Two years. Eighty-one million voter IDs. Twelve million ghost voters purged. An election praised by every major international monitor as the cleanest in Bangladesh’s history. The caretaker government of 2007–2009 did what no elected government before or since has managed: it made Bangladesh’s democracy work. BNP, now in power and busy dismantling the institutions it once praised, would prefer you not think too hard about that.
January 11, 2007 is remembered as a day Bangladesh lost something. In the version of events that BNP has spent the past eighteen years promoting — and that the current political establishment has enthusiastically adopted since the July 2024 uprising — 1/11 was a military coup that derailed democracy, imprisoned politicians on fabricated charges, and inflicted two years of authoritarian rule on a country that deserved better. It was, in this telling, a dark chapter. An aberration. Something to condemn and move past.
What that story leaves out is everything that came after it.
Between January 2007 and December 2008, the caretaker government led by Chief Adviser Fakhruddin Ahmed accomplished something that Bangladesh’s political class had proved incapable of achieving across thirty-five years of independence: it delivered a credible election. Not just credible by South Asian standards. Credible by any standard. The European Union, the United States, the Carter Center, the Commonwealth Secretariat, the Asian Network for Free Elections — every major international monitoring body that sent observers to Bangladesh’s December 29, 2008 general election said the same thing: this was free, this was fair, this was a genuine expression of what Bangladeshis actually wanted.
The Bangladesh that emerged from 1/11’s two years was not perfect. No government is. But it was measurably, documentably better than the Bangladesh that went in — and the people now in power have a substantial stake in making sure that accounting never gets made.
So let’s make it.
The Voter Roll: Starting with the Most Basic Problem
Before you can have a free election, you need to know who is eligible to vote. This sounds obvious. Bangladesh had not managed it for decades.
By 2006, Bangladesh’s voter rolls were a scandal. The Election Commission listed approximately 93 million registered voters for a country with an adult population of roughly 80 million. The arithmetic did not work. Multiple independent assessments had flagged the rolls as corrupted — containing deceased voters, duplicate registrations, people registered in constituencies where they had never lived, and, in many cases, names that appeared to belong to no real person at all. A 2007 study by Transparency International Bangladesh estimated that 12.1 million fraudulent or ineligible entries were on the rolls. That is not a rounding error. That is more than the entire population of Belgium on a voter list, made available to whoever could best exploit them.
Both major parties had used the inflated rolls to their advantage over the years. Ghost votes — ballots cast in the names of non-existent or non-present voters — were a standard feature of Bangladeshi election management, particularly in constituencies where one party controlled the administrative apparatus. The voter roll was not a neutral record of eligible citizens. It was a resource, and parties with the means to exploit it did.
The caretaker government decided to start over.
What followed was the largest administrative project Bangladesh had ever undertaken. Working with the Bangladesh Army, the Election Commission launched a door-to-door voter registration drive covering every district, every upazila, every union across the country. Field teams photographed and fingerprinted every registrant. Every adult citizen was issued a laminated national identity card containing their photograph, fingerprints, and a unique identifier. By the time registration was complete, Bangladesh had issued over 81 million photo voter ID cards — one of the largest such exercises in the history of democratic administration, accomplished in a country of Bangladesh’s population and logistical complexity in under two years.
The 12.1 million fraudulent entries were removed. The rolls went from a corrupted database that neither major party trusted to a verified national registry that every monitoring organization praised as fit for purpose. When Bangladeshis went to vote on December 29, 2008, they voted against a list that actually reflected who they were.
No government before the caretaker period had done this. No elected government had been willing to, because whoever controlled the voter rolls controlled outcomes, and whoever controlled outcomes had an obvious interest in preserving the system. The caretaker government had no such interest. It built a clean list, and it worked.
The Anti-Corruption Drive: Imperfect, but Real
The caretaker government’s anti-corruption campaign is the most contested part of its legacy, and the contestation is, to a large degree, understandable. Some of what it did was genuinely overreach. The emergency powers that enabled arrests without the usual procedural protections were used in ways that were sometimes arbitrary. Cases were filed that would not survive judicial scrutiny. People were held for longer than due process required.
That is a fair criticism. It should be made. It should also be held in context.
The context is this: the government inherited from BNP-Jamaat’s five-year rule (2001–2006) a state that Transparency International had ranked as the most corrupt country in the world for five consecutive years. Not the most corrupt in the region. Not the most corrupt in Asia. The most corrupt on earth, across all nations studied, for half a decade running. The institutions that were supposed to check that corruption — the judiciary, the police, the Anti-Corruption Bureau — had been systematically subordinated to political patronage. The Anti-Corruption Bureau, specifically, had been described by the caretaker government’s own advisers as a body that existed to protect the powerful rather than prosecute them.
Against that backdrop, the caretaker government moved to revitalize the Anti-Corruption Commission (ACC). New leadership was installed. Case files that had been sitting in drawers for years were reopened. Investigators were authorized to pursue targets regardless of their political affiliation — and, critically, that authorization was real rather than notional. Both major parties were hit.
Sheikh Hasina, the Awami League leader, was arrested in July 2007 on charges of extortion and abuse of power. Khaleda Zia, the BNP chairperson and former Prime Minister, was arrested in September 2007 on charges related to the Zia Charitable Trust and the Zia Orphanage Trust — the cases in which she was eventually convicted in 2018 before her conviction was also later overturned. Tarique Rahman, Khaleda’s son and the man accused in WikiLeaks cables of operating a parallel government from Hawa Bhaban, was arrested in March 2007. Against him alone, the ACC filed cases covering money laundering, bribery, and the abuse of political office to extract contract commissions.
More than 160 politicians, businesspeople, and officials were detained during the anti-corruption drive. The list was bipartisan — BNP figures and AL figures faced the same courts. That is precisely what neither party had been willing to do when it held power. Under AL, BNP politicians were prosecuted and AL politicians were protected. Under BNP, the reverse. The caretaker government, having no electoral base to protect, prosecuted both.
The cases were imperfect. Some were dropped. Some convictions were later overturned — though the overturning of those convictions, which happened primarily between 2024 and 2026 under conditions of profound political pressure, is itself a story worth examining. What the caretaker government demonstrated was that accountability was physically possible — that you could investigate a sitting prime minister’s corruption and the country would not fall apart. That was not a small lesson.
Separating the Judiciary: The Reform That Was Never Reversed
For decades, Bangladesh’s judiciary sat formally under the executive branch. The Ministry of Law controlled the appointment, posting, and promotion of subordinate court judges. This was not an abstract constitutional problem — it was a live operational one. Judges who decided cases in ways that pleased the government in power were rewarded. Judges who decided cases in ways that displeased it faced career consequences. The subordination of the judiciary to executive control was one of the structural mechanisms through which Bangladesh’s governments had maintained impunity for their allies.
The separation of the judiciary from executive control had been mandated by the Supreme Court in a 1999 judgment — the Masdar Hossain case — but successive governments had found reasons to delay implementation. BNP, which held power from 2001 to 2006, did not implement it. The caretaker government did.
On November 1, 2007, the caretaker government formally separated the judiciary from executive control, transferring administrative authority over subordinate courts from the Ministry of Law to the Supreme Court. This was not a cosmetic change. It altered the institutional relationship that had enabled executive manipulation of judicial outcomes across multiple governments. It was a structural reform of the kind that takes years to plan and typically gets sacrificed when political calculations make it inconvenient.
The caretaker government did it anyway, because it had no political calculations to protect.
December 29, 2008: The Election That Worked
The ultimate measure of whether the caretaker government delivered on its mandate is the election it held. Bangladesh’s history is full of elections — some stolen, some manipulated, some held under conditions that made meaningful competition impossible. December 29, 2008 is the exception.
Turnout was approximately 87 percent — one of the highest in Bangladesh’s history, and a remarkable figure given the two-year emergency that preceded it. The result was decisive: the Awami League-led grand alliance won 263 seats out of 300, while BNP-led alliance won 32. A landslide by any measure. More importantly, a landslide that nobody seriously disputed as the product of anything other than actual votes.
The Carter Center, which sent a full delegation of election observers, called the election “a significant step forward for Bangladesh’s democracy.” The European Union’s Election Observation Mission said it was conducted “in a peaceful and credible manner.” The Commonwealth Secretariat, the Asian Network for Free Elections, and the US State Department all issued assessments that described the election as meeting international standards for a free and fair democratic exercise.
To understand why this matters, consider what Bangladesh had experienced immediately before 1/11. The Fourteenth Amendment had extended the Chief Justice’s retirement age specifically to ensure a favorable appointment as chief of the neutral caretaker government — a constitutional manipulation that the Supreme Court later struck down. The voter rolls contained 12.1 million phantom entries. Pre-election violence had killed more than forty people. The Awami League had refused to participate in the election scheduled for January 22, 2007, because it judged those conditions to be fundamentally incompatible with a credible result. They were right.
The December 2008 election that the caretaker government eventually organized, after two years of administrative work, was the opposite of what January 2007 would have been. It was run against clean rolls. It was monitored comprehensively. It produced a result that the losing parties accepted — however bitterly. That is what a functional election looks like. Bangladesh had not had one for a long time, and it required a caretaker government to deliver it.
What BNP Says — and Why It Doesn’t Hold
The standard BNP framing of the caretaker period goes something like this: 1/11 was a military coup disguised as a civilian government. Politicians were arrested on fabricated charges. The emergency powers were abused. The whole enterprise was illegitimate, and its architects should be held accountable for what they did.
Elements of this are true. Emergency powers were used beyond what strict due process would permit. Some of the cases filed were weaker than others. The experience of being arrested and detained without normal procedural protections was genuinely harmful to many people, including some who had committed no wrongdoing.
But the framing collapses when you look at what the caretaker government actually delivered. The voter ID system that BNP praised when it helped them and denounced when it didn’t — the caretaker government built it. The judiciary separation that BNP claimed credit for when it was in opposition — the caretaker government implemented it. The December 2008 election that BNP participated in, competed in, and accepted the results of despite a crushing defeat — the caretaker government organized it.
BNP won 30 seats in 2008. It did not claim the election was rigged. It did not boycott the result. It accepted what the voters decided. That behavior is itself a form of institutional validation — you do not participate in an election and accept an adverse result unless you believe the process was legitimate enough to warrant it.
What changed was not the election. What changed was that BNP needed a political narrative. “1/11 was a dark chapter” is a useful narrative for a party whose leadership was arrested during it and whose corruption cases were filed by it. It is not an accurate accounting of what happened.
The People the Narrative Erases
There is a human dimension to this accounting that matters and that the current political moment makes it difficult to discuss.
The officers, officials, and civil servants who implemented the 2007–2009 reforms were not acting without principle. They were people who had watched Bangladesh’s institutional framework deteriorate across decades of partisan governance and who, given an unusual window of political space, worked to fix structural problems that elected governments had repeatedly declined to address. The voter ID project was built by thousands of registration workers, army personnel, and election commission staff operating under difficult conditions across a country of 160 million people. The anti-corruption cases were filed by lawyers and investigators who understood they were taking professional risks.
These people did their jobs. They deserve a fair accounting of what they produced.
The current political environment in Bangladesh is not interested in fair accountings. BNP, which is in power and which has its own reasons to cast 1/11 as illegitimate, has actively promoted the prosecution of officials associated with the caretaker period. The framing of 1/11 as a criminal enterprise requiring legal accountability for its architects serves the party’s political interests. It does not serve historical truth, and it does not serve Bangladesh.
The institutions Bangladesh relies on today — the national identity database built by the voter ID project, the separated judiciary, the Election Commission infrastructure — are products of the caretaker period. You cannot simultaneously use the database and prosecute the people who built it on the grounds that building it was a crime.
The Standard Nobody Wants to Apply
Here is the simplest version of the caretaker government’s report card: it came to power in January 2007, it left power in January 2009, and in between it held an election that the international community considered legitimate and that the losing parties accepted.
By that standard — the standard of ultimately delivering credible democratic process — the caretaker government succeeded at its core mandate. It succeeded at something that BNP’s 2001–2006 government had been conspicuously failing at. It succeeded at something that AL’s 2014 and 2018 elections made a mockery of.
The 2014 election, held under AL’s watch, had an opposition boycott and saw 153 of 300 seats uncontested. The 2018 election was described by international observers and domestic monitors as marred by ballot stuffing, intimidation, and manipulation on a scale that called the result into question. Neither election produced the kind of credible, internationally validated result that December 2008 produced.
The caretaker government did not get everything right. No two-year transitional administration could. But it met the test that actually matters for a transitional government: it handed power back, and it handed it back following a process that Bangladeshis and international observers could recognize as genuine.
BNP is in power today in part because of the institutions and processes the caretaker government built. The voter ID system, the reformed rolls, the Election Commission infrastructure, the precedent of internationally monitored elections — BNP inherited all of it. It won in 2026, in elections that took place against that inherited infrastructure.
Gratitude is not a political requirement. But honesty should be. The caretaker government of 2007–2009 deserves to be assessed on what it actually did, not on the narrative that currently serves the most powerful people in Bangladesh. What it did was fix Bangladesh’s elections. What it left behind was the foundation for the democracy that came after it.
That record exists. It is documented, internationally verified, and written into the history of every election Bangladesh has held since. It will outlast the narratives being written around it today.
Bangladesh Untold is a platform committed to documenting Bangladesh’s recent history with specificity and source integrity. This article draws on reports from the European Union Election Observation Mission (2008), the Carter Center’s Bangladesh election assessment (2008–2009), Transparency International’s Corruption Perceptions Index methodology and Bangladesh country reports (2001–2010), and contemporaneous reporting by The Daily Star, BBC, Al Jazeera, and Reuters. All assertions of documented fact carry identifiable sources.
Leave a Reply