Pakistan’s tax reform debate becomes least demanding at precisely the point where it should become most searching. When revenue disappoints, rearranging departments offers an apparently decisive response. Files move, reporting lines change and new offices acquire responsibility for old targets. The harder questions survive: which incomes escape effective taxation, which concessions remain protected, why compliant businesses face disproportionate demands, and whether the state can distinguish additional revenue from money merely collected earlier or refunded later.

The latest developments make this question immediate. Through notification No. 2525-IR-I/2026 of November 13, 2026, FBR created and filled the post of Member, Sales Tax and FED (Technical), alongside postings to the dedicated technical structure at headquarters.

Business Recorder’s report of November 14 also describes planned separate Sales Tax Zones in the field formations, with taxpayers expected to deal with separate sales-tax and income-tax commissioners from January 2027. The distinction between what has been notified and what remains to be implemented matters: the headquarters appointments are established, while the field reorganisation is a reported rollout. These developments concern the allocation of responsibilities within Inland Revenue; they should not be confused with the creation of an entirely separate civil service.

The proposed return to separate sales-tax and income-tax field structures is a regressive and potentially highly destructive reversal of integration. It makes the taxpayer navigate the administration’s internal divisions again, while creating opportunities for duplicated scrutiny, repeated demands for records and fragmented responsibility.

The same sale is relevant to output tax, turnover, business income and the verification of purchases; separating the officers responsible for these connected questions weakens the organisational basis for examining them together. Technical expertise should be strengthened through specialist teams within integrated administration.

Rebuilding parallel field hierarchies instead treats an implementation failure as a reason to dismantle integration itself. The damage need not await entirely separate databases: competing priorities, uncoordinated notices and divided accountability can reproduce silos even where officers belong to the same service and use the same portal.

Against this background, Salamat Ali’s “Lessons from FBR’s tax reforms”, published on May 27, 2019, advocates separate VAT administration within FBR. To present the November notification as a vindication of that prescription confuses official adoption with policy success.

An appointment order establishes who occupies a post; it does not establish that separation will close the compliance gap, improve audit quality or reduce taxpayers’ costs. The resemblance between the earlier recommendation and the present direction makes the recommendation more urgent to scrutinise, not more persuasive.

Nor does that resemblance establish that Salamat Ali caused the decision. The substantive objection is that the proposed field separation pushes interconnected tax functions back into silos without demonstrating why stronger specialisation within an integrated structure would be inadequate.

A comparison of revenue growth before and after an administrative change cannot, by itself, answer that question. Nominal receipts reflect prices, economic activity, exchange rates, imports, statutory rates, exemptions, refunds and enforcement.

A department collecting more rupees during rapid expansion may be performing less effectively than one collecting fewer additional rupees during stagnation. Unless these influences are separated, a historical comparison supplies a question for investigation, not a verdict on institutional competence.

Consider a purely illustrative example. If a taxable base grows by 20 per cent while collections rise by 18 per cent, revenue performance has weakened relative to the base. If the base grows by 8 per cent while collections rise by 12 per cent, performance has improved on that measure. Ranking the two administrations by their collection growth alone would reverse the conclusion. Even this comparison needs adjustment for changes in rates and coverage. Tax buoyancy, which includes policy effects, must not be confused with an estimate of administrative efficiency.

Likewise, projecting an earlier growth rate into subsequent years does not establish a recoverable revenue loss. Such a calculation assumes that the earlier trajectory would have continued despite changes in economic conditions and tax policy.

A credible counterfactual requires an explicit model, consistent data and sensitivity tests. For Pakistan, the underlying series should distinguish gross from net receipts, domestic from import-stage collection, and changes in the legal base from changes in compliance. Without those safeguards, an impressive monetary estimate can conceal an untested assumption.

The strongest contemporary challenge to an organisation-centred explanation comes from the World Bank’s May 2019 appraisal of Pakistan Raises Revenue. Its account of the earlier Tax Administration Reform Project identified unsupportive tax policy, wavering government commitment, inadequate initial ICT planning, frequent leadership changes, insufficient training and partial reversal of functional reorganisation.

These are materially different failures requiring materially different remedies. A reform only partly implemented cannot be treated as a clean experiment proving that its underlying design was wrong. This does not absolve either FBR or development partners. It makes their accountability more exacting.

Which promised capabilities became operational? Which processes were redesigned before software was purchased? Were staff trained and retained in the functions for which they had acquired expertise? Who authorised reversals, and what evidence supported them?

Counting buildings, computers and training sessions cannot establish that audit selection improved, evasion became harder or taxpayers received better service. Borrowed resources require an independently verifiable account of institutional improvement. The distinction between expertise and separation is fundamental. VAT needs specialists in invoice chains, input credits, refunds and fraud. Income taxation requires knowledge of accounts, deductions and profit allocation.

Customs needs expertise in valuation, classification and border risk. Recognising these differences does not require separate taxpayer identities, repeated information demands or disconnected audit programmes. A hospital can maintain specialist departments without making each department reconstruct the same patient’s entire medical history.

This was not an insight unavailable in 2019. The IMF’s 2010 toolkit for implementing a revenue authority expressly discussed duplication arising from separate VAT and income-tax departments, including multiple audits and disconnected collection action.

Its guidance on functional organisation explained the advantages of organising common administrative work across taxes. Neither document establishes that integration automatically succeeds. Both show why reversing it requires more than dissatisfaction with headline receipts.

Later guidance adds useful qualification. The VITARA discussion of organisational models recognises circumstances in which a tax-specific unit can provide focus and clear accountability, while warning about duplicated functions and higher compliance burdens.

The sensible policy question is how to preserve technical specialisation within shared administration. A specialist VAT risk team can work with common registration, data, taxpayer service and coordinated audits. A separate hierarchy that reproduces these functions must demonstrate benefits exceeding its additional costs. The economic character of VAT also deserves closer attention. Its defining feature is the relationship between tax on outputs and eligible credits on inputs. Comparing taxes simply because their initial calculation uses a percentage of a transaction value misses the importance of that relationship.

For an importing manufacturer, border payment and domestic liability belong to one connected chain. Administration should be designed around that chain and the taxpayer’s actual business, rather than around institutional ownership of a collection point.

Suppose an importer pays 100 units of creditable VAT at the border and subsequently owes 150 units on a domestic sale. Subject to the applicable rules, the additional domestic payment is 50. If border collection rises to 120 and output liability remains 150, domestic payment becomes 30. Total collection is still 150. This illustration does not explain every movement in Pakistan’s receipts; it demonstrates why a changing division between import and domestic collection cannot independently establish deterioration in domestic enforcement. Credit utilisation, refunds and the composition of imports must also be examined.

Refunds create another crucial test. An administration can improve its immediate cash position by postponing money lawfully owed to businesses. That may flatter reported collection while worsening liquidity and increasing financing costs.

The IMF’s 2021 guidance on VAT refunds explains how defective refund management can distort fiscal measurement and create arrears. A serious evaluation should publish refund claims, payments, ageing and risk outcomes alongside receipts. Legitimate refunds and fraudulent claims require different treatment; indiscriminate delay is not effective enforcement.

An equally serious omission from narrowly administrative prescriptions is the difference between a policy gap and a compliance gap. Revenue legally forgone through an exemption cannot simply be recovered by assigning a more energetic officer. Conversely, removing an exemption does not ensure collection where transactions remain concealed.

The IMF’s 2016 analysis of Pakistan linked weak revenue performance to narrow bases, concessions and administrative weaknesses. It also recorded an increase in GST exemptions from 0.1 per cent of GDP in 2012 to 0.8 per cent in 2015. These policy choices belong inside any explanation of collection performance.

The political question follows directly. Which concessions serve a demonstrable public purpose, and which preserve advantages that their beneficiaries could not defend in an open budget debate? Each should carry a published cost, an identified objective, a review date and evidence of results. Estimates of tax expenditure are not cash lying ready for collection: behaviour changes when relief is withdrawn, and some relief protects essential consumption or prevents inappropriate taxation. Nevertheless, transparent scrutiny of concessions reaches a source of fiscal weakness that rearranging offices cannot touch.

Also Read: Islamabad Hosts The Thirtieth Meeting Of The Economic Cooperation Organisation Council Of Ministers

The same care is necessary when discussing the tax base. Income-tax return filers, registered businesses, active sales-tax registrants and people bearing indirect taxes are different populations. Comparing one register with the entire population can obscure children, people below taxable thresholds and households already paying consumption taxes.

The useful questions concern eligible persons outside registration, taxable income omitted from returns and liabilities unpaid after assessment. Enforcement gains legitimacy when it identifies those obligations accurately, rather than treating every absent return as evidence of evasion.

The World Bank’s 2019 appraisal reported that withholding agents collected 60 per cent of income-tax receipts in FY2017–18. That fact complicates any account of institutional performance based on collections alone. Withholding can be appropriate and efficient, particularly where it closely tracks final liability. But transaction-based deductions can also impose costs on people with little or no final liability, while shifting administrative work to businesses. The policy challenge is to retain useful withholding, remove poorly targeted deductions and make credits and refunds accessible.

Fairness requires moving beyond the convenient taxpayer. The IMF’s 2017 discussion of provincial revenue distinguished progress in services taxation from weaker mobilisation through agriculture and real estate. Its institutional recommendations included coordination and consolidation of provincial collection responsibilities. This matters because expanding consumption taxation cannot substitute indefinitely for effective taxation of taxable incomes and property within the appropriate constitutional jurisdictions. A reform agenda should disclose who bears its additional burden, rather than assuming that every extra rupee represents an equitable improvement.

Federalism must be addressed with equal precision. Entry 49 of the Federal Legislative List distinguishes federal taxation of sales and purchases of goods from sales tax on services. Legislative competence, administrative collection and distribution of revenue are separate questions. A common collection platform need not extinguish provincial taxing authority.

Conversely, shifting constitutional taxing powers cannot be accomplished through an administrative order. Any proposed national arrangement must specify its lawful foundation, provincial consent where required, governance and revenue-settlement rules.

The practical cost of fragmented administration was already documented. The World Bank’s Fall 2019 South Asia Economic Focus described businesses facing up to 60 annual sales-tax returns across five jurisdictions, alongside conflicting bases and taxation principles. This is a historical description, not a claim that every business faces precisely that burden today. Its significance lies in the nature of the problem: taxpayers can be required to reconcile the state’s internal divisions at their own expense.

Subsequent developments reinforce the need to evaluate implementation. FBR’s December 2021 announcement of a single sales-tax return envisaged common filing and automated apportionment of adjustments and payments. A shared interface is useful, but its launch does not prove that substantive conflicts have disappeared. The appropriate assessment asks which taxpayers and transactions are covered, whether credits flow correctly, how disputes are resolved and whether compliance time actually falls. Digitising inconsistent obligations can leave their economic burden largely intact.

A workable alternative would begin with a jointly governed federal–provincial administrative framework. A National Revenue Agency could provide shared services under an agreed legal structure while preserving constitutionally assigned revenues and legislative powers. Its value would depend on common taxpayer records, interoperable data, coordinated risk assessment, predictable settlement and enforceable service standards. Its name would confer no special virtue. If it merely added another supervisory layer, it would reproduce the weakness it was created to remedy.

Such an arrangement should be developed through published diagnostics and staged implementation. Existing registrations and data should be reconciled before taxpayers are asked to supply them again.

Specialist staff should have credible career paths and adequate training. Audit selection should follow recorded risks, with human review and a fair opportunity to explain discrepancies. Digital invoicing should come with affordable access, correction mechanisms and protection against disruption. An automated mismatch is a reason to investigate, not conclusive proof of concealed income.

Accountability must also reach the use of coercive powers. Revenue targets should not encourage unsupported demands, premature recovery or pressure on businesses to advance payments. Performance reporting should distinguish amounts assessed from amounts sustainably recovered, disclose assessments overturned on appeal and measure the time taken to resolve disputes.

Independent adjudication, reasoned decisions and accessible remedies are part of revenue administration’s productive capacity. A system that repeatedly collects through error and returns money after prolonged litigation imposes costs that its revenue tables conceal.

Nor can tax reform be divorced from expenditure. Citizens are more likely to regard taxation as legitimate when public money is traceable, procurement is competitive and services improve. Waste does not cancel a lawful tax obligation, but it weakens the political case for asking already visible taxpayers to bear more.

A comprehensive fiscal reform programme should connect revenue mobilisation with expenditure transparency, scrutiny of subsidies and accountability for public assets. Collection is a means to public welfare, not an institutional achievement detached from it.

The November decision should be assessed as a reversal requiring justification, not celebrated as proof that an earlier prescription was correct. Its central weakness is the substitution of institutional separation for a demonstrated remedy to failures of policy, incentives, implementation and political protection. FBR should reconsider the separate field hierarchies and strengthen specialist capacity within integrated administration, with shared taxpayer records, coordinated audits and clear accountability.

Poor implementation of integration calls for correction of that implementation; it does not establish the superiority of fragmentation. Success must mean more lawful and sustainable revenue, fewer avoidable compliance demands and stronger taxpayer protections. A notification creating posts proves none of these outcomes. Rebuilding silos around the same taxpayer threatens to make an already burdensome system more costly and less coherent.

About the author

Dr. Ikramul Haq, Advocate Supreme Court, writer, literary critic, Adjunct Faculty at Lahore University of Management Sciences (LUMS), member Advisory Board and Visiting Senior Fellow of Pakistan Institute of Development Economics (PIDE), holds an LLD in tax laws. He was full-time journalist from 1979 to 1984 with Viewpoint and Dawn. He also served Civil Services of Pakistan from 1984 to 1996.