The Forgotten Hindus: 18,000 Rapes and a Declining Population

In 1941, Hindus made up 28 percent of the population of what would become Bangladesh. By 2011, that number had fallen to 8.5 percent. Between those two dates, roughly 49 million people — or their ancestors — left.

Some went because of the 1947 Partition. Some fled the 1971 genocide. But millions left — or were driven out — because of something slower, less dramatic, and almost entirely absent from the international conversation: the systematic, recurring targeting of Bangladesh’s Hindu minority as a political instrument, most devastatingly in the weeks following the October 2001 general election.

What happened in the autumn of 2001 has a name. It has a number: 18,000 rapes, documented by a judicial inquiry commission. It has 25 names: the ministers and members of parliament from the BNP-Jamaat alliance identified as orchestrators. It has a verdict — rejected almost immediately by the party that won that election and still claims to be a democratic force.

What it does not have, more than two decades later, is accountability.


The Pattern Nobody Admits

Bangladesh’s minority population has been in managed decline for decades. The causes are not mysterious. They are documented, sourced, and largely uncontested in the academic literature. What is contested — bitterly, politically — is who is responsible.

The decline follows a clear pattern: it accelerates in the aftermath of elections where the BNP-Jamaat coalition either wins or loses badly. The logic is brutal in its simplicity. Hindus in Bangladesh vote predominantly for the Awami League. In a country where political violence is a tool of governance, minorities who vote the wrong way get punished. Not randomly. Systematically.

The 1992 Babri Mosque demolition in India triggered riots in Bangladesh that drove thousands of Hindus out of their homes. In 1946, the Noakhali pogrom killed hundreds. After independence in 1971, the Vested Property Act — inherited from Pakistan and modified — allowed the state to seize property from Hindus who had fled or been driven out, creating a legal framework for ethnic dispossession that remained on the books until 2001. By then, the machinery was already in place.

All it needed was a trigger.


October 2001: The Trigger

On October 1, 2001, the BNP-led Four-Party Alliance won Bangladesh’s general election by a landslide. It was a democratic outcome. What followed was not democratic in any sense of the word.

Before the polls even closed, the violence had started. Amnesty International documented it precisely:

“The current wave of attacks against the Hindu community in Bangladesh began before the general elections of 1 October 2001 when Hindus were reportedly threatened by members of the BNP-led alliance not to vote.”

— Amnesty International, “Bangladesh: Attacks on members of the Hindu minority,” AI Index ASA 13/006/2001, December 2001

The US State Department’s International Religious Freedom Report 2002 confirmed the timeline:

“According to a human rights organization, at least 10 Hindu women were raped and a number of Hindu homes were looted by low-level BNP workers a few days before the BNP took power from the non-partisan caretaker government.”

— US State Department, International Religious Freedom Report 2002

The violence intensified after the results were declared. Across more than 20 districts, concentrated in southwestern Bangladesh where Hindu populations were largest, a campaign of terror unfolded that lasted for months. The pattern was consistent: BNP and Jamaat supporters descended on Hindu neighborhoods. Homes were looted, then burned. Temples were desecrated. Women were raped. Families were told to leave — or worse.

This was not a spontaneous eruption of communal hatred. This was organized. The Fair Election Monitoring Alliance, citing field documentation from across Bangladesh, stated plainly: “Most of the violence was committed by BNP activists.”


What the Numbers Actually Mean

It is worth pausing on the number 18,000.

In most countries, a single rape case dominates news cycles for weeks. Eighteen thousand is a number that strains comprehension. It is not one crime or one neighborhood. It is a campaign. It is what happens when political actors with state power decide that an entire community’s bodies can be used as a message.

This figure did not come from an opposition party or an international NGO with an agenda. It came from a judicial inquiry commission ordered by the Bangladesh High Court in 2009 and reporting in 2011. The commission was composed of Bangladeshi judges examining Bangladeshi evidence. Its findings:

“The commission reported that the number of rapes committed exceeded 18 thousand. The report also notes incidents of violence, arson, looting, and torture against the minority Hindu community of Bangladesh.”

— bdnews24.com, April 24, 2011, reporting on the judicial inquiry commission findings

Beyond the rape figures, the commission documented the broader scope:

  • 25 Ministers and Members of Parliament from the BNP-Jamaat alliance identified as having orchestrated the violence
  • 25,000 individuals participated in the targeted attacks
  • Over 10,000 cases of human rights abuses against minorities catalogued (Hindu American Foundation submission to OHCHR)
  • 200+ Hindu women confirmed raped in the immediate post-election period (OHCHR document, Bangladesh Minority Council submission)

The BNP’s response to these findings? They rejected them entirely. “Partisan.” This is the party that asked Bangladesh to make it the government again in 2026.


The Bhola Numbers

Among the worst-documented cases was Bhola District, in the Barisal Division of southern Bangladesh.

On November 16, 2001, The Daily Star — Bangladesh’s leading English-language newspaper — reported that approximately 600 Hindu women had been gang-raped by BNP members in Char Fasson Upazila alone. The youngest victim was 8 years old. The oldest was 70 years old.

Read that again. An 8-year-old child. A 70-year-old grandmother. The violence did not discriminate by age. It was not about desire or individual depravity. It was about terror — about sending a message to an entire community about what their presence in Bangladesh would cost them.

In Lalmohan Upazila, also in Bhola, the documentation is similar. BNP supporters attacked Hindu families and Awami League supporters alike. They looted homes. They cut down trees on victims’ properties — a deliberate economic act, destroying years of agricultural investment. They stripped homes of every item of value. Families were left with nothing.

In Jessore District’s Tuniaghara, six Hindu families were forced to flee their homes permanently. Two women were raped.

Across the affected districts — Barisal, Jessore, Bagerhat, Khulna, Satkhira, Pirojpur, Bogra, Brahmanbaria, Chittagong, Feni, and a dozen more — the same pattern repeated. Attack the Hindu. Loot the home. Destroy the temple. Rape the women. Make them leave.


The Woman Who Dared to Show Up

Among the thousands of victims, one case became the most documented, because it eventually reached a court.

Purnima Rani Shil was a polling agent for the Awami League in Sirajganj District on October 1, 2001. She did what citizens in a democracy are supposed to do: she showed up, she served, she exercised her right to participate. Her reward was gang rape by BNP supporters who knew exactly who she was and why she was there.

It took a decade to get a conviction. In 2011, a court in Sirajganj sentenced 11 individuals to lifetime imprisonment for the crime (BBC News, May 4, 2011). It is one of the very few convictions that emerged from the entire 2001 post-election violence cycle.

The Devpolicy Blog at the Australian National University’s Development Policy Centre described her case as having sent “shockwaves” precisely because it was so documented, so deliberate, and so representative: a Hindu woman punished for participating in Bangladesh’s democracy.

What Purnima Rani’s case also revealed is the gap between exception and rule. She was one of at minimum 18,000 rape victims. Eleven perpetrators went to prison. The architects — the 25 identified ministers and MPs — faced nothing.


The Exodus Behind the Numbers

What does it look like when 18,000 women are raped, 25,000 people attack your community, and the government’s response is to reject the judicial findings?

You leave.

Amnesty International reported in December 2001 that “hundreds” of Hindus had already crossed the border into India in the immediate aftermath of the election. Gulf News confirmed continued flight in February 2002. These were not economic migrants. They were people who had concluded, rationally, that Bangladesh had made its position clear.

The demographic data tells a longer story. Bangladesh’s Hindu population:

  • 1941: 28% of total population
  • 1951: 22% (post-Partition flight)
  • 1961: 18.5%
  • 1971: Dropped sharply during the Liberation War genocide
  • 1974: 13.5%
  • 1981: 12.1%
  • 1991: 10.5%
  • 2001: 9.2%
  • 2011: 8.5%

Each decade represents hundreds of thousands of people who decided they could not stay. Each spike in emigration correlates with a period of political violence or targeted persecution. The 1971 genocide accounts for the largest single drop. The post-2001 period shows continued decline. The question demographers ask is not whether the decline is real — it clearly is — but whether it will stop.

The legal framework that facilitated dispossession — the Vested (Enemy) Property Act, inherited from Pakistani law — allowed the state to classify properties of Hindus who had fled as “enemy property” and transfer them to government hands or to political allies. The Awami League government formally repealed it in 2001 and passed the Vested Properties Return Act. Implementation remained slow and contested. By the time the BNP came to power in October 2001, the tool was nominally off the books — but the culture it had created, of treating Hindu property as fair game, was not.


What International Bodies Documented

The 2001 post-election violence is not a matter of contested history. It is documented by every significant international human rights body that examined Bangladesh in this period.

Amnesty International published a dedicated report within two months of the election. “The current wave of attacks against the Hindu community in Bangladesh” — they used that phrase in December 2001, when the violence was still ongoing.

Human Rights Watch documented BNP-Jamaat alliance supporters carrying out systematic attacks against minorities and Awami League supporters in the weeks following the election.

The UNHCR-affiliated Refworld database contains multiple documentation entries from this period. The Hindu American Foundation submitted formal documentation to the OHCHR. The United States Commission on International Religious Freedom tracked Bangladesh’s treatment of minorities across this period, including specific references to the 2001 violence in subsequent annual reports.

Fair Observer, writing in February 2026, summarized the academic consensus plainly: “Supporters and leaders of the BNP-led coalition and its Jamaat allies [were linked] with targeted violence against religious minorities, including killings, rape, arson and looting.”

The documentation is not sparse. It is voluminous. The problem is not evidence. The problem is consequences.


Twenty-Five People Who Were Never Tried

The judicial inquiry commission that reported in 2011 identified 25 Ministers and Members of Parliament from the BNP-Jamaat alliance government as having orchestrated the post-election violence.

Twenty-five sitting or former members of government. Named. In a judicial report. Submitted to the Bangladesh High Court.

How many were prosecuted? How many went to prison? How many faced any formal legal process for having organized the gang rape of 18,000 women?

The answer is effectively zero.

This is not because the evidence is thin. It is because the political will was absent. The BNP dismissed the commission’s findings as partisan. The Awami League government that commissioned the report did not aggressively pursue criminal cases. The international community, which had expressed concern in 2001 and 2002, had largely moved on by 2011.

And so the 25 orchestrators of the most extensively documented campaign of sexual violence in Bangladesh’s post-independence history walked free. Some remained active in politics. Some continued to serve as party officials. The BNP, when it returned to power in 2026, had people in its ranks who were implicated in the 2001 commission findings.


Why This Is Not Just History

Bangladesh Untold covers the 2001-2007 period with a specific purpose: to document, in sourced detail, what this country experienced under BNP-Jamaat rule so that the record cannot be rewritten by those who benefited from it.

The Hindu exodus question sits at the center of that record. Not as a footnote to an election. Not as a regrettable episode of communal tension. But as a policy outcome — the result of deliberate decisions by people with names, titles, and documented histories of violence.

When you read that Bangladesh’s Hindu population has declined from 28% to 8.5% over eight decades, you are reading the cumulative outcome of those decisions. Each percentage point represents families who concluded that the country of their birth — in some cases their families’ home for generations — would not protect them. That the state, or forces operating with state permission, would use them as targets whenever the political calendar required it.

The 2001 violence did not happen in a vacuum. It happened in a country where the Vested Property Act had been legalizing dispossession for decades. Where police were ordered to stand down during attacks, or actively participated. Where ministers sat in cabinet and organized attacks, then watched judicial inquiry commissions name them and faced no consequences.

It happened in a Bangladesh where the state had a long practice of manufacturing impunity for its own crimes — the same impunity documented throughout this site, in the grenade attack that killed 24, in the arms haul, in the RAB’s 600 “crossfire” deaths.

The pattern is consistent. The victims change. The architecture of protection for perpetrators does not.


The Declining Census

By the 2022 census, Bangladesh’s Hindu population had fallen further — to approximately 7.95 percent, the first time since independence that the figure had dropped below 8 percent.

Demographers note that even this number likely understates the departure of younger generations: the emigration is disproportionately of working-age Hindus who have the means and education to leave, leaving behind an older, poorer community that cannot. What this means for the next generation’s numbers is not hard to calculate.

In absolute terms, Bangladesh has one of the largest Hindu populations in the world — approximately 13 to 14 million people. But it is a community living with the knowledge of what has been done to it, and with the awareness that the people who did it have largely never been held accountable.

The judicial inquiry commission’s 2011 report sits in the Bangladesh court system. The 18,000 figure sits in its pages. The 25 names sit there too.

Bangladesh Untold exists, in part, to make sure those pages are not forgotten.


A Note on Sources

All figures in this article come from primary or verifiable secondary sources: the Bangladesh judicial inquiry commission’s 2011 findings as reported by bdnews24.com; Amnesty International’s December 2001 report (AI Index ASA 13/006/2001); the US State Department’s International Religious Freedom Report 2002; Human Rights Watch documentation from 2001-2002; the OHCHR document citing the Bangladesh Minority Council’s submission; the Fair Election Monitoring Alliance; academic demographic analyses of Bangladesh’s census data published by the International Journal of Population Studies and Devpolicy Blog (ANU); and The Daily Star‘s contemporaneous reporting from November 2001.

The 18,000 figure is from the judicial commission report. The demographic percentages are from census data. The names of identified orchestrators come from the commission’s findings. We have not extrapolated or estimated. This is what the record shows.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *