Minimum tax withdrawal in 36 categories poses a new test for revenue collection

A national budget is not merely a statement of income and expenditure; it is a financial commitment of the government to its citizens. Yet every commitment ultimately hinges on one question: can the government collect the revenue it has projected?

In Bangladesh, one of the strongest foundations of income tax collection has long been tax deducted or collected at source (source tax). In theory, taxpayers earn income, prepare accounts, file returns, and pay tax on the basis of actual income. In practice, however, a large part of income tax is collected much earlier - at the point of transaction. Taxes are deducted when payments are made for imports, exports, contracts, supplies, bank interest, savings instruments, land or property transfers, share transactions and more.

This system ensures a steady flow of revenue to the exchequer. Even when taxpayers delay or misreport returns, the government has already received part of the tax into the public exchequer, ensuring a measure of revenue certainty. As a result, tax deducted at source has become a major pillar of income tax revenue in Bangladesh.

Since 2016, the size of the national budget and the revenue target have increased significantly, and a key basis for this confidence was the introduction and continuation of the minimum tax system. In many important categories, tax deducted at source was treated as minimum tax - meaning taxpayers could not claim refunds simply by showing low profits or losses. This allowed the government to treat a substantial portion of tax collected at source as relatively assured revenue. Put simply, minimum tax ensured that liability could not fall below a legally protected threshold, which acted as a safeguard for government revenue.

Where a category concerns minimum tax, three numbers are usually considered: the tax payable at the prescribed rate, the tax actually deducted at source, and the tax calculated under self-assessment or normal assessment. The highest of these figures becomes the minimum tax.

An example should make it clearer: Suppose a contractor receives Tk10 crore in bills during a particular income year. If the applicable rate is 5%, Tk. 50 lakh should be deducted; in reality, Tk30 lakh is deducted at source. Subsequently, he filed a return showing tax liability at Tk20 lakh. Under the minimum tax system, he cannot claim a refund of Tk10 lakh from the 30 lakh already deducted, nor settle with Tk20 lakh. Since the prescribed rate yields Tk50 lakh, that becomes his minimum tax. Furthermore, if the assessment indicates Tk60 lakh as his tax liability, he is bound to pay Tk60 lakh in taxes. Thus, minimum tax is not a protection for the taxpayer, it is a protection for government revenue.

This protection has now been substantially weakened since the Finance Act, 2026 withdrew minimum tax provisions from 36 categories.  In effect, what was once assured revenue that allowed the government to prepare and implement large budgets with reasonable confidence would now potentially return as refund claims.

The seriousness of our reality has heightened with the new government expanding the budget for the fiscal year 2026-27, continuing the trend of ambitious budgetary estimates. A larger budget requires a stronger and more reliable revenue base. Yet, at the same time, the government has withdrawn a legal safeguard that long ensured the dependability of source tax as a cornerstone of revenue collection. These two moves appear misaligned, with expanded expenditure targets undermined by the removal of a mechanism that guaranteed revenue certainty. 

A more cautious approach would have been to retain minimum tax for at least one year under the new government, observe revenue outcomes, and then decide. Instead, the upfront removal of this protection has potentially exposed the budget to avoidable pressure. 

Refunding excess tax is fair in principle - no taxpayer should suffer from excessive withholding. Where the government collects more than what is legally payable, the excess should be returned, and a modern tax administration must ensure prompt refund of genuine claims.

But the real concern lies elsewhere. Bangladesh's tax administration still struggles to quickly and accurately match income tax returns with import records, VAT returns, bank transactions, and business accounts. In our current state, making large portions of source tax easily refundable risks revenue leakage.

Data from the last five years underscores the concern. Between the fiscal years 2018-19 to 2022-23, total income tax collection was about Tk4.3 lakh crore, of which Tk2.69 lakh crore (nearly 63%) came from source tax. During the same period, imports, exports, contractors, and suppliers alone contributed about Tk1.22 lakh crore in source tax, which is about 45% of the source-based tax. Weakening this stream is no small matter.

Another important reality is that about 93% of income tax collection comes through source deductions, advance tax, and payments with returns. The strength of the system lies not in lengthy assessments but in taxes collected upfront or voluntarily. If the certainty of source tax is weakened, pressure on budget implementation will inevitably rise. 

The greatest vulnerability lies in high-volume business transactions. Importers, exporters, contractors, and suppliers handle large transactions, and their source tax deductions are equally large. The amount of tax deducted or collected from them is also large. If these taxes become refundable, taxpayers may report low profits or inflate expenses to claim large refunds. Some claims may be genuine, but many could be manipulated through accounting practices.  

For instance, a supplier receiving Tk100 crore in a year may have Tk5 crore deducted at source. If this tax is treated as minimum tax, no refund is possible merely by showing low profit. But if it is treated as ordinary advance tax, he could file a return demonstrating actual tax liability at Tk1 crore, and claim a refund of Tk4 crore. In a system where verification is weak, this poses a serious risk.

While the government has retained turnover tax in law, it cannot fully replace the certainty of tax deducted at source. Source tax is collected at the time of transaction, whereas turnover tax depends on the declared turnover in returns - the two are not equivalent. 

The issue is not merely one of taxpayer service but also about revenue certainty. Genuine refunds must, of course, be allowed. Nevertheless, the withdrawal of minimum tax from 36 transaction categories has changed the character of a large part of source-based revenue. What was once a protected revenue stream may now become a field of refund claims.

The decision to withdraw minimum tax from 36 transaction categories in the very first budget appears to be premature. It may please some taxpayers in the short run, but it risks undermining the very revenue base on which the budget depends. If refund claims rise sharply in high-volume categories such as imports, exports, contracts, and supplies, the government may face a serious challenge in achieving its revenue targets.

The central point is simple: tax genuinely collected in excess should be refunded quickly; but a revenue protection mechanism that supported budget implementation for years should not have been removed without first ensuring strong verification capacity. The withdrawal of minimum tax from 36 transaction categories may therefore become not only a tax policy experiment, but a potential test of budget implementation itself. 

Advocate Jahangir is a senior tax lawyer specialising in income tax law and regularly contributes writings on finance and fiscal legislation; he can be reached at jac@thetaxlawsyndicate.com 

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