The referendum design: A plain-language guide to what Bangladesh’s vote authorises —and what happens next

Bangladesh is heading towards a referendum. After years of broken elections and weakened institutions, constitutional reform feels both desirable and necessary.

This piece does not argue for or against the referendum. Its purpose is to explain, in plain language, what the vote authorises in law—and what follows depending on the outcome—so that debate rests on understanding rather than assumption.

What is this referendum about?

The referendum follows a period of political transition that culminated in the July National Charter, a political accord among major parties and key actors.

The referendum will be held on the same day as the general parliamentary election. Voters will receive two ballots: one to elect Members of Parliament and a separate one for the referendum question.

The referendum does not ask voters to approve a new constitution or a final set of constitutional amendments. Instead, it asks voters to approve two things together: the Implementation Order of the July National Charter, and the reform directions set out in the Charter itself (Referendum Ordinance 2025, s.3; Implementation Order, s.4).

In plain terms, this is a vote to approve a process and a roadmap, not a finished constitutional settlement.

The referendum itself does not amend the Constitution or fully incorporate the July National Charter into it. Rather, it authorises a process through which elected representatives may later give constitutional effect to some or all of the Charter's proposals.

What happens if the referendum passes?

If a majority votes "Yes", the legal consequences are immediate. The Implementation Order comes into force, and a Constitutional Reform Council is created with authority to carry out constitutional reforms (Implementation Order, s.7(1)(a)–(b)).

Its members are the same individuals elected as Members of Parliament. They therefore serve simultaneously as ordinary legislators and as members of the Council with authority to carry out constitutional reform (Implementation Order, s.7(1)(b)).

The Council sets its own procedures and must complete its work within 180 working days (Implementation Order, s.7(1)(c)).

Once the Council adopts constitutional reforms, those reforms are treated as final for the purposes of implementation. There is no further approval and no second referendum on the final text (Implementation Order, s.14(1)).

For ordinary readers, the central point is that a "Yes" vote does not itself change the Constitution.

Instead, it authorises a time-bound reform process, led by elected representatives acting under a special mandate, to redesign the Constitution later.

What consequences follow if the referendum succeeds?

A successful referendum has consequences beyond signalling public support for reform. It determines how constitutional authority will be exercised during the reform process and where decision-making power will be located once the vote is over.

First, public consent is exercised once, at the beginning. Voters approve the roadmap and the Implementation Order, but do not vote on the final constitutional text.

After the referendum, constitutional design decisions are made by the Constitutional Reform Council within the authorised framework, with no second referendum. Public influence is therefore concentrated at the start of the process rather than at its conclusion.

Second, the mandate bundles multiple reforms into a single vote. Because distinct issues are combined into one question, it may later be difficult to determine which elements were decisive for voters and which were accepted as part of a package.

This ambiguity may matter when specific reforms become contested during implementation.

Third, the reform process operates under a fixed 180-day deadline. Time limits can encourage decisiveness and reduce prolonged uncertainty.

They may also narrow deliberation on complex institutional questions, particularly where broad consultation would otherwise occur.

Fourth, disagreement does not disappear. A "Yes" vote resolves the question of whether reform should proceed through this mechanism, but it does not settle disputes over substance.

Political contestation may shift from public campaigning to negotiations within the Council, legal challenges, or mobilisation outside the formal reform process.

Fifth, there is a risk of institutional deadlock. The Reform Council is composed of elected representatives with divergent interests and must resolve contested constitutional questions within a limited timeframe.

Even the July National Charter was not unanimously agreed, with several parties registering formal dissent.

If the Council fails to reach agreement on key reforms, the legal texts do not clearly specify what follows. This raises the possibility of incomplete reform or renewed political uncertainty.

Finally, the referendum pathway raises difficult constitutional questions.

Article 142 of the Constitution requires a two-thirds majority of the total membership of Parliament for constitutional amendments.

The Implementation Order, by contrast, authorises the Reform Council—composed of the same MPs—to adopt final constitutional reforms by a simple majority of members present and voting (Implementation Order, s.10(3); s.14(1)).

This creates a parallel amendment pathway that departs from the Constitution's ordinary prescribed method for change.

Proponents argue that popular mandate, expressed through the July uprising and the referendum, stands above written procedure.

Yet the interim government itself operates under the existing Constitution. The President issued the Implementation Order under Article 48, and the referendum ordinance was promulgated under Article 93.

If the Constitution remains in force for these purposes, it is unclear why its amendment rule would not also apply.

Constitutional procedures have been abused wholesale in the past. But that experience alone does not explain why the Constitution's own amendment rule should be set aside now.

Rather, it may reinforce the need for political actors to comply with and respect constitutional rules if reform is to be durable.

In effect, voters are being asked to authorise an alternative amendment pathway that bypasses the Constitution's ordinary two-thirds majority requirement.

The question is whether the need for speedy reform justifies the risk of weakening the very rules meant to protect the state from easy, one-sided changes.

This tension—between claims of popular mandate operating outside the Constitution's amendment procedure and the continued operation of the Constitution itself—may ultimately fall to the courts to resolve.

That would place judicial institutions under significant pressure, regardless of how they respond.

What happens if the referendum fails?

In legal terms, a "No" vote leaves the Constitution unchanged.

Any future amendments would have to be pursued by the new Parliament through the ordinary amendment procedure under Article 142. The Implementation Order and the bundled reform package would lapse.

Politically, however, the legal texts do not specify what follows.

There is no automatic alternative pathway for reform and no agreed fallback mechanism. Reform momentum could stall, even where there is broad agreement that change is needed.

Competing narratives may emerge, with some treating the result as a rejection of reform itself and others as a rejection of the process rather than the substance.

Seen together, the referendum presents voters with two different kinds of uncertainty, shaped by the design of the vote itself.

Why does the design matter?

The importance of the referendum's design becomes clearer when we recall what constitutions are meant to do.

Constitutionalism is not only about expressing values or listing rights. At its core, it is about limiting power so that no single institution can dominate the political system, even when that institution is elected.

In South Asia, this restraint is usually organised through separation of powers. Executives govern, legislatures make law, and courts review legality.

Each is meant to act as a check on the others.

This balance is never perfect. Bangladesh inherited an executive-heavy system designed for command rather than restraint.

Efforts to limit executive dominance have been uneven. Against that background, the specific design of a constitutional reform process matters because it shapes how power is exercised.

One vote, many choices

The referendum presents voters with a single yes-or-no question that bundles together several distinct reforms (Implementation Order, s.4(1)(a)–(d)).

These include changes to how caretaker government and constitutional bodies are formed; whether Parliament becomes bicameral through the creation of an upper house; whether future governments are obligated to implement agreed reforms; and whether remaining Charter proposals will be carried out through party commitments.

Voters cannot approve some elements while rejecting others.

Bundling does not make a referendum illegitimate. However, it makes voter consent less precise.

A voter may strongly support one reform while remaining uncertain about others, yet all are accepted or rejected together.

The scale of what is bundled here is also significant.

The July National Charter contains over 100 reform proposals across the Constitution, judiciary, elections, and other state institutions (July National Charter, pp. 11–57).

Importantly, consensus among the 33 signatory parties was not unanimous.

On key issues such as the creation of an upper house and the model for caretaker government, multiple parties formally recorded dissent during negotiations (see, for example, July National Charter, pp. 22, 26, 34).

A single "Yes" vote may therefore later be interpreted as endorsement of proposals that were contested even among the Charter's drafters.

Can future Parliaments be bound in advance?

The referendum design also raises questions about delegation and pre-commitment.

As a general constitutional principle, one Parliament cannot bind a future Parliament through ordinary legislation.

Each elected legislature is expected to exercise its own judgement on behalf of the electorate at that time.

Constitutions are the main exception. They represent collective acts of pre-commitment, adopted to restrain future majorities and stabilise democratic rules.

It is important to note a technical distinction here.

Section 4(1)(c) of the Implementation Order asks voters to consent to the proposition that "the political parties shall be obligated to implement the thirty reform proposals" set out in the July National Charter.

This formulation reflects a political obligation undertaken by the parties that signed the Charter, rather than a clearly specified legal obligation binding Parliament itself as an institution.

In other words, a "Yes" vote endorses a political promise made by parties, but the legal power to enact constitutional change comes from the separate authority granted to the Reform Council.

The pathway proposed here sits in an ambiguous space between these two categories.

By voting "Yes", the public grants a mandate at the start of the process, authorising elected representatives to draft and adopt final constitutional reforms.

Voters do not, however, get a second opportunity to approve the final constitutional text (Implementation Order, s.14(1)).

In effect, public consent is given in advance, delegated on the basis of trust in the Council and the broad directions of the July National Charter rather than on a settled text.

This delegation is reinforced by the dual role of elected representatives.

The same individuals are tasked with governing under the existing constitutional framework while simultaneously rewriting that framework.

Bangladesh has used elected representatives for constitution-making before, but in 1972 constitution-making was treated as a temporary task, distinct from ordinary governance.

Merging these roles makes it harder to separate long-term constitutional judgement from short-term political calculation.

It also means future Parliaments may be asked to implement constitutional changes that some parties opposed during the Charter negotiations.

A "Yes" vote becomes the Council's mandate to proceed despite those objections.

Time pressure and deadlines

The design of the reform process also concentrates significant control over timing and structure at the outset.

The Constitutional Reform Council has 180 working days to complete its work.

Deadlines can keep discussions focused, but compressed timelines in such high-stakes and often contested discussions tend to narrow deliberation on complex constitutional questions.

This is particularly true where broad consultation or reconsideration might otherwise occur.

Bangladesh's own constitutional history offers a cautionary parallel.

Under Pakistan's Legal Framework Order of 1970, the elected National Assembly was required to frame a constitution within 120 days.

That deadline, imposed by executive order, hardened positions and narrowed political space, contributing to constitutional breakdown.

The lesson is not that deadlines are always wrong, but that timelines imposed from outside the deliberative process can privilege insiders and limit opportunities for correction.

It also matters where this reform process originates.

The President issued the Implementation Order, setting the sequencing, scope, and 180-day timeline of reform.

While the Order invokes the "sovereign will" of the people, the reform process originates from the executive branch during a period when Parliament is dissolved.

In Bangladesh, excessive executive power is not a new problem; it is the problem that constitutional reform is meant to address.

Against that history, it is reasonable to ask whether a process aimed at diluting concentrated power may paradoxically begin by exercising it.

The question that actually matters

The central question before voters is not whether reform is desirable, but whether this particular pathway to reform is one they wish to authorise.

No constitution is perfect, and no design can permanently restrain power on its own.

Even if the reform process is completed, the durability of the resulting constitutional settlement will depend on more than its legal validity.

Constitutions endure when they command broad political acceptance and are resilient under changing political conditions.

Where reforms are adopted through tightly bundled mandates, compressed timelines, and without sustained cross-party buy-in, they may prove more vulnerable once political competition resumes.

Bangladesh's experience suggests that constitutional design can open possibilities—but it is durability, not speed or form alone, that determines whether those possibilities endure under pressure.

Bangladesh's transitional moments have repeatedly produced ambitious reform blueprints that falter once political competition returns.

The experience of earlier efforts, including the post-2008 electoral reforms, suggests that constitutional design can open possibilities, but it is politics that determines whether those possibilities harden into durable structures or dissolve back into familiar patterns.

Design choices nevertheless matter because they shape how power is exercised when trust eventually erodes and pressure returns.

Sketch: TBS

Dr Cynthia Farid is an Advocate of the Supreme Court of Bangladesh and a Researcher with the Chinese University of Hong Kong

Disclaimer: The views and opinions expressed in this article are those of the authors and do not necessarily reflect the opinions and views of The Business Standard.