The Hundred-Billion Rupee Question: The 7.5 Million vs 12 Million 'Data Trap' in Pakistan's Tax System

In Islamabad, a sheet of FBR (Federal Board of Revenue) data was open, and it contained a ratio that could keep any data desk awake at night. Nearly 12 million Pakistanis have registered and filed returns for a one-time property or vehicle transaction in their lifetime, but the number of continuous filers is only 7.5 million. This gap of 4.5 million is no small accounting discrepancy. It is the gap between an administrative system and a genuine taxpayer base. The moment I saw these two numbers side by side, the old line I repeat in every data assignment came back to me: what you are calling a 'taxpayer' might just be a 'registration'.

To understand these numbers, one must first determine what the FBR is actually measuring. When a citizen buys property or a vehicle, the law forces them to record the transaction. That individual registers on the IRIS platform, uploads documents, and files a one-time return. Technically, they enter the 'filer' category. But when income arrives next year, do they continue filing? Mostly not. Their purpose was not tax compliance but completing a property transaction. In my view, this 'one-time filer' category is the biggest data trap in Pakistan's tax system. Until you separate 'registered' from 'active compliant,' every policy you frame will run on estimation, not data.

The FBR keeps issuing notices to these 12 million one-time filers. Notices go out via email. Here an operational flaw emerges that I often see in my work: the system does not consider anyone a 'non-compliant' until that person voluntarily opens their IRIS account. In other words, until you return to the system to conduct a transaction, the system does not even know you are not filing. This is a technical short circuit—the system is blind to its own reporting. The Finance Act 2026 tried to address this directly: if you are not transacting, restrictions on economic activities are designed to ensure you do not need to file a return.

Here an interesting angle emerges. The Finance Act 2026 proposed restrictions on economic transactions, but according to sources, these were diluted due to political-economic pressures. For me, this is a signal that the intent to reform existed but its 'run-time' performance faded on the political field. When you turn a regulatory mechanism into a weapon to expand the tax base, you must also determine how those genuinely unable will be excluded. Low-income citizens who only need to buy a property or vehicle once will fall straight into penalties if they do not know the grievance system. This is an administrative risk, not a moral one.

Contrarian Angle

Here comes the angle I always emphasize: the large size of a data set is not a problem in itself, until you read it incorrectly. The 12 million number is not bad—it is simply 'registered.' The 7.5 million are 'continuous.' The difference is whether you call the 12 million 'taxpayers' or 'transaction holders.' If you put both in the same account, your policy will always remain hazy. A senior once told me: "The biggest enemy of a tax system is not theft, but counting error." I agree with him. The real metric is not 'how many registered' but 'how many are consistently accounting for their income.'

Takeaway

When the FBR releases new data over the next six months, I will look at just one number: is the one-time filer count dropping, or shifting toward continuous filers? If the latter, the system is on the right track. And if not, it means we are still reading the 1.2 crore figure as 'taxpayers,' when it is really just an entry. Just as winning the toss is not winning the match, filing a return is not tax compliance. The question remains: will Pakistan's tax system move from the crowd of 1.2 crore to the precision of 75 lakh?