Beyond tax-to-GDP

Pakistan’s fiscal debate revolves around one apparently simple ratio: tax collection as a percentage of gross domestic product. Whenever the Federal Board of Revenue (FBR) misses its target, the country is reminded that its tax-to-GDP ratio is among the lowest in the region. Whenever collection increases in nominal terms, the same ratio is invoked to demand more taxation. Little attention is paid to the denominator: what does Pakistan’s GDP actually measure, and how closely does it correspond to the legally taxable economy?

This question has acquired fresh relevance after the release of the Finance Division’s Fiscal Operations for July–June 2025–26. The statement reports GDP at Rs. 126.870 trillion and FBR collection at Rs.13.010 trillion, producing an FBR-to-GDP ratio of about 10.3%.

That ratio accompanied official claims of record revenue and exceptional fiscal performance. The weaknesses in those claims—including borrowing, debt servicing and the difference between federal and consolidated deficits—have been examined separately. The denominator now requires scrutiny.

The Rs. 126.870 trillion figure is not the result of a census of economic transactions completed on June 30, 2026. It is the Pakistan Bureau of Statistics’ provisional nominal GDP estimate. PBS explains that provisional annual accounts published in May are based on six to nine months of information and projected to cover the full July–June year. They are revised when full-year information becomes available and finalised later.

The arithmetic is available in the official current-price table. For 2025–26, PBS estimated gross value added at basic prices at Rs. 117.991 trillion. It added Rs. 10.314 trillion of taxes on products and subtracted Rs. 1.434 trillion of subsidies. The result was GDP at market prices of Rs. 126.870 trillion.

This is internationally accepted national accounting, not manipulation. GDP, however, is a measure of production, not a schedule of taxable income. It contains market output, non-market output, production for own use and values for which no separate monetary transaction is observed. PBS itself defines “imputation” as constructing entries where no separate monetary transaction is identified but an economic flow is recognised.

The distinction becomes important with the financialisation of GDP. Dr Fahd Rehman, an independent economist, has drawn attention to the growing significance of finance, insurance and real estate—collectively called FIRE—and its weaker connection with employment and material production.

His broader concern is valid, although one technical point needs refinement. Financial intermediation services indirectly measured, or Financial Intermediation Services Indirectly Measured (FISIM), are not simply fee income. They estimate banking services supplied through the margin between lending and deposit rates where no explicit charge appears.

Real estate accounts contain another imputation: the rental value of owner-occupied dwellings. A homeowner is treated as providing housing services to himself or herself. This prevents comparisons between countries or periods from being distorted merely because one society has more owner-occupiers and another more tenants. It also adds output without generating an equivalent cash receipt or an automatically taxable income stream.

According to the PBS sectoral table, financial and insurance activities contributed Rs. 2.377 trillion in fiscal year (FY) 2025–26, while real estate and ownership of dwellings contributed Rs. 4.535 trillion. Together they represented about 5.45% of GDP at market prices. If both sectors were crudely removed, FBR collection would rise from 10.25% to approximately 10.85% of the remaining denominator.

It is also important not to attribute the latest GDP increase entirely to finance. Financial-sector value added fell from Rs. 3.669 trillion in FY 2023–24 to Rs. 2.145 trillion in FY 2024–25 before recovering to Rs. 2.377 trillion. Financialisation is a structural issue, not a one-year explanation. That calculation is only an illustration. Banks, insurers and real-estate businesses provide real services and include taxable entities.

Deleting their entire value added would be methodologically indefensible. The exercise demonstrates something narrower: the tax-to-GDP ratio can move because of the statistical composition of GDP even when cash tax collection remains unchanged. The exact effect cannot be known until PBS separately discloses FISIM, imputed rent and other non-cash components for the relevant year.

The informal economy creates a second complication. The PBS national-accounts glossary says the “non-observed economy” conceptually includes illegal, hidden, underground and informal activities. The word “conceptually” is decisive. Legitimate informal production is captured partially through household and small-enterprise surveys, agricultural estimates, commodity-flow methods, trade margins and extrapolations. These techniques estimate economic activity that never appears in an income-tax return or formal business account.

International rules also place some illegal production within the conceptual boundary of GDP. The UN handbook includes genuine production of goods or services exchanged voluntarily despite being unlawful, such as narcotics or smuggling services. Theft and extortion are not consensual transactions and do not constitute production. Pakistan’s published headline tables provide no separately quantified estimate of criminal activity. It would thus be incorrect to claim that Rs. 126.870 trillion comprehensively incorporates the criminal economy.

The policy conclusion is not that Pakistan should abandon GDP or construct a politically convenient smaller denominator. The conventional ratio remains useful for macroeconomic and cross-country comparison. It should not be confused with a direct measure of FBR efficiency or taxable capacity.

Pakistan needs a transparent bridge from GDP to the potential tax base. PBS and FBR should jointly identify market and non-market output, imputed rent, FISIM, government value added, informal-sector estimates, legally exempt income and sector-wise tax payments. Income-tax receipts should also be compared with taxable household and corporate income; sales tax with private consumption and its efficiency ratio; and property taxation with assessed land and property values.

Such disclosure will not absolve FBR. Its collection remains dominated by withholding, advance payments and indirect extraction from documented persons. Nor does the denominator explain why Rs. 6.948 trillion of interest consumed more than half of FBR receipts in FY 2025–26. It will, however, prevent a broad statistical aggregate from being presented as the precise measure of national tax compliance.

National accounting is designed to measure production. Taxation must mobilise actual capacity according to law, equity and democratic consent. Conflating the two allows governments to celebrate nominal “records”, demand more from those already captured and leave privilege untouched. Before prescribing another tax-to-GDP target, Pakistan must first disclose how much of its GDP is observed, imputed, informal, exempt—and realistically taxable.

_____________________________________________________________

Dr. Ikramul Haq, Advocate Supreme Court, was full-time journalist from 1979 to 1984 with Viewpoint and Dawn. He served Civil Services of Pakistan from 1984 to 1996.