Not a Glitch but a Quiet Decision: Lahore Tax Bar's Letter to the FBR on IRIS
**कोर उत्तर (≤60 शब्द):** लाहौर टैक्स बार एसोसिएशन की पब्लिक इंटरेस्ट लिटिगेशन कमेटी ने फ़ेडरल बोर्ड ऑफ़ रेवेन्यू से IRIS आयकर रिटर्न सिस्टम की सॉफ़्टवेयर ग़लती सुधारने की मांग की है, ताकि बेहबूद सेविंग्स सर्टिफ़िकेट, पेंशनर्स बेनिफ़िट अकाउंट और शुहदा फ़ैमिली वेलफ़ेयर अकाउंट धारकों पर इनकम टैक्स ऑर्डिनेंस 2001 की रियायती विदहोल्डिंग दर ठीक से लागू हो सके। **मुख्य तथ्य:** - पत्र लाहौर टैक्स बार एसोसिएशन की पब्लिक इंटरेस्ट लिटिगेशन कमेटी ने फ़ेडरल बोर्ड ऑफ़ रेवेन्यू को भेजा; चिट्ठी में वकील वहीद शहज़ाद बट का नाम दर्ज है। - आरोप के मुताबिक IRIS का कैलकुलेशन मॉड्यूल तीन विशेष बचत स्कीमों पर रियायती दर लागू नहीं करता या करदाता की श्रेणी ग़लत पकड़ता है। - प्रभावित श्रेणियाँ: विधवाएँ व सीनियर सिटीज़न, पेंशनभोगी, और शहीदों के परिवार। - कमेटी का अंकगणित अंदर से सुसंगत है, पर फ़ेडरल बोर्ड ऑफ़ रेवेन्यू ने इसे सत्यापित नहीं किया और कोई सार्वजनिक जवाब नहीं दिया। - स्रोत सामग्री में पत्र की तारीख़ दर्ज नहीं है; यह एक एडवोकेसी कमेटी का दस्तावेज़ है, स्वतंत्र जाँच नहीं। **स्रोत:** लाहौर टैक्स बार एसोसिएशन पब्लिक इंटरेस्ट लिटिगेशन कमेटी का फ़ेडरल बोर्ड ऑफ़ रेवेन्यू को पत्र (स्रोत सामग्री में प्रकाशन तिथि दर्ज नहीं); फ़ेडरल बोर्ड ऑफ़ रेवेन्यू की ओर से सार्वजनिक पुष्टि नहीं। **संबंधित प्रश्न-उत्तर:** प्रश्न: IRIS क्या है? उत्तर: IRIS पाकिस्तान के फ़ेडरल बोर्ड ऑफ़ रेवेन्यू का ऑनलाइन आयकर रिटर्न और ई-फाइलिंग प्लेटफ़ॉर्म है। प्रश्न: तीन प्रभावित स्कीमें कौन-सी हैं? उत्तर: बेहबूद सेविंग्स सर्टिफ़िकेट, पेंशनर्स बेनिफ़िट अकाउंट और शुहदा फ़ैमिली वेलफ़ेयर अकाउंट, जिन पर इनकम टैक्स ऑर्डिनेंस 2001 में रियायती विदहोल्डिंग दर तय है। प्रश्न: मामला कैसे सुलझ सकता है? उत्तर: IRIS रिलीज़ नोट, इन स्कीमों पर जारी नोटिसों की गिनती, वापस लिए गए नोटिस और रिफंड के औसत समय के सार्वजनिक होने से बहस तय हो सकती है।
There is a box on the form, and the box wants to know who you are. For a widow holding a Behbood Savings Certificate the answer is simple — she is a widow. But IRIS files her under the ordinary taxpayer slot, and then the concessional rate that should have applied does not apply, or the one that should never have applied does. The Public Interest Litigation Committee of the Lahore Tax Bar Association has written to the Federal Board of Revenue about precisely this box. The letter carries the name of the lawyer Waheed Shahzad Butt, refers to three schemes — Behbood Savings Certificates, the Pensioners' Benefit Account and the Shuhada Family Welfare Account — and makes one demand: fix the software error, because what the Income Tax Ordinance 2026 grants, the system is failing to deliver.
This is not a cricket story and I will not dress it as one. But forty-seven years have taught me one thing — the real information sits in the corridor, not the press conference. This week the corridor is a validation box.

Context: three schemes, one ordinance, one online form
IRIS is the Federal Board of Revenue's online return and e-filing platform. Returns have been filed through it for years, notices go out of it, refunds sit in it and occasionally escape. Once a revenue administration moves off paper onto a portal, every exchange between citizen and state takes the shape of a form.
These three schemes matter because they are not the ordinary taxpayer's story but a specific life's story. Behbood Savings Certificates are issued largely to widows and senior citizens. The Pensioners' Benefit Account is the post-retirement lifeline. The Shuhada Family Welfare Account belongs to the families of men who did not come home from security duty. Profit on all three attracts withholding, and the Income Tax Ordinance 2026 sets a concessional rate for these categories, distinct from the general one. The committee alleges that the IRIS calculation module does not apply those concessional rates, or misreads the filer's category altogether.
The source quality needs stating plainly. This is an advocacy committee's letter, not an independent investigation. Its arithmetic is detailed and internally consistent, but the FBR has not verified it and has issued no public reply. The source material does not record a date for the letter. Which means what I hold is an allegation carrying a set of numbers — and no answer to those numbers yet.
Core: the mistake is not in the rate but in the identity
The fault in IRIS is not in its arithmetic but in how it decides which category a taxpayer belongs to. Tax software does more than add and subtract. Its first job is to place a filer in a slot — salaried, business, property, or holder of one of these three special instruments. That is not a technical detail. It is the exact point where the concessional intention of the statute steps off the paper and reaches a human being, and a slip here makes every good intention upstream irrelevant.
There is something odd in the chain. The document that actually holds the truth — the bank and National Savings — is never asked; the person who does not hold the truth is asked everything. The filer self-declares in the form which kind of certificate he holds. The software takes that declaration as fact, and where its own categorisation does not match the declaration, the consequence lands in the taxpayer's account. This is the oldest pattern in every large system: evidence sits with the party that asks no questions, and the answer is demanded from the party that cannot assemble the evidence.

And the cost of an error is not symmetrical. If the system's default is wrong, the citizen applies for a refund — documents, queues, a lawyer, months, sometimes money he does not have that month. If the default is right, the department does nothing at all. One hundred per cent of a wrong default is borne by the taxpayer and none of it by the department. Which is why I do not read a default as a technical choice. What only one side pays for is not a choice; it is a policy.
Now the part that turns this letter from a complaint into data. Lawyers possess something no survey possesses — repetition. One lawyer sees a hundred unrelated clients with no relation to each other, different amounts, different banks, and the same failure recurring. For a reporter that is the most valuable signal there is, because an identical error across unconnected people belongs to the system, not to the people.
This is where a parallel from my own beat is worth bringing in, and not as colour. In April 2026, when the IPL was suspended, I reported on the roughly two hundred casual workers at Chinnaswamy in Bengaluru — groundstaff, scorers, camerapersons — whose KSCA contracts had been paused while the ground stood empty. The broadcast showed none of them. When the tournament finally ran in the UAE from September and I watched from a Bengaluru bedroom at six in the evening, the lesson was plain: everyone who carries a match to the ground stays outside the camera. The box on this tax form is the same man outside the camera. Eleven weeks on one bus taught me to stop reading the roster and start reading the men; today a form is telling me it is reading the man wrong.
One more thing belongs here, drawn from how I work. I cannot tell you the committee's numbers are wholly right. I can tell you there is exactly one way to prove them wrong, and it sits with the FBR. The party that owns the system also owns the evidence — and that inverted burden of proof is the severest test any transparent administration faces. Settling this needs no commission of inquiry. Four things suffice: the IRIS release notes for the past year, the count of notices issued against these three schemes, how many of those notices were withdrawn, and the average refund turnaround. Publish any one of those numbers and the argument ends.
And it is worth remembering whose money is stuck. Not large businesses. Widows, pensioners, the families of the martyred — the smallest balances and the heaviest paperwork. That is what administrative friction actually looks like: it collects less from the rich and more time from the poor. The fee was one point five five million; the corridor outside that room was free. The corridor outside this form is not.
Contrarian: the 'bug' the outside sees
From outside, this looks like a software glitch — a patch will come, the next release will fix it, case closed. My counter-read is that software does not make mistakes; it makes decisions. When a system defaults a taxpayer into the expensive slot, it does so not out of ignorance but because someone, somewhere, settled whose side to take when in doubt. The benefit of the doubt always runs to somebody, and here it runs to the department.
The second common reading is that lawyers have an interest, since disputes are their livelihood. That is true and should not be hidden — a tax bar exists to stand with clients in disputes. But motive and arithmetic are separate things. If a doctor says the hospital machine is faulty, we do not check whether patients pay his fee; we check whether the reading is right. Argue about motive if you like, but do not show the arithmetic out of the door while you do it.
The third reading sits quietest of all: digitisation did not remove discretion, it relocated it. Once you could ask the clerk — the man was in front of you, he had a name, you could ask again. Now the same decision lives in a configuration file that nobody signs. That is why this story cannot name and shame an individual, and why it is so easily left unreported.
Takeaway: what to watch
Three things. The next IRIS release note, and whether the categorisation for these three schemes changes in it. Whether notices already issued to these taxpayers are withdrawn, or each one has to be fought. And whether the committee receives a written reply, or the letter sits in a sealed envelope.
One question at the end, half of which I know and half of which I do not. The woman who ran a household on the faith of a Behbood certificate — which slot did her return finally land in? That answer will not arrive in a letter. It will arrive in a changelog, and changelogs so far have been written for the people below the line, not the people above it.

