A fractured institution, a violent year: the case for arming Dhaka’s election safety net

The question of whether soldiers belong anywhere near a ballot box has animated democratic theorists since Athens first debated who should guard the assembly. Bangladesh’s Election Commission has now stepped into that old argument, quietly amending the legal definition of “law enforcement agency” to include the armed forces ahead of local elections due to begin later this year. The decision deserves scrutiny on its merits rather than reflexive alarm.
Start with what the Commission actually decided on July 22. Five measures were finalized: MPO-listed teachers barred from candidacy, loan defaulters and their guarantors disqualified, candidates with unpaid utility bills ruled ineligible, party symbols removed from local contests entirely, and the legal groundwork laid for possible military deployment, mirroring the framework already used in national elections. Read together, these are the instruments of an institution trying to raise the floor of electoral integrity, not lower it.
Critics of the army provision tend to argue from abstraction. They invoke the specter of intimidated voters, of a chilling effect on turnout, of civilian authority quietly ceding ground to men in uniform. These are legitimate theoretical concerns in the vacuum of political science seminars. Bangladesh, however, is not choosing between an idealized peaceful election and a militarized one. It is choosing between institutional preparedness and institutional naivety, and both the data and recent history make that choice fairly stark.
There is a structural reality here that critics often skip past. The police department that would ordinarily shoulder the full weight of election security has not fully recovered from the institutional collapse that followed the July-August 2024 uprising. In the chaos of that period, police stations across the country were attacked and vacated, officers went into hiding or walked off duty entirely, and the force effectively ceased functioning in large parts of the country for days. Rebuilding that institution, its manpower, its morale, its standing with the public, has been a slow process rather than an overnight fix. Asking a police force still mid-repair to single-handedly secure a nationwide cycle of local elections, without any reserve option, is not caution; it is wishful thinking dressed up as principle. Keeping the armed forces available as a legal backstop is a direct, sensible response to a real capacity gap, not an ideological preference for uniforms over badges.
Layered on top of that institutional strain is the raw violence data. The Human Rights Support Society’s July 2026 report counted 189 deaths from various forms of violence in the first half of this year, 56 of them explicitly political and 133 arising from mob violence. Over 5,000 people were injured in factional political disputes over the preceding twelve months. A country carrying that burden of unrest into a nationwide cycle of local elections, spanning union parishads through city corporations, cannot reasonably be asked to rely on an under-strength police force and hope as its security strategy.
What the Commission has actually built, though, is a contingency, not a mandate. Officials have said plainly that there is no intention to station troops at every polling booth as a matter of routine. Deployment, where it happens at all, will follow an assessment of how the first phase unfolds and what local conditions demand. Civilian police retain primary responsibility for order; the military exists as a reserve option written into law rather than an occupying presence written into practice. That distinction, between having the authority to act and choosing to exercise it, is precisely what separates prudent contingency planning from creeping militarization. Confusing the two flattens a nuanced policy into a caricature.
There is also a historical dimension worth weighing. A significant share of local government seats had been stripped of elected representatives and handed to appointed administrators under the previous administration, which makes the coming vote unusually consequential for restoring genuine local governance. Political parties, BNP among them, are already working at the grassroots to settle on single candidates and avoid the kind of internal rebellion that has fueled violence in past cycles. An electoral process this significant, arriving after both a representational vacuum and a policing vacuum, warrants more institutional caution than an ordinary vote might, not less.
None of this forecloses debate about where the line should sit, or about what safeguards should govern any eventual deployment, transparency in the decision-making process, clear rules of engagement, oversight mechanisms that prevent abuse. Those are the questions worth pressing the Commission on. But the underlying premise, that a country recording triple-digit political and mob-violence deaths within six months, with a police force still rebuilding from institutional collapse, should keep every available legal tool on the table for its most consequential local elections in years, is difficult to argue against in good faith.