Beyond Public Finance: Towards  Constitutional Political Economy of Pakistan—III  The Myth of Neutral Tax Policy

The second part of this series explained why public finance and Constitutional Political Economy (CPE) ask different questions. Public finance ordinarily examines what taxes governments should impose and how revenue should be spent. CPE asks who makes those choices, under what rules, for whose benefit and subject to what restraints. That distinction leads to an important conclusion: there is no politically neutral tax system.

Tax policy is often presented as a technical exercise. Economists compare direct and indirect taxes, estimate elasticity, calculate effective rates and recommend reforms intended to improve efficiency. Governments describe exemptions as incentives, withholding provisions as enforcement mechanisms and consumption taxes as instruments of broad-based revenue mobilisation. Such terminology creates the impression that taxation operates independently of political power. It does not.

Every tax decision identifies those who will pay, those who will collect, those who will receive concessions and those whose activities will remain beyond effective enforcement. A tax may be neutral between two products in an economic model, but the process through which it is enacted and administered can rarely be neutral between organised interests, social classes or political constituencies.

The Organisation of Economic Cooperation and Development (OECD) itself recognises that taxes affect taxpayers differently according to their income and other socio-economic characteristics. They alter behaviour and influence the distribution of income both directly and through the public expenditure they finance. Taxation is, thus, not merely a device for transferring money to the treasury. It changes economic opportunities and affects the relationship between citizen and state.

The celebrated Mirrlees Review sought to design a coherent tax system in which similar activities were treated consistently, and economic choices were not distorted without good reason. This is a valuable objective. Neutrality can reduce arbitrary discrimination and prevent tax considerations from dominating productive decisions. Nevertheless, even the most carefully designed system must decide which activities are alike, which differences justify special treatment and how equity should be balanced against efficiency. Those choices necessarily embody judgments about society.

The idea of neutrality becomes more problematic when it is transferred from theory to a state characterised by unequal political influence. Consider a general sales tax (GST). In theory, a broad-based value added tax (VAT) imposed at a uniform rate minimises distortions and preserves the chain of documentation. In practice, exemptions, reduced rates, special schedules, fixed taxes, withholding taxes (unique in Pakistan even for VAT/GST!) and sector-specific arrangements are introduced during the political process. The final statute may bear little resemblance to the neutral instrument initially proposed.

The same is true of income tax. Horizontal equity requires persons with similar ability to pay to bear comparable burdens. Vertical equity requires those possessing greater capacity to contribute more. In Pakistan, however, the legal character assigned to income often determines the burden more decisively than the taxpayer’s actual economic capacity. Salary, business income, capital gains, dividends, property income and agricultural income may all be subjected to different regimes, rates or jurisdictions.

These distinctions are not always indefensible. Different types of income may require different collection methods. The constitutional distribution of taxing powers must also be respected. However, a CPE analysis asks why particular differences survive, who benefits from them and whether their stated rationale corresponds to their actual effect.

Tax exemptions provide the clearest illustration. Governments describe them as instruments for attracting investment, supporting industries, protecting vulnerable groups or promoting exports. Some concessions may serve legitimate public purposes. Others constitute expenditure conducted through the tax system without the scrutiny ordinarily applied to direct spending.

Pakistan’s official Tax Expenditure Report 2026 estimated revenue forgone during fiscal year 2024–25 through income tax, sales tax and customs concessions at approximately Rs. 2.353 trillion (excluding sales tax on POL products to deprive provinces of their constitutional right, replacing it with petroleum levy). Of this amount, sales tax concessions accounted for about Rs. 1.274 trillion, income tax concessions for Rs. 579.70 billion and customs concessions for Rs. 499.14 billion. These are not accounting curiosities. They represent choices about which persons, sectors and transactions receive preferential treatment, and which taxpayers must bear the resulting revenue burden.

A concession granted through the tax law is economically similar to public expenditure. If the state collects Rs. 100 from a citizen and transfers it to an industry, the transaction appears in the budget. If the state allows that industry to retain Rs. 100 that would otherwise have been payable, the distributive effect may be comparable, but the benefit is less visible.

This opacity is politically useful. Direct subsidies attract public attention and legislative scrutiny. Tax concessions are buried in schedules, exemptions and statutory notifications. Their beneficiaries are often concentrated and organised, while the cost is dispersed across millions of taxpayers and consumers.

Constitutional Political Economy explains why such arrangements persist. A concentrated group has a strong incentive to lobby for a benefit worth billions of rupees. Each member of the general public, bearing only a fraction of the cost, has little incentive or capacity to oppose it. What appears to be an anomaly in tax design may be the predictable result of unequal political organisation.

Withholding taxation presents another example. It is defended as an efficient method of collecting revenue from an economy with weak compliance. In limited circumstances, deduction/collection at source is entirely justified. Salary taxation (pay roll taxes) and payments to non-residents commonly require withholding systems in many jurisdictions.

Pakistan, however, has transformed withholding from a collection technique into a parallel tax regime. According to the Revenue Division Year Book 2024–25, withholding taxes contributed 60 percent of total income tax collection in that year. Collection through withholding reached approximately Rs. 3.382 trillion. This reliance changes the institutional character of income taxation.

The tax administration increasingly obtains revenue from transactions rather than determining taxpayers’ actual net income and ability to pay. Banks, employers, utilities, property registrars, businesses and other intermediaries become unpaid tax collectors. Persons already operating within the documented economy bear recurring deductions/collections in advance, compliance costs and the burden of seeking adjustments or refunds (hardly allowed automatically in our context unless palms are greased or intervention of Federal Tax Ombudsman is sought).

Meanwhile, the state may report increased income tax collection even though much of it resembles transaction-based taxation and advances received for next tax year (cumulative around 90 percent). The label remains “direct tax”, but the economic burden may be shifted, capitalised or passed on in prices (increasing incidence on the marganilised segments of society). Classification can therefore conceal as much as it reveals.

The distinction between filers and non-filers illustrates the same problem. Higher rates for persons outside the Active Taxpayers’ List are presented as incentives for documentation. In reality, many such deductions/collections operate as revenue measures imposed on transactions, without establishing the taxpayer’s income or final liability (in most of the cases enabling the passing on of the burden to end consumers). The state may collect more (in billions) from non-compliance without necessarily eliminating it.

A system can consequently develop a fiscal interest in preserving the very informality it claims to combat. If non-filers generate higher withholding receipts, the administration may prefer continuing differential rates to undertaking the difficult institutional work of assessment, enforcement and expansion of the genuine tax base. This is a classic CPE problem: the incentives of the collecting institution may differ from the declared objective of tax policy.

The same principle applies to indirect taxation. A consumption tax may be described as neutral because it applies at a uniform statutory rate. However, its social effect depends on what is taxed, what is exempt, whether input tax is allowed, how refunds operate and how households at different income levels spend their resources.

A tax on essential consumption absorbs a larger share of the income of a poor household than of a wealthy household. An exemption may provide relief, but it may also break the value added tax chain and benefit producers more than consumers. A reduced rate may appear compassionate while generating classification disputes, evasion and lobbying.

There is no escape from choice. The relevant question is not whether tax policy can be made entirely neutral, but whether its distributive choices are transparent, constitutionally authorised and publicly defensible. This insight was not discovered exclusively by modern Western economists.

In the eighth century, Abu Yusuf, in Kitab al-Kharaj, treated taxation as a question of justice, administrative responsibility and protection against oppression. He emphasised the conduct of revenue officials and the need to preserve productive capacity rather than maximising immediate extraction. His work demonstrates that fiscal policy has long been understood as part of the moral and institutional relationship between ruler and citizen.

The underlying principles are universal: public burdens should be just; officials should be accountable; property should not be appropriated arbitrarily; wealth should circulate rather than remain concentrated through privilege; and revenue should support collective welfare, particularly the protection of vulnerable citizens. These principles can be expressed in constitutional language without converting fiscal analysis into theological argument.

Pakistan’s tax debate must move beyond the recurring contest between higher rates and lower rates, direct and indirect taxes, or enforcement and incentives. Before recommending any measure, we must ask four questions. Who bears its real burden? Who receives the benefit? Who possesses the power to secure an exception? Who is accountable when the measure is applied selectively?

A tax system should be judged not merely by how much it collects but by the constitutional and political relationship it creates. A system that raises revenue by repeatedly burdening those already documented, while accommodating those possessing influence, may succeed administratively and fail constitutionally.

Tax policy is never neutral. It either restrains privilege or protects it; broadens citizenship or deepens exclusion; strengthens consent or normalises coercion.

In the next part, we shall examine why technically sound tax reforms fail, and how political incentives alter reforms between their conception, enactment and enforcement.

[To be continued]

____________________________________________________________________

Dr. Ikramul Haq, Advocate Supreme Court, 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.