On 19 January 2026, Muhammad Yunus posted on social media that his government had investigated 645 incidents involving minorities in Bangladesh during the previous year. Of those, he said, only 71 could be classified as communal in nature. The remaining 574, the government maintained, were ordinary criminal or social disputes — land conflicts, theft, sexual violence, unnatural deaths — bearing no meaningful relationship to the religion of the victims.
Monindra Kumar Nath, the Acting General Secretary of the Bangladesh Hindu Buddhist Christian Unity Council, read that post and described it, in public, as “absurd and misleading.” His organisation had spent the entire year of 2025 monitoring national newspapers and media outlets. They had counted 522 incidents of communal violence — not 71. Those 522 incidents had resulted in 61 murder cases producing at least 66 deaths, 28 cases of sexual violence including rape and gang rape, 95 attacks on places of worship, and 102 attacks on minority homes and businesses. There were also cases of abduction, torture, blasphemy-linked arrests, and systematic forcible occupation of land and commercial property.
Neither set of numbers can be right. What separates them is not just methodology but definition — specifically, a government definition that appears, as Nath put it, to exclude killings, rapes, arson attacks on homes, and land grabbing from the category of communal violence unless they occur inside a temple.
A Year Measured in Incidents
The Unity Council’s figures represent one of the most comprehensive records attempted of what Bangladesh’s minorities experienced in 2025. They are drawn from media reporting across the country, which means they are likely an undercount: incidents in areas without consistent press coverage, cases where victims chose not to speak publicly, and episodes that local police declined to record at all would not appear in newspaper columns.
A separate investigation by the Human Rights Congress for Bangladesh Minorities documented 116 minority deaths between 6 June 2025 and 5 January 2026, across all eight administrative divisions and at least 45 districts. This figure encompasses a broader range of deaths than homicides alone — it includes lynchings, suspected killings, and deaths under circumstances the organisation assessed as connected to the minority status of the victim. The two organisations’ figures are not directly comparable: they use different time periods, different methodologies, and different definitions of what counts as a minority-related death. What they share is the conclusion that this is not a local or isolated phenomenon but a nationwide pattern.
The geography alone makes that clear. The violence has been recorded in Satkhira and Natore, in Savar and Gopalganj, in Dinajpur, Noakhali, Mymensingh, Faridpur, Narsingdi, Rajbari, Rangpur, Chittagong, and Pirojpur. When the Unity Council asked the government how killings in all these districts could be dismissed as non-communal, it received no satisfactory answer.
Counting and Not Counting
In March 2026, India’s Ministry of External Affairs told the Rajya Sabha that approximately 3,100 incidents of violence against Hindus and other minorities had occurred in Bangladesh between August 2024 and February 2026, based on data from human rights organisations. That figure incorporates the 2,010 incidents the Unity Council documented in just the first fortnight after Sheikh Hasina’s ouster in August 2024, as well as the 522 communal attacks recorded in 2025. India has consistently raised these figures with Bangladesh at the highest diplomatic levels.
The Yunus government’s response to such figures has been consistent: the numbers are exaggerated, the incidents are misclassified, the violence is political rather than communal, and the narrative is being driven by those with an interest in destabilising Bangladesh or undermining its relationship with India. There is a genuine methodological debate to be had about how different organisations classify incidents and what threshold transforms a criminal act into a communal one. But the government’s position — that investigations found not a single one of 22 minority killings examined to be communal in nature — has struck rights organisations as straining credibility. When Yunus described reports of widespread minority violence as “baseless” at the United Nations in late 2025, the Unity Council issued a statement calling his remarks “a denial of truth.”
The Structure Beneath the Statistics
What both the Unity Council data and the HRCBM mapping describe is not a random distribution of crime. The elements recur: a mob assembles, often after a provocative social media post or a blasphemy allegation that may be fabricated; temples, businesses, and homes are targeted; sexual violence and public humiliation are used to terrorise entire communities rather than simply injure individuals; and land or commercial assets frequently change hands after families are displaced. The perpetrators calculate, generally correctly, that the cost of this violence is low.
That calculation is not made in a vacuum. It is informed by decades of experience in which communal attacks on minorities have been reclassified as routine crime, prosecutions have stalled, and survivors who pursue cases have faced pressure to withdraw them. Ain o Salish Kendra, one of Bangladesh’s most respected human rights organisations, documented 160 incidents targeting the Hindu community between January and September 2025 and stated in August that the interim government had failed to take effective measures against mob violence. Their year-end report described “mob terrorism” as a defining feature of 2025, with mob violence killing 197 people across Bangladesh over the course of the year — not all of them minorities, but all of them victims of a system in which extrajudicial violence had become normalised.
What the Numbers Are Really Saying
Bangladesh’s constitutional promise of equality for its religious minorities is not being kept. That conclusion follows from the data of multiple independent organisations using different methodologies across a full calendar year, and it is not seriously disputed by anyone who has examined the evidence closely. The question is what follows from it.
The government’s preferred framing — that most violence is criminal rather than communal, and that attributing it to religious hatred is a politically motivated distortion — serves primarily to avoid the harder conversation. If 574 out of 645 incidents involving minorities are simply crime, then no special response is required, no structural failure needs to be acknowledged, and no political accountability attaches to the pattern. If they are communal attacks — if the religion of the victim is the reason they were chosen — then something more serious is being said about Bangladesh’s direction of travel.
The figures suggest the latter. A country in which 95 places of worship were attacked in a single year, in which blasphemy accusations are a reliable mechanism for triggering mob violence against an entire community, and in which a rights organisation feels compelled to describe the demographic trajectory as pointing toward Bangladesh becoming a “Pakistan-style” state rather than the plural democracy its founders envisioned — that country has a problem that statistical reclassification cannot wish away. What it requires is honest accounting, followed by prosecution, followed by the kind of sustained political will that has, so far, been absent.