years after to officially recognise it. YouTube similarly implemented its 2019 hate speech policy, which prohibited content promoting violence or hatred based on various attributes, including caste.[5] This gave dominant caste groups ample time to plant their casteist roots and to let them flourish. A research article showed that Dalit history and content are often made invisible as platforms push biased algorithms, which influence upper-caste individuals into pushing casteist content, leading to a rise in anti-caste hatred. Compounding this, platform policies and reporting systems do not adequately protect Dalits from digital hate speech, nor do they allow their concerns to be meaningfully addressed. Furthermore, Dalit users are forced to face trolling, abuse and discrimination on social media, constantly leading Dalit users into silence or withdrawal.[6]

Conclusion

Nearly a century after Ambedkar first warned that paternalistic safeguards handed down by the powerful would never substitute for the power held by the marginalised themselves, his scepticism has only been proven to be relevant even to this day. The Poona Pact that was won through Gandhi's coercion and the repeating pattern it set still shows its effects today. In legal safeguards like the SC/ST Act, which exists on paper as federal protection, yet conviction rates, inconsistent case law, and FIR barriers show that legal enforcement remains hollow in practice. And when casteism moved into digital spaces, the same pattern repeated once more. Platforms took over a decade to even acknowledge caste-based hate speech, while algorithms bury Dalit voices instead of protecting them. Some may argue that the existence of legal and constitutional safeguards itself demonstrates the success of the social welfare policies when compared with the past. However, measuring standards against the past sets a benchmark far too low. The real measure is whether these safeguards have abolished the inequalities they were designed to address, and the evidence suggests that inequalities have remained deeply rooted. Across three different periods of time and three entirely different domains: government, law, and technology. The shape of the failure has stayed parallel. A safeguard is granted, then diluted and then ignored by the very structures meant to uphold it. This is precisely what Ambedkar foresaw when he rejected the idea that Dalits could rely on protections managed by others rather than governed by themselves. Constitutional and legal safeguards were never going to be enough to get rid of a hierarchy that is this deeply rooted, and 75 years of evidence has shown it. Until protection is built by and for the marginalised rather than granted to them, the gap between the promise of equality and the reality of caste will keep finding new spaces to survive in.

Works Cited


1 M. Palani Kumar, "Caste in the Classroom," People's Archive of Rural India (PARI), 2026, https://ruralindiaonline.org/article/caste-in-the-classroom.

2 Gail Omvedt, "Ambedkar Confronts Gandhi," Seeking Begumpura (blog), August 7, 2012, https://seekingbegumpura.wordpress.com/2012/08/07/ambedkar-confronts-gandhi/.

3 Government of India, Ministry of Social Justice and Empowerment, "The Scheduled Castes and Scheduled Tribes (Prevention of Atrocities) Act, 1989," https://socialjustice.gov.in (site restricts automated access; description of Act provisions cross-confirmed against secondary source in note 4).

4 Sandeep Madhukar Tamgadge and Pravina Khobragade, "A Critical Review of the SC/ST Act's Performance in Preventing Atrocities Against Marginalized Communities," ShodhKosh: Journal of Visual and Performing Arts 5, no. 7 (2024): 1740–1745, https://doi.org/10.29121/shodhkosh.v5.i7.2024.6419.

5 Social & Media Matters, "Caste-Hate Speech in the Age of Digital Society," National Campaign on Dalit Human Rights, April 2024, http://www.ncdhr.org.in/wp-content/uploads/2024/04/Caste-Based-Abuse-Report.pdf.

6 Pardeep Singh Attri, S. M. Ramya, Sarah Glozer, and Vivek Soundararajan, "' You Belong to Gutters, Not Facebook or Twitter': Recovering Dalit Histories From the Shadows of Social Media," Business & Society, February 9, 2026, https://doi.org/10.1177/00076503261415776.

7 Office of the Registrar General & Census Commissioner, India, Census of India 2011 (Scheduled Castes and Scheduled Tribes population figures).

8 Himanshu, "SC/ST and Women in the Judiciary: A Closer Look at Representation," Legal Bites, February 16, 2025, https://www.legalbites.in/legal-news/scst-and-women-in-the-judiciary-a-closer-look-at-representation-1112561.

9 "Not Enough SCs & STs Take Up MBBS & Dental Studies, But Exceed Their Quotas in Nursing," ThePrint, January 23, 2019, citing All India Survey on Higher Education (AISHE) data, https://theprint.in/india/governance/not-enough-scs-sts-take-up-mbbs-dental-studies-but-exceed-their-quotas-in-nursing/181056/.

10 Rajya Sabha written reply by Union Minister of State Dr. Jitendra Singh, February 12, 2026, reported in "Union Govt Withholds Caste-Wise Data on IAS, IPS, IFS, Shares Recruit Numbers," The News Minute, February 17, 2026, https://www.thenewsminute.com/news/union-govt-withholds-caste-wise-data-on-ias-ips-ifs-shares-recruit-numbers.

11 Department of Personnel and Training (DoPT), Annual Report 2024-25, Government of India, reported in "66% Group C Sanitation Workers Employed With Union Government Are From SC, ST, OBC Communities: Report," The Wire, January 29, 2026, https://m.thewire.in/topic/caste/66-group-c-sanitation-workers-employed-with-union-government-are-from-sc-st-obc-communities-report.

12 Ministry of Social Justice and Empowerment, written reply to Rajya Sabha by Minister of State Ramdas Athawale, December 1, 2021, https://static.pib.gov.in/WriteReadData/specificdocs/documents/2022/may/doc202251656401.pdf.

13 Ministry of Social Justice and Empowerment, Waste Picker Enumeration data under the NAMASTE Scheme, reported in "60.3% of India's Waste Pickers Are Dalits," India Herald, February 2026, https://www.indiaherald.com/Breaking/Read/994880432/-of-Indias-Waste-Pickers-Are-Dalits-Upper-Castes-ZERO-Visibility-in-Official-Data.

14 "Digital 2023: India," DataReportal, February 2023, https://datareportal.com/reports/digital-2023-india.