Where Does Power Reside, and Where Does Accountability Lie?
People do not vote for bureaucrats; they vote for political leaders. Therefore, in a democratic system, the government must bear political responsibility for any failure. But although the government holds the state’s overall authority, it does not control every function of the state directly. Failing to understand this distinction leaves any explanation of Bangladesh’s state system incomplete.
Our common assumption is that whoever gives the final signature on a file holds all the power. In reality, however, the power behind an administrative or policy decision does not lie only in that final signature. Which information is placed before a minister, what is left out, which alternatives are prepared, and which proposal is described as realistic or risky—these are all part of where power operates. Even how long a file remains at a particular desk can become important. Power, therefore, does not exist only in the final decision; it is also spread throughout the process through which that decision is made.
A large part of Bangladesh’s state machinery operates through this permanent administrative system. Article 55(2) of the Constitution states that the executive power of the Republic shall be exercised by or on the authority of the prime minister. Under sub-rules 4(1) and 4(4) of the Rules of Business, 1996, the minister in charge holds political authority over the activities of a ministry or division. At the same time, under sub-rule 4(5), the secretary is the administrative head of a ministry or division and is responsible for its administration, discipline and proper execution of assigned functions. The secretary is also responsible for keeping the minister informed about the activities of the ministry or division.
In legal terms, bureaucrats are not above political leaders. But much of the lengthy administrative process through which political decisions are turned into state action is handled by permanent officials. This creates a clear difference in power. A minister has political legitimacy, while a senior official has years of institutional experience. The official knows where a document can get stuck, which rules apply, whose opinion is required and where a proposal can remain stalled indefinitely. The minister knows what he or she wants to achieve; the official knows where the real complexities lie.
When these two strengths work together, the state moves forward. When they become disconnected, an invisible gap emerges between policy and implementation. The real problem arises when this balance is disrupted. Weak political leadership can lead to bureaucratic dominance, while weak bureaucracy can increase the misuse of political power. A healthy state system should ensure that neither side can hold the other hostage.
The major question now is where and to what extent decision-making power is concentrated in the state. This is not a new debate in Bangladesh. According to government data on promotions published on August 28, 2025, 198 of the 268 officials promoted to deputy secretary were from the administration cadre. The Public Administration Reform Commission, however, had recommended a 50:50 ratio between the administration cadre and other cadres for promotion to the post of deputy secretary.
This is not a question of the personal competence or integrity of individual officials. The central issue is whether other forms of professional knowledge are receiving enough space when the same pattern of professional promotion becomes too dominant at important levels of state decision-making.
Health Ministry decisions require medical knowledge. Education policy requires deep experience in education and pedagogy. The same applies to energy, banking, industry, agriculture and the labour market. The administration’s role is to coordinate the state. But if that coordination traps expert opinion in layers of bureaucratic approval, questions about the quality of decisions are inevitable.
Therefore, the issue should not be reduced to a conflict between “bureaucrats and politicians”. The relationship between politics and administration is much more complex. During the long tenure of any government, mutual interests and dependence can develop between political leaders and the administration. Sheikh Hasina’s long rule was no exception. The political leadership must certainly bear responsibility for the failures of that period. But no long-lasting government survives through political instructions alone. The cooperation, loyalty and implementation capacity of permanent institutions also play an important role.
Therefore, the entire situation cannot be explained by portraying politicians as victims and bureaucrats as villains. Political leaders depended on the administration, while many parts of the administration also depended on political decisions, appointments and benefits. This mutual dependence can sometimes help run the state, but it can also create a closed circle of power.
So, does the state change when the government changes?
A change in political leadership can certainly change many things. Policy priorities change, ministers change, appointments change and some decisions are reconsidered. But if the methods of work remain the same—the movement of files, stages of approval, or the discretionary powers of officials—the new government may also become trapped in the same old problems.
In other words, the government may change without changing the machinery that drives the state. After some time, the same complaint may emerge again: the government is unable to work. The government may indeed bear some responsibility. But another question will arise: how much power did it actually have, and how high were the bureaucratic barriers to exercising that power?
Time is another major issue. Administrative delays are often dismissed as part of the “procedures and rules”. But if a project is delayed by six months, investment is delayed. Passports, land records, hospital approvals and documents required by migrant workers are familiar examples for people in Bangladesh.
For the administration, it may simply be a file. For an ordinary citizen, however, it can mean income, medical treatment, future employment or even a struggle for survival. If a migrant worker’s clearance is held up, an entire family’s income and dreams may also be held up. The state does not have unlimited time, and neither do citizens. Therefore, when evaluating an administrative decision, it is not enough to ask whether the work was eventually completed. How quickly and within what timeframe it was completed must also be considered.
The solution, however, is not to weaken the administration cadre. A modern and effective state cannot function without a skilled, stable and professional administration. At the same time, replacing the dominance of one cadre with the dominance of another would not be a real solution. It would merely shift the centre of power while leaving the basic problem unchanged.
What is needed is a balanced distribution of power. Economists, doctors, engineers, educators, agricultural scientists, technology experts and other specialists should have a meaningful role in policymaking where their expertise is required. The government should also keep open the option of bringing in skilled professionals from outside the state system when necessary.
The administration should remain a strong pillar of coordination and implementation, but it should not become the only gateway to every issue. At the same time, the discretionary powers of bureaucrats should be made more transparent and accountable. There should be clear records of where a decision was delayed, why it was delayed, who had the authority to act and how much time was required. Power can exist, but the exercise of that power should not remain hidden.
The Public Administration Reform Commission made several important recommendations on administrative reform and the promotion of government employees. Along with a 50:50 ratio between the administration cadre and other cadres for deputy secretary positions, the commission recommended reducing the number of secretaries and equivalent officials from 80 to 60.
Therefore, the issue should not be limited to what percentage of positions should go to a particular cadre. The central question should be how state decision-making can become more efficient, diverse and accountable. Political leaders also have a major responsibility here. Leadership does not end with giving instructions. Political leaders must understand how the administrative machinery through which those instructions are implemented actually works, where power is concentrated and which responsibilities can be delegated.
Experienced, farsighted and strong political leadership does not need to enter into conflict with the administration. It can get things done by understanding the language of rules and procedures. A major and uncomfortable dilemma exists in a democratic state. A person who is directly elected and accountable to the people does not always have the same level of effective administrative power. At the same time, those who hold much of the state’s practical and day-to-day administrative power are not directly held accountable by voters.
The wider this gap becomes, the more difficult it is to distinguish between government failure and state failure. The problem becomes more complicated when political leaders are inexperienced, do not understand administrative rules and procedures, or become excessively dependent on the administration for personal or group interests. Political leaders then become dependent on the administration not only for information but also for navigating the process.
As a result, elected political authority may exist only on paper, while the administrative structure begins to determine the pace of actual state decisions. The core of reform in Bangladesh, therefore, is not simply about who will form the government. The real question is how effective the power brought by an elected government remains within the state.
Changing the government through elections is the first step of democracy. But changing rulers alone does not change the state. To change the state, the process of decision-making must also change. Power that is unnecessarily concentrated must be distributed. The administration must become more accountable, not weaker. Politics must become effective, not all-powerful. And expert knowledge must be brought from the formal advisory room to the main decision-making table.
On election day, people can change the government. But the state changes through the speed of files, balanced distribution of power and accountability. Without this internal change, new faces will appear and new promises will be made, but the old structure of the state will remain unchanged.
Author: Associate Professor, Shaheed Buddhijibi Government College, Rajshahi
Email: hhafij@yahoo.com
