Bangladesh's tax conundrum: Why do reforms fail to bring change?

In Bangladesh, every budget sets an ambitious revenue target, and by the end of the year, it's the same story—the target hasn't been met. There are familiar explanations: tax evasion, weak administration, corruption, a large informal economy, limited use of technology, and reluctance to pay taxes.

These reasons are true. But after so many reforms, new laws, assistance from development partners, and digital systems, why do the problems remain? The National Board of Revenue (NBR) is not unaware of the problem; their plan includes a blueprint for modern tax administration. Yet, reforms get stuck.

In the NBR's presentation to the parliamentary standing committee on 23 August, there was a startling piece of information. The amount of tax exemptions is 6. 41 per cent of GDP, while total tax collection is only 6.8 per cent. In other words, we give away almost as much in revenue exemptions as we collect in taxes.

NBR has identified this issue too. They want to run the tax administration based on information, rules, and risk assessment, instead of personal decisions by officials. In other words, who gets noticed will be determined by data and rules, not by the officials.

The obstacle is not unknown to them either. Income tax, VAT, and customs still operate in separate realms; a taxpayer's complete profile is not available in one place. Moreover, tax policy and tax collection are housed in the same institution, with powers for assessment, audit, and enforcement being quite centralised. Information is dispersed, and decision-making power is centralised. The question is, why is finding a solution so difficult, even when the problem is known? To answer that, we need to look at the tax reforms of the last three decades.

In 1991, with the introduction of VAT, tariff reductions, and trade liberalisation, major economic reforms began. Initially, there were positive results. The government's revenue as a percentage of GDP rose from around 6.5 per cent in the 1980s to about 9. 6 per cent in the fiscal year 2014-15. But that momentum did not continue. Subsequently, initiatives such as automation, large taxpayer units, audit reform, and expanding the tax net were undertaken. However, tax exemptions, negotiations, and opportunities for personal decisions by officials remained. The VAT law of 2012 followed the same path—its implementation was delayed, weakened, and eventually enacted in 2019.

One trend is clear: Bangladesh has adopted various elements of a modern tax system, but reforms that threaten entrenched benefits and personal power have been difficult to sustain. Thus, the question is not just about what reforms are needed. What have other countries done in the face of similar challenges that Bangladesh hasn't been able to achieve?

China's path was to clarify authority and responsibility. In their 1994 reforms, tax policy was kept under the Ministry of Finance, while a strong national tax administration directly under the State Council was established for tax collection. Central and joint tax collections were removed from local government control. As a result, the boundaries of who makes the policy, who collects the tax, and who holds the information became clear. Later, digital systems became effective atop this structure.

India's path involved compromise and digital connectivity. For GST, compensation and the GST Council played crucial roles in aligning central and state interests. Simultaneously, linking Aadhaar with the Permanent Account Number (PAN) for taxpayers, online returns, and tax assessments without face-to-face interaction brought taxpayers into a unified information system. In corporate tax, there was room to give up many exemptions for lower rates.

Nepal's path was driven by necessity. Despite being poorer and having a larger informal economy than Bangladesh, its tax-to-GDP ratio was about 14 per cent in 2023. A new federal structure, local government, and social spending increased the need for revenue; simultaneously, efforts are ongoing to reduce tax exemptions, widen the tax base, and make tax compliance easier with digital systems.

The paths of the three countries are different, but the common ground is crucial: more revenue didn't just come by increasing tax rates. Expanding the tax net, reducing special privileges, unifying information, and reducing opportunities for personal decisions played a role too.

For Bangladesh, the question is not just how much tax to collect, but how it is to be collected. The NBR's proposed ''integrated NBR structure'' aims to bring income tax, VAT, and customs under a single administrative system to unify information and risk management. This is important but not sufficient. This requires separating the policy work of determining tax rates, bases, and exemptions from tax collection. The NBR has itself identified this weakness. Similarly, clear boundaries are needed between assessment, audit, enforcement, appeal, and taxpayer services.

But this is also only half the story. The NBR says the tendency for voluntary tax payment is weak, and inconsistencies in administration reduce taxpayer trust. But trust doesn't just depend on how taxes are collected; it also matters how the tax money is spent. Therefore, increasing only the power to collect taxes will not suffice. Tax reforms are more sustainable when they are part of a broader revenue consensus—citizens will pay more taxes in exchange for better public services and see that their money is being spent in public interest, not for the benefit of special groups. Thus, the capacity to collect revenue and accountability in government expenditure must both grow together.

When an interim government issued an ordinance to separate tax policy from tax collection, there was intense resistance from within the NBR. Some of the officials' objections were logical, especially regarding their professional role and leadership in the proposed new structure. But the protest quickly escalated to a work stoppage and shutdown of revenue operations; work at ports was also disrupted. Eventually, the government had to return to negotiations and amend the ordinance.

The lesson from this experience is that institutional reform doesn't happen through ordinances alone. If those whose power will diminish hold the keys to keeping revenue and trade going, they effectively have veto power; even more so in politically vulnerable times. If the three branches come together, institutional boundaries will change, influence zones will shift if policy and administration are separated, and personal power will reduce if information is unified. Making everyone adversaries creates a coalition against reform.

The challenge is made even tougher by the need to transition from LDCs. If the dependence on import duties decreases, the shortfall will have to be made up from domestic taxes. But that doesn't mean taking more from existing taxpayers. The tax net needs to expand, and revenue lost in negotiations, loopholes, and personal decisions must be reduced.

The question is not just what reform is urgent but where to touch so as to start shaking the political foundation of the current system. Tax exemptions are one of those areas. The problem is not exemptions themselves; many exemptions have logical reasons. The issue is that discretionary power to grant or retain exemptions aligns the interests of two sides—those inside who want to retain that power and those outside who benefit from it. This can create an alliance to resist reform.

Therefore, the first task is not to abolish exemptions but to bring them within rules. A comprehensive account will reveal who is getting how much benefit and how much revenue the state is giving up. Every exemption should have a term and public rationale; it should end unless newly approved. This makes it the beneficiary's responsibility to prove the justification for keeping the benefit, not the government's burden to remove it.

Ultimately, the pressure for change may come from financial necessity. Bangladesh has not lacked reform plans. The question is, does the political cost of maintaining the old system now exceed the cost of changing it?

#Zahid Hussain is the former lead economist of the World Bank's Dhaka office.

*The opinions expressed here are the author’s own.

#This article, originally published in Prothom Alo print and online editions, has been rewritten in English by Rabiul Islam