The Income-tax Act 1961 is the statute through which the Indian state makes its largest single claim on private resources, and it is also the statute that Parliament has altered more often than any other on the book. Every other Indian Act can be read, with effort, as a document that was enacted on a date and then amended from time to time. This one cannot. It has been amended in substance by the Finance Act of almost every year since it commenced, so that the text a reader opens is a sediment of more than six decades of budgets, each layer laid down for the fiscal and political reasons of its own year and rarely removed when those reasons expired. To understand the Act is therefore to understand two things at once: the machine its sections build for charging, computing, collecting and disputing a tax on income, and the annual process that kept rebuilding that machine while it was running.

The position a reader must fix first is the succession. The Income-tax Act, 2025, Act 30 of 2025, received the assent of the President on 21 August 2025 and operates from 1 April 2026, replacing the 1961 Act for the tax years that begin on or after that date. The repeal does not erase the older statute. Income earned in any period up to 31 March 2026 is charged, assessed, appealed and penalised under the 1961 Act, and every proceeding relating to those periods continues under it.

The scale of the statute’s operation should be held alongside its structure. Income tax is the largest single head of the Union’s direct tax revenue, and the department that administers this Act deals with the population of the country’s formal economy: returns filed in crores each year, deductions at source recorded against practically every salary, interest payment and professional fee in the organised sector, and an information system into which banks, registrars, exchanges and employers continuously report. A statute of that reach is not merely a legal text. It is a piece of national infrastructure, closer in its daily significance to the payments system than to an ordinary regulatory code, and the consequences of its design choices, a definition widened, a threshold frozen, a procedure made faceless, are felt by millions of people who will never read a section of it. That is the standard against which the pillar judges it: a code of this reach owes its subjects legibility as a matter of governance and not merely of drafting taste, and its failure to remain legible is the substantive charge in the case against it. For a generation of assessments, appeals and prosecutions, the 1961 Act remains the governing law, and it will be cited in tribunals and courts for decades after its replacement because the disputes of a tax year take years to die. This pillar therefore treats the 1961 Act neither as history nor as a rates card. It treats it as a working statute whose structure a reader may need at any time, and whose design explains both its own replacement and the shape of the statute that replaced it.

What stood before: the 1922 Act and the making of the 1961 statute

Income tax in India is older than the republic. A tax on income was first levied in 1860, and the first comprehensive code was the Income-tax Act, 1922, enacted under British rule and carried into independent India with the Constitution’s continuance of existing law. The 1922 Act governed for four decades, and by the 1950s it was widely regarded inside and outside government as a statute that amendment had overtaken. It had been passed for a smaller economy, its language belonged to an earlier drafting era, and decades of piecemeal change had left it difficult to navigate even for the department that administered it. The case for replacement was not that its policy was disgraced. It was that its text no longer served the administrators, the courts or the taxpayers who had to use it, a rationale that will recur in this article, because the 1961 Act eventually died of the same condition.

The path to the 1922 code itself had been incremental. The levy of 1860, introduced by James Wilson, the first finance member of the Viceroy’s council, to repair the finances strained by the revolt of 1857, lapsed and returned more than once before a permanent income tax was placed on the book in 1886. An Act of 1918 revised the structure during the First World War, and the Act of 1922 consolidated the law into the code independent India inherited, influenced by the English model and administered through a central board created in its wake. The 1922 Act was itself substantially amended, most heavily in 1939, and by the time the republic’s first governments reviewed it, administrators and taxpayers shared the view that a modernising economy needed a code drafted as a whole rather than a consolidation patched for forty years. The importance of that pre-history for the 1961 Act is that its draftsmen were the second generation to attempt the same cure, and the failure mode they were warned against, accretion defeating arrangement, was documented in their own source material.

The drafting basis for the replacement was laid by two exercises of the late 1950s. The Law Commission took up the revision of the income tax law and reported in 1958, examining the structure of the 1922 Act and the case law that had accumulated around it. Running alongside it, the Direct Taxes Administration Enquiry Committee, chaired by Mahavir Tyagi and appointed in 1958, reported in 1959 on the administration of the direct taxes, the machinery of assessment and collection, and the grievances of taxpayers. Between them, the Commission supplied the legal architecture and the Tyagi Committee supplied the administrative critique. The Income-tax Bill built on both was introduced, examined and enacted as the Income-tax Act, 1961, Act 43 of 1961, which received assent on 13 September 1961 and came into force on 1 April 1962, the first day of the financial year 1962-63. The 1922 Act was repealed, with the usual savings for proceedings relating to earlier years, establishing the pattern that its successor’s repeal would follow sixty-four years later.

The new code extended to the whole of India from its first day, including the territories where its predecessor’s application had historically required separate handling, and that territorial completeness is easy to overlook because it was never controversial: a single national market in goods, services and employment was the premise of the republic’s fiscal design, and the income tax was its principal instrument. The Act’s commencement structure was equally simple, the whole code operating from one date, a simplicity later tax legislation abandoned in favour of staged notifications, and the contrast between the clean commencement of 1962 and the staggered commencements of the codes that followed it, including its successor’s single date deliberately restored in 2026, is a small lesson in what the draftsmen of that era assumed a statute was: a complete instrument that began at once and was then maintained, rather than a framework switched on in pieces.

The aspiration recorded for the new statute was a clear and compact code. The Act as enacted contained 298 numbered sections arranged in twenty-three chapters, running from the preliminary definitions through the basis of charge, the computation of income under its heads, the machinery of assessment and collection, and the chapters on appeals, penalties and offences. It was not a short statute, because a national income tax cannot be short, but it was a legible one, drafted in the belief that orderliness was achievable and could be preserved. That belief is the poignant fact of the Act’s history. Nothing in the 1961 text was careless. What overwhelmed it was not drafting failure but the amendment mechanism described later in this article, which guaranteed that clarity would be consumed year by year. Readers who keep that distinction in mind will judge both the 1961 Act and its 2025 replacement more accurately than readers who treat complexity as a drafting sin rather than a structural product.

The intellectual climate of the drafting deserves a sentence, because the Act’s assumptions about the taxpayer came from it. The economist Nicholas Kaldor had surveyed Indian taxation for the government in 1956 and recommended a wider, more integrated direct tax system, including taxes on wealth, expenditure and gifts that Parliament enacted in the following years as companion statutes, most of them later repealed. The 1961 Act was thus born as the centrepiece of a family of direct taxes administered together, and it outlived its siblings, absorbing functions and expectations that the family was designed to distribute. When later committees, the Direct Taxes Enquiry Committee under Justice Wanchoo in 1971 on the causes of evasion and the tax arrears of that era, the Chelliah Committee on tax reforms of 1991 to 1993, and the task forces of the following decades, proposed rebuilding the structure, each was in part proposing to recover by design what the family’s collapse had left the income tax carrying alone.

The 1961 Act also inherited and restated doctrines the courts had built under the 1922 Act, and its early decades were shaped by litigation over language carried across from the older statute. Where the new Act repeated the words of the old, the courts treated the settled interpretations as travelling with the words, a continuity principle that matters again at the 2026 succession, when courts will have to decide how much of the 1961 Act’s enormous case law travels into a successor that deliberately uses different words for the same concepts. The draftsmen of 1961 simplified the arrangement; they did not pretend the past could be discarded, and the Act’s structure, examined next, is best read as a reorganisation of a living body of law rather than a fresh invention.

The architecture of the Income-tax Act 1961

The Act is arranged as a sequence that follows the life of a tax liability. It opens with the preliminary chapter, which fixes the short title, the territorial extent across the whole of India, and the definitions in section 2 on which everything else depends, including the definitions of income, person, assessee, previous year and assessment year. It then states the basis of charge: the charging provision, the scope of total income, the rules of residence that decide whose worldwide income is caught, and the catalogue of incomes that do not form part of total income at all. Next comes the computational heart, the chapter that classifies income under five heads and gives each head its own code of computation, followed by the chapters on income of other persons included in an assessee’s total income, on the aggregation of income and the set-off and carry-forward of losses, and on the deductions allowed in computing total income. A long series of chapters then deals with special provisions for classes of cases and classes of taxpayers, before the statute turns to machinery: the procedure for assessment, the collection and recovery of tax, refunds, appeals and revisions, penalties, and offences and prosecution. The arrangement is logical enough that a reader who knows where a question sits in the life of a liability can usually guess the chapter that answers it, which is precisely the quality that the annual accretion of inserted chapters and lettered sections gradually obscured.

How many sections does the Income-tax Act 1961 really have?

The Act as enacted in 1961 contained 298 numbered sections, and that figure is still quoted as the Act’s size, but it stopped describing the real statute long ago. Because amendment proceeded by inserting lettered sections rather than renumbering, sections such as the long run from 115A onwards multiplied inside the numbering.

The Finance Ministry’s own comparison, published with the replacement exercise, counts 819 effective sections spread across forty-seven chapters, fourteen schedules and a text of about 5.12 lakh words. Both figures are accurate answers to different questions. The first counts the numbers Parliament printed in 1961. The second counts the provisions a practitioner must actually navigate.

That numerical puzzle is not a curiosity. It is the visible form of the Act’s central design fact. A statute that grows by insertion, because renumbering would break every cross-reference in the department’s forms, the courts’ judgments and the profession’s memory, will always present two sizes: the clean figure on its face and the sedimented figure in use. The 819 count includes whole inserted chapters on subjects the 1961 draftsmen never contemplated, from transfer pricing to special regimes for classes of companies and funds, each inserted at the point in the numbering where space could be found rather than where logic would place it. A reader searching the Act for the rule on a modern question therefore has to know not only the rule but the decade in which the rule arrived, because the decade determines the neighbourhood of the numbering in which it hides. The replacement statute’s decision to renumber everything, accepting the disruption the 1961 method avoided, is the clearest evidence of how heavily that method’s costs had come to weigh.

The definitions chapter deserves separate notice because so much litigation is definition litigation. Section 2 defines income inclusively, so the list it gives illustrates rather than exhausts the concept, and the courts have spent decades deciding which receipts possess the character of income and which are capital in nature and outside the charge unless a specific provision pulls them in. The definition of person is enumerative and wide, covering an individual, a Hindu undivided family, a company, a firm, a limited liability partnership, an association of persons and body of individuals, a local authority and every artificial juridical person not falling elsewhere, which is why entities unknown to 1961 have been absorbed without structural amendment. The definition of assessee extends beyond the person liable for his own tax to persons liable for the tax of another and persons against whom proceedings are pending, a breadth that explains why the machinery chapters speak of assessees where a lay reader would say taxpayers. A statute’s definitions are its load-bearing walls, and the 1961 Act’s walls were built generously, which is one reason the building could be extended so often without being rebuilt.

Two features of the arrangement complete the map. The first is delegated legislation. The Act is worked through the Income-tax Rules framed by the Board under the rule-making power in section 295, together with forms, notifications and circulars, and a significant share of the law a practitioner applies on any given day sits in that subordinate layer rather than in the sections themselves, a dependence that grows as schemes such as faceless assessment are structured by notification. The second is the schedules. The Act carries fourteen schedules, several of them machinery of real weight: the First Schedule prescribes the computation for insurance business, and the Second Schedule prescribes the entire procedure by which a Tax Recovery Officer recovers arrears by attachment and sale, so that a reader who stops at the sections has not read the whole statute. A pillar account records both features because the criticism that the Act is unreadable is partly a criticism of a text whose operative content is distributed across sections, rules, forms and schedules that no single volume keeps in step.

The constitutional settlement beneath the Act

The Act’s existence and its boundaries are fixed by the Constitution, and three provisions of the Seventh Schedule settlement explain features of the code that otherwise look arbitrary. Entry 82 of the Union List gives Parliament the power to tax income other than agricultural income, and Entry 46 of the State List gives the state legislatures the corresponding power over agricultural income, which is why the Act’s charge stops at the farm gate and why section 10(1) reads as it does. The Constitution itself, in Article 366, defines agricultural income by reference to the law of income tax, so the boundary between the two legislatures is drawn by a definition inside the Union’s own statute, a circular arrangement that the courts have managed for seventy years and that gives the definition in section 2 a constitutional weight no other definition in the Act carries. States have used their Entry 46 power sparingly, chiefly on plantation income, so that in practice agricultural income sits largely untaxed by either level, a policy outcome produced by the distribution of powers rather than by any single legislature’s choice, and periodically criticised on exactly that ground.

The second constitutional feature is sharing. Income tax is not the Union’s exclusive revenue in yield as it is in legislation. Under Article 270, taxes on income other than agricultural income form part of the divisible pool shared with the states on the recommendation of the Finance Commission, which is why every state government has a fiscal interest in the Union’s income tax law and why the rates and base decisions taken in the Finance Bill reverberate through state budgets that Parliament does not vote. The surcharge the Union levies for its own purposes stands outside the divisible pool, a distinction that has grown in importance as cesses and surcharges have formed a rising share of the gross collection, reducing the proportion states receive and generating a federal grievance that attaches to this Act’s Finance Act amendments though it is created by the constitutional text. The third feature is the one already examined at the mechanism level: the whole of this law moves through Money Bills under Article 110, so the constitutional design that shares the proceeds between two levels of government concentrates the law-making in one House of one legislature. A reader who holds the three features together sees the Act whole, a Union tax on a Union definition, shared with states that cannot shape it, amended annually by a procedure that subordinates deliberation to supply.

The charge, the scope and the person taxed

What does the Income-tax Act 1961 actually charge tax on?

Section 4 is the charging provision, and its structure explains the whole statute. It provides that income tax shall be charged for every assessment year at the rates prescribed by the Finance Act of that year, in respect of the total income of the previous year of every person.

Three consequences follow from that single design. The charge is annual, so rates are not in the Act itself but in the First Schedule to each year’s Finance Act. The charge falls on the total income of a defined year, which is why the previous year and assessment year distinction organises all compliance. And the charge falls on persons as the Act defines them, which is why residence and the scope rules that follow decide the real reach of the tax.

One mechanical definition does daily work throughout the Act and belongs to its foundations. The previous year, defined in section 3, is the financial year in which income is earned, and since the reforms that took effect from assessment year 1989-90 it has been the uniform financial year for every assessee, ending the older liberty under which businesses closed their accounts on divergent dates and the department administered a calendar of staggered years. Uniformity of the previous year is an invisible precondition of everything described later in this article: of withholding matched to a common year, of a single filing season, of information statements that aggregate a person’s dealings across payers, and of the assessment year in which the whole country’s income of one period is determined together. The 2025 Act’s celebrated single tax year is, on inspection, a renaming of this settlement rather than a new idea, which does not diminish it: it extends the uniformity of 1989 from the accounting period to the vocabulary in which the period is discussed.

Section 5 then fixes the scope of total income, and it does so by residence. For a person resident in India, total income includes income received or deemed to be received in India, income accruing or arising or deemed to accrue or arise in India, and income accruing or arising outside India. For a person not ordinarily resident, foreign income is included only if it accrues from a business controlled in India or a profession set up in India. For a non-resident, only the Indian-source categories are included. Section 6 supplies the residence tests, built principally on physical presence counted in days across the previous year and the years before it, with separate rules for individuals, companies and other persons, and section 9 deems specified categories of income to accrue or arise in India, including income from business connections, property, assets and shares situated in India, salaries earned for services rendered in India, and dividends paid by Indian companies. The deeming provision is where the Act’s territorial ambition lives, and its amendment is where the famous retrospective controversy arose, which belongs to the article on that episode and is linked at the anti-avoidance discussion later in this pillar.

Beneath the scope rules lies the distinction on which the whole charge rests, between capital and revenue receipts. The Act taxes income, and a receipt that is capital in nature, the price of an asset sold, compensation for the loss of the source itself, a capital contribution, is not income and falls outside section 4 unless a charging provision, the capital gains head being the great example, expressly draws it in. The courts developed the tests for the distinction over the life of the 1922 Act and applied them under this one, asking whether the receipt replaces a source of income or flows from it, whether it is a fixed or circulating asset in the recipient’s hands, and what the parties’ documents and conduct show. The distinction is not a technicality at the margin of the Act. It is the boundary of the tax, and the deeming provisions and the specific charging sections that pierce it, on capital gains, on specified premiums and perquisites, on gifts above a threshold received without consideration, are best understood as deliberate parliamentary crossings of a boundary the courts would otherwise police. Drafting that enlarges the base proceeds by exactly such crossings, year by year, which is one more route by which the text accumulated.

The gift provisions are the cleanest modern example of a crossing. For decades a gift of money received without consideration was a capital receipt in the recipient’s hands and outside the charge, a gap through which income could be routed as generosity. Parliament closed it in stages, first for money, then for immovable property and specified movable property, and the provision settled in section 56(2)(x), recast by the Finance Act of 2017, under which money or property received without consideration, or for inadequate consideration, beyond a threshold of fifty thousand rupees is taxed as income from other sources in the recipient’s hands, with exceptions for relatives, marriage, inheritance and other specified occasions. The provision sits in the residuary head rather than in any anti-avoidance chapter, which is the point: a device that began as avoidance was answered not by a doctrine but by redefining the base, and the redefinition then applied to every genuine gift as well, subject to its exceptions. Base-broadening by crossing is the Act’s characteristic legislative act, and a reader who can spot the crossings can date the anxieties of successive decades from the text alone.

Between the charge and the computation sits the chapter of incomes that do not form part of total income, of which section 10 is the working core. Its best known clause excludes agricultural income, a consequence of the constitutional distribution of taxing power, under which taxes on agricultural income fall to the states while Parliament taxes income other than agricultural income. The exclusion is definitional as much as concessional: agricultural income is defined in section 2 and carved out of the Union charge by section 10(1), though for individuals and certain other assessees it is taken into account for determining the rate applied to the non-agricultural remainder, a device called partial integration that has operated since the 1970s. Around that central exclusion, section 10 gathers dozens of others, for categories of receipts and categories of recipients, from specified allowances to the incomes of particular institutions. The volume of section 10 is itself a lesson in the Act’s growth, because every Parliament that wished to favour a receipt found it easier to add a clause to section 10 than to restructure the charge.

Which taxpayers does the Income-tax Act 1961 bind, and where?

The Act extends to the whole of India and binds every person whose income falls within the scope rules, whether that person is a citizen, a resident foreign national, a company incorporated in India, a company incorporated abroad with income arising here, a firm, a trust, or a local authority. Its grip follows income and residence rather than nationality.

A citizen living abroad with no Indian-source income is outside the charge. A foreign company with no presence in India but income deemed to accrue here through an Indian asset is inside it. The Act also binds the machinery of the state itself in a second sense, because the department that administers it, headed by the Central Board of Direct Taxes under the Ministry of Finance, draws every one of its powers, from calling for information to searching premises, from the Act’s own text.

The Board’s place in the structure is worth fixing because readers habitually confuse the statute with its administrator. The Income-tax Act creates the liability and the machinery. The Central Board of Direct Taxes, a statutory board constituted under separate legislation of 1963, directs and superintends the department that works the machinery, and it holds powers the Act itself confers, including the power to issue circulars binding on the department and, historically, a contested space between interpreting the Act and softening it. Courts have held that the Board’s circulars bind the officers who administer the Act but cannot override the Act itself, a distinction that has generated its own shelf of litigation. Below the Board, the Act’s later chapters create the hierarchy of authorities, from the Assessing Officer who makes assessments to the Commissioners who supervise and revise, and the appellate authorities described later in this article. The statute is thus both a tax code and an administrative charter, and many of its longest chapters are about the powers and duties of the state’s own officers rather than the liabilities of taxpayers.

The five heads of income and why classification decides everything

Section 14 commands that, save as otherwise provided, all income shall be classified under five heads: salaries; income from house property; profits and gains of business or profession; capital gains; and income from other sources. The command looks like a filing convention and is in fact the Act’s most consequential structural decision, because the heads are not labels but computational codes. Each head has its own charging section, its own rules for what counts as a receipt under it, its own list of permissible deductions, its own treatment of losses and its own rate consequences. The same rupee received by the same person can bear different tax depending on the head under which it is assessed, and a receipt that fits no head and no charging provision escapes tax altogether, however income-like it appears. Classification is therefore the first question in every assessment, the first dispute in most litigation, and the organising idea of the table that follows.

Two corollaries discipline the scheme and are worth stating before the detail. The heads are mutually exclusive in the sense that a receipt properly chargeable under a specific head is not charged under the residuary one, and where Parliament has provided a specific computational code for a receipt, that code exhausts the treatment, deductions included: a deduction not found in the head’s own list is not imported from another head’s generosity. And the heads presuppose a source. Income for this Act is not any accretion of wealth but a flow from a source, employment, property, business, asset or the residuary sources the fifth head gathers, and receipts that cannot be related to any source and fall outside every charging and deeming provision remain the hardest cases at the Act’s edge, argued over unexplained credits and gifts in kind. The scheme’s elegance is real. It is a periodic table of the ways money reaches a person, built in 1961 and extended rather than redesigned as new ways appeared.

Head of income	The receipts it captures	The basis of computation	Principal deductions available	The commonest classification dispute
Salaries	Pay, wages, annuity, pension, gratuity, fees, commission, perquisites and profits in lieu of salary received from an employer	Charged on salary due or received, with the employer-employee relationship as the gateway condition	A standard deduction from salary, taxes on employment actually paid, and the exemptions housed in section 10 for specified allowances and retirement receipts	Whether the relationship is employment at all, or a contract for services taxable as business income, which decides withholding, deductions and the head itself
Income from house property	Rent and the annual value of buildings and lands appurtenant owned by the assessee	Charged on annual value, meaning the letting potential of the property, rather than only on rent actually received, with municipal taxes paid allowed in arriving at it	A flat thirty per cent of annual value in place of itemised repair and collection costs, plus interest on borrowed capital used for the property	Whether a letting is the exploitation of property by its owner or a business of letting, which moves the receipt between this head and profits and gains
Profits and gains of business or profession	The profits of any trade, commerce, manufacture, profession or vocation carried on by the assessee	Charged on profits computed under the ordinary accounting and statutory rules, on the broad base of receipts less the expenditure the Act allows	Rent, repairs, depreciation, interest, insurance and the general expenditure deduction for outgoings laid out wholly and exclusively for the business, subject to statutory disallowances	Whether an outlay is revenue expenditure deductible at once or capital expenditure locked out of the general deduction, the oldest argument in the Act
Capital gains	Profit on the transfer of a capital asset, from shares and securities to land, buildings and goodwill	Charged on the difference between the full value of consideration and the cost of acquisition and improvement, adjusted by the Act’s indexation and holding rules as they stand from time to time	Costs of transfer and acquisition, indexed cost where the Act allows it, and the reinvestment exemptions that defer or discharge the charge on specified substitutions	Whether an asset held is a capital asset or stock in trade, which decides whether the gain is a capital gain at special rates or a business profit at ordinary rates
Income from other sources	Interest, dividends and every residue of income that no earlier head captures and no provision excludes	Charged on the residuary principle, sweeping in receipts of an income character that fail the tests of the four specific heads	Expenditure laid out wholly and exclusively to earn that income, other than capital expenditure and personal expenses	Whether a receipt is income at all, or a capital receipt outside the charge, since the residuary head tempts the department to treat every unexplained credit as taxable

Salaries, dealt with in sections 15 to 17, is the head most citizens meet. Its charge turns on the relationship of employer and employee rather than on the nature of the payment, which is why the boundary between salary and business income is litigated by consultants, retainers and professionals whose engagements are drafted to sit on one side or the other. The head defines salary inclusively to catch perquisites and profits in lieu, values many perquisites by rule rather than by market price, and delivers most of its concessions through the exemptions in section 10 and the standard deduction, a flat amount allowed in place of the itemised employee expenses that earlier versions of the Act struggled to verify. The machinery of withholding at source, examined later, is attached principally to this head, so for the salaried majority the Act is experienced almost entirely as a deduction from monthly pay and an annual reconciliation.

Section 17 defines salary with deliberate breadth, sweeping in wages, annuity and pension, gratuity, fees and commission payable by an employer, advances, amounts from provident funds beyond the employee’s own contribution, and the aggregate of perquisites, the value of benefits provided by reason of employment, from rent-free accommodation and cars to interest-free loans and club membership. Because perquisites resist market valuation in an assessment, the Rules value them by formula, and the formulas, rent-free accommodation computed as a percentage of salary with caps and adjustments being the standard example, decide the real income of senior employees more than the charging words do. Retirement receipts are handled by a mixture of charge and relief: pension is salary, commuted pension is partly exempt under section 10, gratuity is exempt within limits, and leave encashment carries its own exemption, so that the end of a career is taxed through a series of negotiated exceptions accumulated over decades of representation. The head is also where the Act’s sensitivity to the form of engagement is sharpest, because every rupee of tax, withholding and deduction turns on whether the payer is an employer or a client, and a large advisory practice exists to draft engagements onto the preferred side of that line.

Income from house property, in sections 22 to 27, is the Act’s most conceptual head, because it charges not rent received but annual value, the sum for which the property might reasonably be expected to let. Rent actually received is evidence of that value rather than the measure of it, and properties that are not let, beyond the limited relief the Act grants for self-occupied houses, can still carry a charge on their notional value. From the annual value the Act allows municipal taxes actually paid and then grants two deductions only, a standard thirty per cent and interest on borrowed capital, an austerity that makes this head the simplest computation in the Act and the one whose simplicity the replacement statute most visibly preserved. Its perennial dispute is the boundary with business income where letting is organised as a commercial activity, a line the courts have drawn by asking whether the letting exploits the property as an owner or deploys it as the stock of a trade.

The head’s mechanics repay a closer look because they show the Act simplifying by fiction. Ownership is the gateway, and section 27 deems specified persons to be owners, including a transferor who has parted with possession under an agreement to sell without a registered conveyance, and the member of a housing society allotted a flat under its scheme, so that the charge follows enjoyment rather than the register of title. Where a property is co-owned and the shares are definite, section 26 assesses each co-owner separately on his share rather than the co-owners together as an association, one of the few places where the Act prefers the fragment to the whole. Vacant property and self-occupied property required repeated legislative attention, and the number of houses a person may claim as self-occupied without a charge on notional rent was raised from one to two by the Finance Act of 2019, a small amendment that mattered to thousands of families and illustrates the scale at which much of the Finance Act traffic operates. Interest on money borrowed for a let property is deductible without the cap that applies to a self-occupied house, an asymmetry that shapes borrowing decisions and shows the head quietly performing housing policy.

Profits and gains of business or profession, in sections 28 to 44, is the head under which the economy’s productive income is taxed and the head with the deepest computational code. It charges the profits of any business or profession carried on at any time during the previous year, then spends a long run of sections deciding what may be deducted in computing those profits: rent and repairs, depreciation on assets used for the business, interest on borrowed capital, insurance, bad debts actually written off, and the residuary general deduction in section 37 for expenditure laid out wholly and exclusively for the purposes of the business, provided it is neither capital nor personal in nature. Around the general deduction stand the disallowances, provisions that deny deductions for payments made without withholding tax, for cash payments above prescribed limits, for unpaid statutory dues, and for categories Parliament wished to discourage. Section 37 is also where the Act’s treatment of corporate social responsibility sits, because an Explanation inserted by the Finance (No. 2) Act, 2014 declares that expenditure on activities relating to corporate social responsibility under section 135 of the Companies Act is not deemed to have been incurred wholly and exclusively for business, and therefore cannot reduce taxable profits, which is the point at which the statutory CSR obligation in section 135 of the Companies Act meets the tax code and is told that social expenditure, however mandatory, is not a business cost.

Within this head sits depreciation, the deduction by which the cost of wasting assets is recovered against profits, and its modern form is a Finance Act creation. The Finance Act of 1987 replaced the asset-by-asset computation with the block of assets system from assessment year 1988-89, under which plant and machinery, buildings and intangible assets are grouped into blocks by rate, depreciation is allowed on the written down value of the block, and individual assets lose their identity for this purpose until the block itself empties or is sold. The reform ended a generation of disputes about individual assets at the price of a new arithmetic of adjustments on purchase and sale, and it is the pattern of much reform under this Act: a litigated case-by-case inquiry exchanged for a formula, simpler to administer and cruder in its fit to any particular business. Alongside the ordinary computation stand the presumptive schemes, under which small businesses, transporters and specified professionals may offer a prescribed percentage of their receipts as income and be spared the full apparatus of books and audit: the lineage runs from the schemes of the 1990s to section 44AD for businesses, section 44ADA for professionals, added by the Finance Act of 2016, and section 44AE for goods carriage operators. Presumption is the Act’s admission that its full computation costs more to apply to a small enterprise than the enterprise’s tax is worth, and the steady enlargement of the schemes’ thresholds is the quiet counterweight to every enlargement of the statute’s demands elsewhere.

Capital gains, in sections 45 to 55, taxes the profit made on transferring a capital asset, a category the Act defines negatively by excluding stock in trade, personal effects within limits, and specified rural agricultural land. The computation deducts from the consideration the cost of acquiring and improving the asset, and for long periods the Act adjusted that cost for inflation by indexation, while applying concessional rates to gains on assets held beyond prescribed periods; the exact rates, holding periods and the availability of indexation have been rewritten repeatedly by Finance Acts, including in the rationalisation carried by the Finance Act of 2024, and are stated in this pillar as a mechanism rather than as figures, because any figure printed here would be the property of one Finance Act and false in the hands of a reader of another year. The head’s characteristic litigation is the character question, whether shares or land were held as investment or traded as stock, answered from the taxpayer’s conduct, frequency, volume and intention, and its characteristic reliefs are the reinvestment exemptions, which discharge the charge where the gain is committed to a house or specified assets within the Act’s conditions and periods.

The computation rules around the charge show the same accretion. Section 55 fixes what counts as the cost of acquisition and improvement, and its history includes the device of allowing an assessee who acquired an asset long before to substitute its fair market value at a statutory base date, originally 1 April 1981, as the cost, a base the Finance Act of 2017 advanced to 1 April 2001 so that appreciation across the earlier decades escaped the charge altogether. Section 50C deems the stamp duty value adopted by the state registration authority to be the consideration on a transfer of land or buildings where the declared consideration is lower, a deeming aimed at the endemic understatement of prices in property transactions, with a reference to a valuation officer as the safety valve. Each device answers a real abuse or a real unfairness, and each generated its own jurisprudence, so that the capital gains computation became a sequence of statutory layers in which the year of acquisition decides which layers apply. No head better supports the article’s thesis that the Act is read as sediment or not read accurately at all.

Income from other sources, in sections 56 to 59, is the residuary head, and its residuary character is the point. It charges income that is not chargeable under any of the four specific heads, and it expressly gathers interest on securities, dividends and a list of specified receipts, while allowing deductions for expenditure laid out wholly and exclusively to earn the income. Because it is a sweep provision, it is the department’s natural home for receipts it cannot place elsewhere, and the courts have had to police its edge, holding that the residue catches only what is income in character, so that a capital receipt does not become taxable merely because no head fits it.

Parliament progressively converted the residuary head into a specified one, and the conversion is instructive. Section 56(2) enumerates receipts deemed to be income under the head: compensation and enhanced compensation interest, sums forfeited from advances toward capital asset transfers that fall through, the gift charge described earlier in this article, share premium in excess of fair value received by closely held companies on the issue of shares, and other categories added as transactions revealed them. Each enumeration is a small statute of its own with its own definitions and exceptions, and several, the share premium provision being the celebrated instance, drafted as anti-abuse measures against the routing of unaccounted money into companies as capital, caught genuine investment in their width and had to be tempered by notification and safe harbour. The residuary head’s history thus recapitulates the whole Act’s: a general power of capture, disciplined by the courts, supplemented by specific charges, and widened by each specific charge beyond the abuse that prompted it. The discipline the courts imposed is fundamental to the whole Act: the heads classify income, they do not manufacture it, and a receipt outside the concept of income and outside every charging provision is outside the tax, however convenient its taxation would be.

A final structural feature of the heads should be named before the machinery that surrounds them, because it governs the rate a rupee of income ultimately bears. The heads are the unit of computation, but the rate structure of each Finance Act applies to total income as an aggregated sum, so the heads matter to rates chiefly through their own special provisions: capital gains taxed at concessional rates prescribed for the category, specified incomes of non-residents taxed at flat rates under the special chapters, and casual receipts taxed at a flat rate under their own charging provision. The design consequence is that classification disputes are rarely only about arithmetic. Where a concessional or a penal rate attaches to the head, the dispute about the head is a dispute about the rate, and the parties’ energy in the long case law on heads reflects that stake. The replacement statute preserves the architecture entire, heads for computation and special rates riding on identified categories, which is further evidence for the continuity thesis this article has pressed: what was replaced in 2025 was the expression of the system, not the system.

Around the five heads the Act builds the aggregation machinery. The chapter on income of other persons, sections 60 to 64, clubs specified income back into the hands of the person who arranged the transfer, attacking the simplest avoidance device of diverting income to a spouse, a minor child or a revocable trust while retaining the benefit. The chapter on set-off and carry-forward, sections 70 to 80, permits losses under one head to be absorbed against income under another within rules that vary by the character of the loss, and allows unabsorbed business losses and depreciation to be carried forward for defined periods, with speculation losses quarantined in their own compartment. These chapters are where the Act’s anti-fragmentation policy lives, and they explain why classification disputes matter beyond the year in question: the head under which a loss arises decides against what income, and for how long, it can be used.

The deductions chapter and the courts’ two rules for reading reliefs

Chapter VI-A, sections 80C to 80U, is the Act’s incentives shelf. Where section 10 removes receipts from total income altogether, the deductions chapter allows specified payments and outlays to be subtracted from total income after it is computed, and the distinction between the two mechanisms is structural rather than verbal. An exemption means the receipt never enters the computation. A deduction means it enters and is then allowed against it, usually up to a monetary ceiling, usually on conditions, and usually for a stated policy purpose: to encourage savings and insurance, housing, medical cover, education borrowing, donations to specified funds and institutions, and investment in designated regions and industries. Many of the chapter’s deductions carried sunset dates and have expired on schedule, which is why the chapter is a museum of the fiscal fashions of successive decades as much as a code of reliefs, and why the alternative tax regime created in 2020 could define itself largely by listing the deductions it required the taxpayer to surrender.

Section 80C, the chapter’s anchor, has its own instructive biography. Until the Finance Act of 2005 the encouragement of household savings ran through a rebate, a reduction of the tax itself under the former section 88, graduated by income and withdrawn as income rose. The 2005 recasting converted the rebate into a deduction of up to a ceiling amount for a basket of specified outlays, life insurance premia, provident fund contributions, principal repayment on housing loans, tuition fees, specified savings instruments and others, available across income levels and subtracted before the rate applied. The change of instrument altered the benefit’s distribution, since a deduction is worth most to the taxpayer in the highest slab, and the debate between the rebate’s progressivity and the deduction’s simplicity was conducted, characteristically, in the schedule of a Finance Bill rather than in any standalone consideration of savings policy. The basket itself became the annual object of lobbying by every industry whose product sought admission to it, and the list in the section is best read as a map of which savings products successively persuaded the finance ministry of their national usefulness.

The courts built two interpretive rules for this terrain, and a reader who holds both can predict most deduction litigation. For exemption provisions, the settled position, confirmed by a Constitution Bench in the Dilip Kumar case of 2018 concerning exemption notifications, is strict construction: the claimant must bring himself squarely within the words of the exemption, ambiguity is resolved in favour of the revenue, and the burden of proving eligibility lies on the person claiming the relief. For incentive deductions in Chapter VI-A, an older and different instinct survived in the line of authority associated with the Bajaj Tempo decision of 1992, that a provision intended to encourage an activity should be construed so as to advance the incentive rather than defeat it on a technicality. The two rules coexist because they govern different instruments, a carve-out that withdraws the state’s own revenue on conditions, read strictly, and a behavioural incentive Parliament enacted to be used, read purposively. Much of the skill in reading the Act consists of knowing which rule a given provision attracts, and many appeals have been won and lost on that classification alone.

One further doctrine completes the interpretive frame, and it concerns the Act as a whole rather than its reliefs. The courts have consistently held that a taxing statute is construed by its words, that there is no equity about a tax, and that a subject is not taxed unless the charging words clearly impose the charge, a principle descended from English authority and restated across the Act’s life. The principle does not licence obstruction; it disciplines the department to find the charge in the text, and it disciplines the taxpayer to find the relief in the text, and it is the deep reason why the drafting style of the Act, its provisos, explanations and deeming fictions, matters so much. In a field where argument from fairness is formally worthless, the text is the whole case, which is why a statute whose text became unreadable presented a problem that no amount of judicial goodwill could cure.

One further interpretive tool governs the deeming provisions scattered through the Act, the fictions by which income is deemed to accrue in India, property is deemed to be owned, or a value is deemed to be a price. The settled approach is two-sided: a fiction must be carried to its logical conclusion for the purpose Parliament enacted it, because a half-applied fiction defeats the provision, and it must not be extended beyond that purpose into fields the provision was not addressing. The two-sidedness explains the characteristic shape of deeming litigation, in which the department presses the fiction outward and the taxpayer presses it back to its purpose, and the court restates both halves of the rule. It also explains why the Act’s draftsmen reached for fictions so often: a deeming provision converts a factual inquiry the department would lose, where income really accrued, who really owned, what price was really paid, into a legal rule it can administer, at the cost of a standing argument about the fiction’s proper edge.

A word on the life cycle of reliefs completes the picture. Deductions and exemptions under this Act are rarely permanent settlements. Many carry express sunset dates, inserted because the incentive was aimed at a class of investment Parliament wished to attract for a period, and when the period closes the provision remains in the printed Act for years, spent but unrepealed, still consulted by readers who cannot tell from its face that no claim can arise under it for later years. Others are withdrawn prospectively with grandfathering for commitments already made, so that two taxpayers in identical businesses are taxed under different relief codes because their projects began in different years. Still others are withdrawn and partly restored, or capped, or conditioned on fresh approvals, each variation enacted against the text laid down by its predecessors. The relief chapters are therefore the least self-sufficient part of the statute: a provision states its promise, but the Finance Acts of the intervening years state whether the promise is still being kept, and only a reader who checks the trail knows which.

The second code inside the Act: special provisions and special taxpayers

Alongside the general code of charge, computation and machinery, the Act grew a second code of special provisions, and by bulk it rivals the first. Chapter XII and its lettered successors prescribe the determination of tax in special cases: special rates and computational rules for non-residents on royalties, fees for technical services and capital gains, presumptive schemes for shipping, aircraft, mineral oil and other industries whose income cannot practicably be computed by ordinary accounts, and optional regimes for companies that surrender exemptions for lower rates. Chapter XII-A created a self-contained optional code for non-resident Indians on investment income, with its own rates, its own withholding and a facility to dispense with returns in defined cases. A separate chapter taxes units and funds through bespoke pass-through and levy arrangements that have been rewritten as the financial sector invented new vehicles. The pattern in every instance is the same. A class of taxpayer or transaction appears for which the general code fits badly, and instead of redesigning the general code, Parliament bolts a special regime onto its side, preserving the misfit and the remedy together.

The minimum alternate tax is the largest single exhibit. Introduced in an early form in the late 1980s and recast as section 115JB by the Finance Act of 2000, it addresses the scandal, as successive finance ministers described it, of profitable companies paying no tax because exemptions and depreciation reduced their taxable income to nothing while their book profits and dividends flourished. The provision computes a floor tax on book profit where the ordinary computation yields less, with credit carried forward for the excess, and an alternate minimum tax on the same pattern was later extended to firms, limited liability partnerships and other non-corporate taxpayers. The device is an admission that the incentive structure had outrun the charge, and its history of extensions, rate changes and sectoral carve-outs, including the eventual bargain under which companies electing the concessional regimes of 2019 exit the minimum tax altogether, is a compact history of Indian corporate tax policy in itself.

Charitable and religious trusts occupy another special world, in sections 11 to 13, where income from property held for charitable or religious purposes is exempt on conditions of application, accumulation, registration and audit that have been tightened in waves, most heavily when Parliament concluded that the exemption sheltered commercial activity dressed as charity. Universities, hospitals, funds and institutions are handled through approval-based exemptions in section 10 and the deduction provisions for donations, creating an interlock between the tax code and the regulatory state for the non-profit sector. Firms and limited liability partnerships are assessed as entities with a flat rate and a deduction for partner remuneration and interest computed on a statutory formula, while the partners’ shares of profit are exempt, an architecture designed to tax the collective once and the members not at all on the same income. Hindu undivided families survive as a taxable person in their own right, a recognition of the joint family as an economic unit that no simplification exercise has yet dared to abolish. Each of these regimes is defensible, several are elegant, and their accumulation is precisely the point this pillar has been making: the Act did not become two codes by anyone’s decision. It became two codes by sixty years of decisions, each reasonable in the year it was taken.

The dividend provides the special code’s neatest parable of arrival and departure. For most of the Act’s life, dividends were taxed in the shareholder’s hands, with deduction at source. The Finance Act of 1997 moved the charge to the company distributing the dividend, creating the dividend distribution tax under section 115-O, collected once at a flat rate while dividends were exempt in the shareholder’s hands, a design defended for its collection efficiency and attacked for taxing profits twice in economic substance and for dulling the price signal of the distribution decision. The Finance Act of 2020 abolished the distribution tax and restored taxation in the shareholder’s hands at his own slab rate, with withholding, so that within one professional generation the same receipt was taxed under three different architectures, each announced as a reform of the last. Nothing in the episode was irrational at any step. Its lesson for the reader of the special provisions is simply that their permanence should never be assumed: they are the most frequently rewritten stratum of the Act, because they are where each government’s preferences about who should bear the visible burden of the tax are recorded.

The special code’s outer edge lies outside the Act altogether, and a pillar should mark it. When the digital economy produced income that the Act’s residence and source rules could not satisfactorily reach, Parliament did not force the new levy into the old code. The Finance Act of 2016 enacted the equalisation levy as a self-contained chapter of that Finance Act itself, a charge on specified online advertising payments to non-residents, extended by the Finance Act of 2020 to e-commerce operators, administered through machinery borrowed from the Income-tax Act but legally standing apart from it, so that its charge did not disturb treaty entitlements that attach to income tax. The levy’s later partial withdrawals belong to trade diplomacy rather than to this article, but its placement is the structural point: by 2016 the Act had become so dense a thicket that a government legislating a genuinely new tax preferred to build beside it rather than within it. The 2025 consolidation, which gathers the field back into one code, is in part a judgment on that preference.

The annual Finance Act: amendment as the design, not the defect

The namable claim of this pillar can be stated in one sentence. Annual amendment is the design of the Income-tax Act, not a defect that befell it, because the Indian budget process writes tax policy into the statute itself every year, and any statute so written will accumulate complexity as a mathematical certainty. The mechanism is constitutional. The annual financial statement is presented to Parliament, the Finance Bill follows it as a Money Bill, and the Finance Act that emerges does two distinct things. It fixes the rates of income tax for the coming year, which section 4 of the Income-tax Act deliberately leaves to it, setting them out in its First Schedule. And it amends the Income-tax Act in substance, inserting, substituting and omitting provisions across the whole code, so that rates legislation and structural legislation travel in the same vehicle.

The First Schedule itself deserves a reader’s attention, because it is where the political content of the tax is printed. Its parts prescribe the rates for the year for individuals and Hindu undivided families, for firms, for companies and for the special categories, with the surcharge and the cess that ride on the computed tax, and it is reprinted annually with the year’s figures, so that the operative rate law of India is a schedule to a statute that itself expires with the year. Alongside the Schedule, the budget speech performs the announcement function: changes in exemptions, thresholds and regimes are declared by the finance minister on budget day, take legal effect when the Finance Act passes weeks later with retrospective reach to the year’s first day, and are debated in the interval as policy while remaining inoperative as law. That gap between declaration and enactment, familiar to every practitioner who has advised on a budget proposal, is a further consequence of the annual vehicle: the tax law of the country is, for several weeks of every year, a statement of intention attached to a Bill. A Parliament that wished only to set rates could have left the Act alone; the Indian practice combined the two, and the combination ran for over six decades.

The consequences compound. Because the Finance Bill is a Money Bill, its passage is driven by the budget timetable and dominated by the lower House, and its clauses amending a permanent code receive a fraction of the scrutiny that a standalone amendment Bill would face in committee. Because the amendments take effect from the first day of the new financial year, the profession and the department absorb each year’s changes at speed, and transitional rough edges are repaired, if at all, by the next year’s Finance Act rather than by considered revision. Because insertion is administratively cheaper than renumbering, new provisions arrive as lettered sections, new chapters arrive as lettered chapters, and qualifications arrive as provisos to provisos and explanations to explanations, the Finance Ministry’s count of roughly 1,200 provisos and 900 explanations in the closing text being the statistical portrait of the method. And because every year’s amendment is drafted against the text as it stands that year, provisions enacted for a closed fiscal purpose, a temporary incentive, a transitional rate, remain in the book long after their purpose closes, since repealing dead law requires the same scarce legislative effort as enacting live law and rewards nobody within a budget cycle.

The mechanism’s constitutional character deserves one more paragraph, because it explains why scrutiny failed. A Money Bill is defined by Article 110 of the Constitution, certified by the Speaker of the Lok Sabha, and insulated by design: the Rajya Sabha may only recommend amendments, which the Lok Sabha may reject, and the Bill must be returned within fourteen days. That insulation is appropriate for supply, the raising of the year’s revenue, for which the directly elected House bears responsibility. Its extension to the permanent amendment of a code governing every economic transaction in the country is the accident at the centre of this article. Amendments of lasting structural effect, to anti-avoidance law, to assessment procedure, to the rights of searched persons, travelled with the budget’s timetable and the Money Bill’s procedural shelter, debated under the pressure of the financial calendar and rarely examined by a select committee of the kind that considered the replacement Bill of 2025 at length. Criticism of that practice came from parliamentarians of several parties across the decades and from commentators on the budget process, and it was sharpened when Finance Acts began carrying amendments to statutes outside taxation altogether. The counter-argument, made by governments of both complexions, is that tax policy is inseparable from the budget and that the Finance Bill is its natural and only timely vehicle. Both arguments are honestly held, and the sediment in the Act is what their unresolved contest looks like when printed.

It follows that the familiar indictment of the Act’s draftsmen misses its target. The statute did not become difficult because its language was neglected. It became difficult because it was the most amended text in Indian law, amended by more than sixty Finance Acts and numerous Taxation Laws Amendment Acts, carrying several thousand individual changes, and because the amendment vehicle was constitutionally guaranteed to return every year regardless of whether the code needed changing. The 2025 replacement accepts this diagnosis in its design: it restructures, renumbers and rewrites, but it leaves untouched the annual Finance Act mechanism that produced the sediment, since the mechanism is a feature of the budget process and not of the tax statute. The prediction this pillar records is therefore structural rather than cynical. The new Act begins clean, and the same process that filled the old one begins filling it from its first Finance Act, which is why the simplification of 2025 is best understood as a reset of the text rather than a cure for the accumulation.

Practitioners who work with both statutes keep their own apparatus for surviving this process, because the official text alone does not reveal which year’s version of a section governed which year’s income. Annotated editions, amendment histories and personal provision maps are the profession’s standard equipment, and readers building their own study files on the Act, whether provision maps, amendment chronologies or collections of circulars and orders in spreadsheet and document form, can open Excel, Word and PowerPoint files in the browser through the office file viewer when a file arrives in office format and no desktop suite is at hand. The detail matters for a reason beyond convenience. In this field the operative text is never the Act in the abstract; it is the Act as it stood in the previous year relevant to the dispute, and the reader who cannot reconstruct that year’s text cannot test any statement made about it, including the statements in this article.

Collection at the source: withholding, advance tax and refunds

A charging statute that waited until the year ended to collect would fail administratively and fiscally, so the Act builds collection into the flow of income itself. Chapter XVII is the machinery, and its centrepiece is tax deducted at source. The design obliges the payer of specified categories of income, salary, interest, dividends, rent, professional fees, commission and many others, to deduct tax at prescribed rates when paying or crediting the income, and to deposit that deduction with the government as an advance collection on the payee’s eventual liability. The payee claims credit for the deduction in his assessment, and the payer faces interest, fee, penalty and disallowance of the corresponding expenditure for default. Tax collected at source extends the same logic to specified receipts of sellers. The effect is that the department receives most of its individual revenue before any assessment occurs, acquires a reporting trail for every significant payment stream, and converts the country’s employers, banks and businesses into unpaid collection agents, a transfer of administrative cost to the private sector that is one of the Act’s defining policy choices and one of its least debated.

Withholding works because the Act first builds an identity and reporting system for the money it follows. Section 139A requires specified persons to obtain a permanent account number, the ten-character identifier that anchors every return, deduction, deposit and high-value transaction report in the system, and obliges its quotation across a widening list of dealings. Section 285BA requires banks, companies, registrars and other specified persons to report high-value transactions in an annual statement of financial transactions, and section 285BB, inserted by the Finance Act of 2020, built on that reporting the Annual Information Statement, a consolidated account of a taxpayer’s reported financial dealings placed before him at filing time. The tax credit statement, long known as Form 26AS, collects the deductions made against a number in one view. Together these provisions converted the Act from a code that waited for the citizen’s account of himself into one that assembles the state’s account of the citizen in advance and asks him to reconcile it, a transformation in the balance of information between taxpayer and department that is arguably the deepest change the machinery underwent in the Act’s lifetime, deeper than any change in rates.

The withholding machine is completed by its certificate and credit apparatus. The deductor issues certificates, Form 16 for salary and its counterparts for other payments, the credits are gathered in the payee’s tax credit statement, and discrepancies between the credit claimed and the credit recorded became, in the system’s middle years, a grievance of the first order, since tax deducted but not properly reported by the deductor was tax the payee had to chase rather than use. The Act answered with the deductor’s reporting duties enforced by fee and penalty, with section 206AA, brought into force from 2010, requiring deduction at a higher rate where the payee furnishes no permanent account number, and with the faceless reconciliation of the information statement era, in which the credit assembles itself from the payers’ filings. The design lesson generalises: every collection mechanism in this Act generates a verification mechanism behind it, and usually a litigation genre behind that, and the sections providing the three layers sit adjacently in the chapter for a reader who knows to look.

Advance tax completes the design for income that no payer withholds upon. Assessees whose liability exceeds a modest statutory threshold must estimate their income and pay the tax in instalments across the previous year, on the principle the department calls pay as you earn, with interest charged under sections 234A, 234B and 234C for late filing, short payment and deferment of instalments respectively. The interest code is compensatory in stated design and penal in felt effect, and it operates automatically, without any officer’s discretion, which makes it the part of the machinery taxpayers most often meet and least often contest. Refunds close the circle in Chapter XIX: where collections exceed the liability finally determined, the excess is refundable with interest in defined cases, and the history of refund delay and adjustment against disputed demands is a standing item in every critique of the department’s administration, because a system that collects first and adjudicates later holds the taxpayer’s money while it decides.

Assessment: return, scrutiny, reopening and the search power

The assessment chapters, principally Chapter XIV, convert liability into a determined demand. Ahead of the return stands the department’s power to gather material, which conditions everything that follows. Under section 131 the income tax authorities hold the powers of a civil court for discovery, inspection, enforcing attendance and compelling production of books, and under section 133 they may call for information from firms, financial institutions and other persons in defined classes. These inquiry powers, exercised through notices under section 142 once a case is taken up, mean that assessment is the visible tip of a continuous intelligence function, fed by the reporting system described earlier and capable of reaching a taxpayer who has never been selected for scrutiny at all.

The sequence begins with the return of income under section 139, which specified persons must file when their income exceeds the basic exemption threshold and which many others file to establish refunds, losses or a compliance record. The return is processed centrally under section 143(1), with arithmetical and prima facie adjustments, and a minority of returns are selected for scrutiny, ending in an assessment under section 143(3) in which the Assessing Officer examines the return, calls for evidence and determines the total income. Where no return is filed or the officer cannot proceed on the material, a best judgment assessment under section 144 follows. The design distributes the burden characteristically: the taxpayer self-declares, the system processes by default, and the officer’s full adjudicatory power is reserved for the selected minority, which is why the criteria and manner of selection have always been as contested as the assessments themselves.

The routine end of the process is as administrative as its beginning. Returns are processed at the department’s centralised processing centre, which issues the intimation under section 143(1), computes refunds and raises the demands that follow from prima facie adjustments, handling the great majority of the country’s returns without any Assessing Officer’s involvement. Mistakes apparent from the record may be rectified under section 154 on the application of either side, a narrow power confined to errors that need no argument to establish, and the Commissioner’s revision powers under sections 263 and 264 stand behind the process for orders that are erroneous and prejudicial to the revenue, or for grievances the appellate route does not fit. Parliament has also compressed the clock throughout: the period allowed for completing a scrutiny assessment was reduced by stages across successive Finance Acts, reaching nine months from the end of the assessment year under the Finance Act of 2021, a compression that changed the practical character of scrutiny, forcing selection, evidence and hearing into a window the original Act’s draftsmen, who allowed years, would have found unrecognisable. Speed became a statutory value alongside accuracy, and the friction between the two is a recurring ground of appeal.

Finality is qualified by reassessment. Where the Assessing Officer has reason to believe that income has escaped assessment, sections 147 to 151 permit the completed position to be reopened within defined limitation periods, three years in the ordinary case and a longer period, extending to ten years, where the escaped income is substantial and represented in the form of an asset. The reassessment code was substantially recast by the Finance Act, 2021, which rebuilt the notice procedure, interposed a mandatory show-cause stage under section 148A before a reopening notice may issue, and recast the concepts on which the officer’s belief must rest. The recast was Parliament’s response to a generation of litigation in which reopening, designed as an exceptional power against escaped income, had become a routine second assessment and the single largest source of the department’s disputes, and the courts’ supervision of the new procedure has continued the argument in fresh vocabulary. Reopening is where the Act’s two values collide most openly: the revenue’s claim that no escaped income should be immunised by an officer’s earlier error, and the taxpayer’s claim that a tax system must let closed years stay closed. The statute’s answer is a compromise written in limitation periods and procedural gates, and it moves whenever either side persuades Parliament that the compromise has failed.

What powers does the Income-tax Act give the department during a search?

Section 132 confers the Act’s most intrusive power. Where the department has information giving reason to believe that a person is in possession of money, bullion, jewellery or other articles representing income that has not been or would not be disclosed, authorised officers may enter and search premises and vehicles, seize books, documents and assets found.

Authorised officers may also restrain the disposal of assets found but not seized, and the power extends to breaking open locks and to requisitioning material in another authority’s custody. Section 133A confers the lesser power of survey, entry into business premises during working hours to inspect and record, without seizure of assets. Both powers descend from the 1922 Act, both are hedged by the requirement of recorded reasons and authorisation at senior level, and both have been supervised by the courts on the sufficiency of the belief that triggers them.

Search carries its own assessment procedure behind it. A block assessment scheme in Chapter XIV-B, introduced in 1995, first gathered the years surrounding a search into a single proceeding on the undisclosed income found, and the Finance Act of 2003 replaced it with the procedure in sections 153A to 153D, under which the income of the years surrounding a search is assessed through a special track, with abatement of pending assessments for those years and approval requirements at senior level before the orders issue. The special track was Parliament’s answer to the chaos of parallel proceedings that searches under the ordinary procedure had produced, and its history repeats the Act’s rhythm: an exceptional power creates exceptional litigation, the litigation is answered with a procedural code of its own, and the code generates fresh questions, about the material found in the search, about years in which nothing was found, about the effect of a search on a third party’s assessment, that the courts then spend a decade settling. Where the search finds nothing incriminating, the machinery still runs, which is the sharpest form of the taxpayer’s complaint against the power and the point on which its defenders answer that undisclosed income is, by definition, found only where it is looked for without notice.

Recovery is the machinery’s final stage and its least visible. An assessment crystallises into a notice of demand under section 156, and Chapter XVII’s later parts arm the department to collect it: the Tax Recovery Officer proceeds under the Second Schedule’s code of attachment and sale of movable and immovable property, the Assessing Officer may direct a third party holding money for the defaulter to pay the department instead, in the garnishee procedure of section 226, and the department ranks as a creditor with statutory protections for its dues. The recovery code is deliberately older in spirit than the assessment code, closer to the revenue recovery law of the nineteenth century than to the faceless present, and disputes at this stage are correspondingly about property and priority rather than about interpretation. A pillar records it because the life of a liability is not complete at assessment, and because the state’s willingness to use these powers against the small defaulter while large demands wait in appeal is a standing criticism that any honest account of enforcement must carry.

The search power matters structurally because it stands outside the cooperative design of the rest of the machinery. Assessment proceeds on returns and evidence; search proceeds on suspicion and surprise, and the material it gathers feeds a special assessment track for the years it covers. Its existence shapes behaviour far beyond the premises actually entered, which is precisely the deterrent logic its defenders cite, while its critics, including committees that examined the department across the decades, have questioned whether the yield of searches justifies the disruption and whether the power’s breadth invites its use as pressure in unrelated disputes. The honest position records both: the power is the Act’s necessary answer to income that no return will ever disclose, and it is the power whose misuse does the most reputational damage to the system that wields it, a tension the replacement statute inherited entire rather than resolved.

The most consequential recent change to the machinery is the removal of the human interface. Faceless assessment began as an experiment in electronic assessment and was generalised by scheme in 2019 and 2020, then given a statutory home in section 144B, inserted by the Taxation and Other Laws (Relaxation and Amendment) Act, 2020 and operating from 1 April 2021. Under the faceless design, assessments are made through national and regional centres, the case is allocated without regard to the taxpayer’s location, the officers who assess never learn the identity of the assessee and the assessee never meets them, notices and replies travel through the electronic portal, and draft orders pass through risk management and review units before issue. The territorial Assessing Officer, the fixed point of the 1961 design, disappears from scrutiny assessment altogether. The stated purposes were to eliminate discretion, reduce the corruption that personal interface made possible, and distribute work nationally. The recorded experience, stated at full strength by both camps later in this article, includes each of those gains and a corresponding loss, the taxpayer’s difficulty in explaining a complex case to an unseen and rotating reader. Parallel faceless schemes extended the model to appeals at the first stage, and the 2025 statute absorbed the faceless architecture as its default rather than as an exception, which settles the direction of travel even where particular grievances remain live.

The appellate ladder: from the first appeal to the Supreme Court

Chapter XX builds a four-rung ladder, and its structure reflects a deliberate judgment about where tax disputes should be fought. The first appeal lies under section 246A to the Commissioner (Appeals), or to the Joint Commissioner (Appeals) for the classes of cases assigned to that authority, a departmental appellate authority independent of the assessing hierarchy for this purpose, which may confirm, reduce, enhance or annul the assessment. The second appeal lies under section 253 to the Income Tax Appellate Tribunal, a judicial tribunal composed of judicial and accountant members, which is the final authority on facts: it may decide any question of fact afresh on the record, and its findings of fact bind the courts that sit above it. From the Tribunal, an appeal lies under section 260A to the High Court, but only on a substantial question of law, and from the High Court under section 261 to the Supreme Court where the case is certified or leave is granted. Alongside the ladder stand the Commissioner’s powers of revision under sections 263 and 264, and the mechanisms for rectification of mistakes apparent from the record.

Two specialised channels branch off the ladder. For a defined class of cases, principally those involving foreign companies and transfer pricing references, section 144C, inserted by the Finance (No. 2) Act of 2009, interposes a Dispute Resolution Panel before assessment is completed: the Assessing Officer forwards a draft order, the assessee may object to a collegium of Commissioners, and the panel’s directions bind the officer, a design intended to give high-stakes international cases a senior internal hearing before positions hardened into appeal. The second channel looks forward rather than back. Chapter XIX-B created an advance ruling mechanism in 1993, under which non-residents, and later specified residents, could obtain a binding ruling on the tax treatment of a proposed transaction from an Authority for Advance Rulings, a device imported to give investors certainty before they committed capital. The Authority was replaced by a Board for Advance Rulings under the Finance Act of 2021, a change of institutional form that kept the function, the promise of an answer in advance being one a tax system competing for investment cannot easily surrender, whatever the backlog at the door.

The first appellate authority’s powers define the practical experience of appeal. Under section 251, the Commissioner (Appeals) may confirm, reduce, enhance or annul an assessment, and in other cases may set the matter aside or pass such order as the case requires, and the power of enhancement, exercised after notice to the appellant, means that an appeal is not a riskless option but a reopening of the assessment to an authority that may find more in it than the officer did. The Tribunal above it is the system’s workhorse and its most distinctive institution, a bench of a judicial member and an accountant member combining the two kinds of expertise a tax dispute joins, sitting in benches across the country, and empowered under section 254 to pass such orders as it thinks fit after hearing both sides. Because the Tribunal is final on fact and because its members specialise, its orders constitute the largest single body of applied tax law in the country, cited between benches as persuasive authority and binding on the department’s officers within the jurisdiction unless reversed above. Reform proposals have repeatedly contemplated a national reorganisation of the Tribunal’s place, and the article on the appeals institution owns that debate.

Between assessment and appeal sits the question of payment, and its treatment shapes the system’s fairness as directly as any hearing. A demand is payable notwithstanding an appeal, and stay of recovery while the appeal pends is a discretion exercised under section 220, on the assessee’s application, by the Assessing Officer subject to the appellate authorities’ control. Administrative instructions of the Board standardised the exercise, ordinarily expecting payment of a proportion of the disputed demand, commonly twenty per cent, as the condition of stay, with senior officers empowered to vary the proportion on the merits shown. The practice is a compromise between the revenue’s claim that adjudicated demands should be collected and the taxpayer’s claim that collection during appeal can destroy the business whose income is in dispute before any independent authority has heard him, and complaints that the compromise hardens into a tariff, applied without regard to the strength of the case, are the standing criticism of the recovery side of the ladder.

The design deserves attention because it allocates trust. Facts are exhausted at the Tribunal, a specialist body whose members hear tax appeals in volume and whose benches have built much of the working law of the Act through orders that bind the department within their jurisdiction. Law alone travels upward, so the High Courts and the Supreme Court shape doctrine on a diet of pure legal questions, which is why the jurisprudence of the Act is disproportionately a jurisprudence of interpretation. The ladder’s chronic disease is volume and time, and the department’s historically low success rate in appeals it chooses to file has itself become a policy problem, addressed in part by monetary thresholds below which the department is directed not to appeal, an administrative admission that the ladder was being climbed for sums the climb could not justify. The full treatment of the ladder as an institution, its pendency and its reforms belongs to the article on taxpayer appeals and the tax tribunals, which owns that ground; this pillar records the structure because the structure determines what an assessment means, an assessment being less a conclusion than the opening position in a negotiation the Act expects to continue upward. The account given here states the general structure of the ladder and is not advice on any individual case, and limitation periods in tax appeals run from dates that are unforgiving of delay.

Penalties and prosecution: the deterrent chapters

Chapter XXI imposes penalties, and its history tracks the Act’s broader movement from discretion toward formula. The older architecture centred on the concealment penalty under section 271(1)(c), quantified as a percentage, between one and three times, of the tax sought to be evaded, imposed after hearing and litigated endlessly on the mental element the courts read into the word concealment. The Finance Act, 2016 replaced that centrepiece for later years with section 270A, which penalises under-reporting of income at fifty per cent of the tax payable on the under-reported income and misreporting at two hundred per cent, classifying misreporting by a statutory list of cases such as misrepresentation or suppression of facts and failure to record investments or receipts. The reform’s logic was to trade a culpability inquiry for a taxonomy, reducing argument about intention by specifying conduct. Around the centre sit penalties for failure to file returns, to maintain and produce accounts and documents, to deduct and deposit tax at source, and to comply with notices and information calls, together with the interest provisions already described, which frequently exceed the penalty in money terms. A parallel system of fees handles the routine defaults of the withholding and reporting machine: a fee under section 234E for each day of delay in furnishing deduction and collection statements, and a fee under section 234F, introduced by the Finance Act of 2017, for late filing of the return itself, graded by income and capped, so that the everyday frictions of compliance are priced automatically rather than prosecuted through penalty proceedings. The fee device is the penalty system’s own faceless turn, replacing an officer’s discretionary response to delay with a tariff, and its spread across the Act’s reporting obligations measures how far the administration had moved from adjudication toward automation even before the faceless schemes formalised the change. Immunity mechanisms allow an assessee to settle the penalty exposure by paying the tax and interest within defined windows, a pragmatic recognition that the system’s interest lies in collection rather than in the multiplication of penalty litigation.

The character of penalty proceedings was settled early and matters to how the chapters are read. In the Hindustan Steel decision of 1969, the Supreme Court held that penalty is not imposed for every technical or venial breach, that an authority must exercise its discretion on the conduct shown, and that the penalty power is quasi-criminal in the sense that it fastens on blameworthy conduct rather than on arithmetic alone. That tempering doctrine ran through the concealment era’s litigation, where the department’s burden and the assessee’s explanation were argued in nearly every case, and the 2016 recasting into the under-reporting and misreporting taxonomy of section 270A is best understood as Parliament narrowing the field on which the older discretion operated, specifying the conduct that counts as misreporting so that the dispute shifts from the assessee’s state of mind to the category his case falls in. Practitioners report both gains and losses from the exchange, certainty about the rate of penalty against the loss of a forum in which an honest error could be forgiven entirely, and the argument between those values is the penalty chapters’ own version of the argument that runs through the whole Act.

Chapter XXII adds prosecution, reserving the criminal law for the conduct at the far end of default: wilful attempt to evade tax, failure to deposit tax withheld, failure to file a return in specified circumstances, and false statements in verification, punishable with imprisonment graded by the amount involved, and gated by the requirement of sanction at a senior level under section 279 and by administrative compounding. Prosecutions under the Act have always been few relative to the size of the tax base, and their function is exemplary rather than systemic. The two chapters together state the Act’s theory of enforcement with unusual clarity. The ordinary defaulter is met by automatic interest and formula penalties, the disputing taxpayer by the appellate ladder, and the fraudulent by the criminal courts, and the boundaries between those categories, routinely tested, are where the department’s credibility is won and lost.

Anti-avoidance: transfer pricing, GAAR and the long argument about form and substance

The Act’s specific anti-avoidance code begins with transfer pricing. Chapter X, sections 92 to 92F, inserted by the Finance Act, 2001 and operating from assessment year 2002-03, addresses the problem that associated enterprises of multinational groups can shift profit out of India by pricing their internal transactions away from market terms. The chapter requires income from an international transaction to be computed by reference to the arm’s length price, prescribes the methods by which that price is determined, imposes documentation and reporting obligations including an accountant’s certificate, and empowers reference to a specialist Transfer Pricing Officer. It converted a general suspicion about multinational pricing into an audit industry of its own, generated a distinctive body of litigation about comparables and adjustments, and demonstrated the Act’s mature pattern: a targeted code, with its own definitions, machinery and appellate experience, inserted wholesale into the parent statute when a gap became undeniable.

The general anti-avoidance rule took the opposite path to arrival, through the courts’ refusal to build one. The interpretive arc is the intellectual history of the Act in miniature. In the McDowell decision of 1985, a Constitution Bench speaking through Justice Chinnappa Reddy condemned colourable devices and tax avoidance as a public evil, and for a period the revenue read the judgment as a general licence to look through transactions to their substance. The tide turned with the Azadi Bachao Andolan decision of 2003, which upheld the Indo-Mauritius treaty framework and the Board’s circular recognising residence certification, and restated that the McDowell observations did not license disregarding lawful arrangements or treaty entitlements. It turned further in the Vodafone judgment of 2012, where the Supreme Court declined to tax an offshore transfer of shares with underlying Indian assets under the Act as it then stood, and treated the revenue’s substance arguments with reserve. Parliament’s answer to Vodafone came in two parts: the retrospective amendment of the deeming provision in section 9, which belongs to the article on the retrospective taxation episode and is treated there in full, and, in the same Finance Act of 2012, the enactment of a statutory general anti-avoidance rule, Chapter X-A, sections 95 to 102, supplying by legislation the doctrine the courts had declined to supply by interpretation.

Chapter X-A empowers the department to declare an arrangement an impermissible avoidance arrangement where its main purpose is to obtain a tax benefit and it bears one of the tainted elements the chapter specifies: it creates rights or obligations not ordinarily created between persons dealing at arm’s length, it results in misuse or abuse of the Act’s provisions, it lacks commercial substance in whole or part, or it is carried out by means not ordinarily employed for bona fide purposes. The consequences include disregarding parties, reallocating income and expenditure, denying treaty benefits and recharacterising the arrangement. Enacted in 2012, the chapter was then deferred in stages on the recommendation of the expert committee chaired by Parthasarathi Shome, which advised a monetary threshold, procedural safeguards, grandfathering of earlier investments and a postponement to allow preparation, and it finally took effect for assessment year 2018-19, with arrangements below an aggregate tax benefit of three crore rupees excluded by rule, and investments made before 1 April 2017 grandfathered. The deferral history is itself instructive: the same state that enacted the rule in a budget also accepted, repeatedly, that applying it at once would damage the certainty on which investment depends, and the safeguards rather than the powers are what distinguish the Indian GAAR from the judicial doctrines it replaced.

The procedure for invoking GAAR, in section 144BA, is part of its substance and should be recorded in a pillar. An Assessing Officer who considers an arrangement impermissible may not proceed alone: the case travels by reference through the Principal Commissioner to an Approving Panel, constituted with senior officers and a retired judge among its members in the design as enacted, whose directions bind the officer. The assessee is heard, the panel’s approval is a condition of the declaration, and the chapter’s provisions for corresponding adjustments seek to prevent the same income being taxed twice where an arrangement is recharacterised across parties. The proceduralisation was the price of the power. Where the courts had refused a general anti-avoidance doctrine partly from a fear of uncontrolled administrative discretion, Parliament supplied the doctrine wrapped in a committee, and the early operation of the rule, measured in references rather than in mass application, suggests a power held in reserve for arrangements of scale, which is what its threshold and its history both predict.

GAAR does not stand alone, and its place is clearer beside the specific anti-avoidance rules that preceded it and the treaty system it overlays. Specific rules against identified devices are as old as the modern Act: provisions against bond washing and dividend stripping, of which sections 94(7) and 94(8) address the purchase of securities and units shortly before a dividend or bonus to manufacture a loss, sit in Chapter X alongside the transfer pricing code and descend from the 1922 code’s experience. Residence itself acquired an anti-avoidance edge when the Finance Act of 2015 introduced the place of effective management test in section 6(3), making a foreign company resident in India where its effective management sits here, aimed at shell incorporation abroad with control exercised from India, and operating from assessment year 2017-18. Over both sits the treaty network built under sections 90 and 90A, under which the Act or the applicable double taxation agreement applies to the assessee’s advantage, whichever is more beneficial, a rule that made treaty entitlements a real part of domestic liability and their abuse a real target: the Finance Act of 2012 required a tax residency certificate as a necessary condition of treaty relief, treaties themselves acquired limitation of benefits clauses, and GAAR was expressly empowered to deny treaty benefits where an arrangement is impermissible. The architecture that results is layered, specific rules for known devices, a residence test for shell structures, treaty conditions for treaty claims, and a general rule behind them all for the arrangements nobody foresaw, and the layering is the considered answer of a system that had tried judicial doctrine alone and found it unstable.

A boundary of the anti-avoidance code should be marked, because readers routinely conflate regimes that Parliament kept separate. Undisclosed foreign income and foreign assets were taken out of the ordinary assessment framework and placed under a dedicated statute, the Black Money Act of 2015, with its own charge, its own steep rate and penalty structure and its own one-time compliance window, while the Income-tax Act retained the taxation of disclosed foreign income in the ordinary way and the reporting obligations that feed information to the department. The two statutes are halves of one policy, detection abroad and disclosure at home, and the pillar notes the division so that the anti-avoidance architecture is not misread as a single code.

Two regimes inside one Act

The Finance Act, 2020 inserted section 115BAC, creating an alternative computation for individuals and certain other persons: lower prescribed rates in exchange for surrendering most exemptions and deductions, including the great majority of the Chapter VI-A shelf and the section 10 allowances that structured salaried tax planning. The device had already been used for companies in sections 115BAA and 115BAB of 2019. The Finance Act, 2023 then inverted the default, making the alternative regime the normal computation from assessment year 2024-25, with the older computation available on election for those it still favoured. The episode is a small, complete illustration of this article’s thesis. A policy argument about simplification was conducted entirely through annual amendment, the statute acquired a second parallel code inside itself rather than a cleaned single one, and the taxpayer’s annual question became not what the Act says but which of its two versions applies to him, a complexity the older structure never imposed. When the replacement statute arrived, it carried forward the default regime as its own basic computation, ratifying the direction of 2023 while discarding the dual structure that produced it.

The dual regime is also the end of a longer arc that a pillar should record, the arc of the rates themselves. The Act began in an era of confiscatory marginal rates at the top, reaching, with surcharge, 97.75 per cent in the mid 1970s, a figure that functioned less as revenue policy than as a standing invitation to evasion and to the black economy the state then legislated against elsewhere. Successive decades lowered the top rate in steps, through the reforms of the 1970s associated with the Wanchoo Committee’s diagnosis of evasion, the moderate structure of the 1990s, and the corporate rate reductions of 2019 that traded exemptions for a lower headline rate, until the structure stabilised at levels intended to make compliance cheaper than concealment. The lesson the arc teaches is that rates in this Act are never merely rates. They are the point where fiscal need, political economy and the enforceability of the rest of the code meet, and the draftsmen of each generation set them by judging how much of the code’s machinery the year’s rates would leave standing.

The succession: replacement in 2025 and the law of the transition

The replacement had been attempted before. A Direct Taxes Code was drafted in 2009, introduced as a Bill in 2010 and examined by the Standing Committee on Finance, before lapsing with the dissolution of the House in 2014, and a second task force exercise later in that decade produced a report that was never enacted. That earlier history, the code that failed and the reasons it failed, belongs to the article on the Direct Taxes Code Bill, which owns the replacement story in its political dimension. What belongs to this pillar is the narrower and more useful point: the successful replacement, when it came, chose the opposite method from the failed one. The 2024 Union Budget announced a comprehensive review of the 1961 Act aimed at making its text concise, lucid and easy to read and understand, to be completed in six months, and the exercise was conducted inside the revenue administration rather than as a public renegotiation of tax policy. The Income-tax Bill that followed in February 2025 passed through a select committee, was withdrawn and reintroduced in revised form, and was enacted as the Income-tax Act, 2025, Act 30 of 2025, assented to on 21 August 2025 and in force from 1 April 2026, with 536 sections in twenty-three chapters and sixteen schedules and a text roughly half the word count of the closing 1961 text.

The making of the replacement differed from the making of its predecessor in one respect worth recording. The Bill introduced in February 2025 was referred to a Select Committee of the Lok Sabha chaired by Baijayant Panda, which examined it clause by clause, received the administration’s explanations and returned several hundred recommendations, the great majority of which the government accepted when the Bill was withdrawn and reintroduced in revised form in August 2025. A tax code was therefore, on this occasion, given the committee scrutiny the annual Finance Bill process denies to tax amendments, and the episode stands as a quiet institutional verdict on the mechanism described earlier in this article: when Parliament wished to legislate about the tax system as a whole, it reached for the deliberative instrument, not the budget vehicle. The February Bill and the August Act differ in particulars the committee record explains, and the Act as passed carries 536 sections against the Bill’s original scheme, renumbered through the committee’s work.

The new Act’s own account of itself, which the record supports, is continuity of policy with simplification of language and structure. It preserves the five heads of income, the residence-based scope, the collection machinery and the appellate ladder, reorganises them, renumbers them, converts provisos and explanations into numbered sub-sections and tables, replaces the paired concepts of previous year and assessment year with a single tax year, and carries the faceless architecture into the statutory text as the default. It does not rewrite the rates, which rest where the Finance Acts had placed them, and it does not renegotiate the distribution of reliefs. A reader moving from the old Act to the new is therefore learning a new map of familiar territory, and the profession’s transitional labour consists mostly of concordance, matching each familiar provision to its new number and confirming that the words, where changed, changed the law or only its expression. Where Parliament intended no policy change, courts will be asked for years to decide whether altered words in a simplification statute altered meaning, the same question the courts of the 1960s answered when the 1922 Act’s language was carried into the 1961 Act, and the answers will draw on the older case law precisely because the concepts survived.

The transitional position itself can be stated cleanly, and it answers the question readers ask most in this territory. The repeal of the 1961 Act by the 2025 Act is accompanied by savings of the standard and deliberate kind. The 1961 Act continues to govern the charge, assessment, appeal, penalty and prosecution for every previous year ending on or before 31 March 2026. A return for the financial year 2025-26 is filed under the 1961 Act, an assessment for that year is made under it, a notice of reassessment for an earlier year proceeds under it, and an appeal filed in 2028 against an assessment of a pre-succession year is heard under its provisions, by authorities applying its text as it stood for the year in issue. Rights accrued, liabilities incurred and proceedings pending under the repealed Act are preserved, on the pattern of the savings that attended the repeal of the 1922 Act in 1962 and of the general savings doctrine of the General Clauses Act. The two statutes therefore run in parallel for a long transitional period, the new one taxing the future and the old one litigating the past, and the working lawyer’s shelf will carry both for the professional lifetime of anyone practising at the succession.

The transition reaches below the statutes to the subordinate layer. Rules framed under the 1961 Act, forms notified under it and circulars issued in its administration continue so far as they are consistent with the successor, on the ordinary savings principle, until the new administration replaces them, and the department’s electronic systems, the filing portal, the information statements and the permanent account number itself, pass across unchanged, since the citizen’s identity in the system was never a creature of the 1961 Act alone. Even the name survives in shortened public usage, the department remaining the Income Tax Department and the charge remaining income tax, which is a reminder that what a repeal of this kind actually changes is the organisation of a continuing institution’s law. The reader who grasps the transition in these terms will not be misled by headlines in either direction, neither by the claim that an era ended on a single April morning nor by the claim that nothing changed because the tax did not.

Reading the Act: versions, commentaries and the working text

A pillar on this statute owes the reader one practical education that bare-act sites never provide, which is how the Act is actually read by those who work with it. Nobody applies the Income-tax Act from a single printed text, because the question is never what a section says but what it said for the previous year in issue, and the working apparatus of the profession is built around that fact. The standard commentary tradition, of which the treatise founded by Kanga and Palkhivala is the senior example, tracks each section’s legislative history, the Finance Act that last touched it and the case law that construes it, and is consulted in that order: text, history, authority. Annotated bare Acts print the amendment trail in footnotes for the same reason. Circulars and instructions of the Board are read alongside the sections they gloss, both because officers are bound by the beneficial ones and because the boundary between interpretation and legislation by circular is itself litigated. Judgments are read with their years attached, because a decision of 1975 construes a text that four Finance Acts may have altered since, and citing it against the altered text is the beginner’s characteristic error in this field.

The discipline has a direct consequence for how the succession will be lived. The 2025 Act arrives with no commentary tradition, no amendment trail and no case law, and its early readers will work with concordance tables mapping its sections back to the 1961 provisions they restate, reading the new text through the old one’s accumulated meaning until the new text accumulates its own. That dependence is not a weakness of the replacement. It is the normal condition of a consolidated statute in a system where meaning is made jointly by Parliament and the courts over time, and it confirms this article’s opening claim. The 1961 Act’s real content was never contained between its own covers. It lived in the trail of amendments, circulars and judgments that surrounded the text, and a reader who understands that understands why the Act could be at once unreadable and indispensable, and why its history, rather than any summary of its sections, is the knowledge that transfers to its successor.

The honest assessment: a text that failed its readers, and a law most readers never meet

Two counter-readings must be taken seriously before a verdict, because the standard account of the 1961 Act is true and misleading in opposite directions at once. The standard account says the Act became unusable through complexity. Of the text, that is true, and the state’s own replacement rationale said so: the Finance Ministry defended the 2025 Act on the ground of the old Act’s volume and difficulty, expressly disavowing any major policy change, which is an unusually honest stated reason for replacing a statute and is taken at face value in this article. But the account misleads about the law as experienced. The sediment is not evenly distributed. The provisions that generate the bulk of the 819 sections, the special regimes, the inserted chapters for classes of transactions most taxpayers never enter, the anti-avoidance codes, the international provisions, touch a small minority of assessees. An ordinary salaried taxpayer’s entire relationship with the Act runs through a handful of provisions: the salary charge, a short list of exemptions and deductions, withholding by the employer, the return, and processing. The unreadable statute and the workable tax system coexisted for decades because they were, for most people, different objects, and reform rhetoric that ignores the distinction promises citizens a simplicity of experience that a simpler text alone cannot deliver.

The second counter-reading concerns motive, and it is the perennial argument that Indian tax law is kept complex because complexity sustains a compliance industry of advisers, representatives and intermediaries whose livelihoods depend on the citizen’s inability to proceed alone. The argument deserves its fair hearing. Complexity does create that industry, the industry is large, and every simplification exercise from the Tax Administration Reform Commission of 2014, chaired by Parthasarathi Shome, which criticised the volume, unpredictability and retrospectivity of tax amendments, to the review of 2024 has had to reckon with the profession’s ambivalence toward a code its clients could read unaided. The fair answer distinguishes effect from design. The record assembled in this article shows complexity arising from a mechanism, the annual Finance Act applied to an expanding economy’s transactions, each year’s insertion defensible in its own terms, rather than from any demonstrated intention to manufacture work for advisers. A conspiracy would have produced a more uniform obscurity. What the record shows is a legible core buried under specialised accretion, which is the signature of legislative workload, not of design against the reader. The compliance industry is a beneficiary of the sediment. The sediment’s cause is the budget process, and remedies aimed at the beneficiary rather than the process have failed accordingly.

A third measure belongs in the assessment, the litigation record, stated as a pattern because precise pendency figures vary by forum, period and source and are contested whenever quoted. Across the Act’s later decades the appellate system carried a backlog that the department’s own reporting placed in lakhs of pending appeals across its forums, disputed demands formed a significant fraction of the tax collected in the years they arose, and the revenue’s success rate in the appeals it pursued was low enough that governments of both parties treated the reduction of litigation as a stated policy goal, raising the monetary thresholds for departmental appeals, creating dispute settlement schemes and, in the taxpayers’ charter adopted in 2020, promising faceless and time-bound disposal. The pattern matters to a pillar because it qualifies the design described in the earlier sections. The ladder is well built. Its landings are crowded, and a reader deciding what an assessment is worth must discount it by the years the ladder takes to climb, as every practitioner advising on a disputed demand already does.

The faceless experiment concentrates both readings in one institution. Its defenders, led by the revenue administration that designed it, state the case at its strongest: personal interface between taxpayer and officer was the historical site of discretion, delay and rent-seeking, allocation by algorithm and anonymity by design removed the opportunity, and national distribution of work ended the tyranny of the local jurisdiction. Its critics, including the associations of tax practitioners who challenged and commented on the scheme, state the opposing case at its strongest: complex assessments turn on explanation, documents acquire meaning in conversation, and a system in which a case is read in fragments by unseen officers who cannot be addressed produces orders that recite the record without engaging it, multiplying appeals rather than resolving disputes. Both accounts describe real effects of the same design, and the evidence available, appeal volumes and the department’s own reporting, does not yet crown either. What can be said is that faceless assessment is the first reform of the Act that changed the experience of the ordinary assessee more than the text of the law, and that its absorption into the replacement statute as the default settles the institutional question while leaving the evaluative one open.

One further honesty is required about where simplification actually reached the citizen during the old Act’s last decades, because it did not arrive through the statute at all. The pre-filled return, the information statement placed before the filer, the centralised processing that turned most assessments into an automated reconciliation, the faceless interface: the lived simplification of Indian income tax was delivered by administration and technology working under the unreformed text, while the text itself grew harder. That sequence carries a warning for the successor statute’s reception. A clean text is a genuine good, and this article has treated it as one. But the citizen’s experience of the tax will continue to be decided by systems, selections and interfaces, and a reform that is judged five years on will be judged on those, on which the new Act’s words are a necessary but distinctly partial influence.

The verdict of this pillar follows from its spine. The Income-tax Act 1961 was a competently drafted code destroyed as a readable text by the process constitutionally assigned to maintain it, and it remained, throughout its unreadability, a functioning system that collected the Union’s largest revenue, sustained a sophisticated appellate jurisprudence, and gave most of its subjects a workable annual routine. Its replacement is a genuine simplification of language and arrangement and no simplification of policy, and readers who understand that distinction will not be disappointed by the 2025 Act, because they will not have expected it to change what they pay. The cluster’s other articles take the story forward from here: the failed code, the retrospective amendment, the appeals institution and the separate foreign assets regime each own their ground, and this pillar’s work is done when the reader can hold the whole structure, charge, heads, machinery, appeals, deterrents and anti-avoidance, in one view, and can say which statute governs any given year and why.

Frequently Asked Questions

Q: What did the Income Tax Act 1961 replace and when did it come into force?

The Income-tax Act 1961 replaced the Income-tax Act 1922, the code enacted under British rule that had governed income tax in India for four decades and continued in force after independence. The 1961 Act is Act 43 of 1961. It received the assent of the President on 13 September 1961 and came into force on 1 April 1962, the first day of the financial year 1962-63, applying from assessment year 1962-63. Its drafting rested on the Law Commission’s revision work of 1958 and the report of the Direct Taxes Administration Enquiry Committee chaired by Mahavir Tyagi, and its repeal of the 1922 Act carried the standard savings for proceedings relating to earlier years, the same pattern its own repeal would follow in 2026.

Q: What are the five heads of income under the Income Tax Act?

Section 14 of the Income-tax Act 1961 classifies all income under five heads: salaries; income from house property; profits and gains of business or profession; capital gains; and income from other sources, the residuary head for receipts that no specific head captures. The classification is not a filing label. Each head has its own charging section, its own computation rules and its own deductions, and the same receipt can bear a different tax depending on the head under which it falls. A receipt that is not income in character, or that no charging provision reaches, is not made taxable by classification alone, which is why disputes about the correct head, and about whether a receipt is income at all, dominate assessment litigation.

Q: How is the Income Tax Act amended every year?

The Act is amended through the annual Finance Act. Each year the budget is presented to Parliament with a Finance Bill, classified as a Money Bill, and the resulting Finance Act does two things at once. It fixes the rates of income tax for the coming year in its First Schedule, because the charging provision of the Income-tax Act charges tax at the rates the Finance Act prescribes. And it amends the Income-tax Act in substance, inserting, substituting and omitting sections across the code. This is why the Act accumulated lettered sections, inserted chapters, and provisos to provisos over six decades: structural amendment rode on the annual budget vehicle, driven by the budget timetable, rather than travelling as standalone legislation examined at leisure.

Q: What is GAAR and when did it take effect?

GAAR is the general anti-avoidance rule in Chapter X-A of the Income-tax Act, sections 95 to 102, enacted by the Finance Act 2012. It empowers the department to declare an arrangement an impermissible avoidance arrangement where the main purpose of the arrangement is to obtain a tax benefit and it bears a tainted element, such as lacking commercial substance or creating rights not ordinarily created between persons dealing at arm’s length, and to deny the benefit, disregard parties or reallocate income accordingly. Enacted in 2012, it was deferred in stages following the recommendations of the expert committee chaired by Parthasarathi Shome, and it took effect from assessment year 2018-19. Arrangements producing an aggregate tax benefit below three crore rupees are excluded by rule, and investments made before 1 April 2017 are grandfathered.

Q: How does faceless assessment change an income tax proceeding?

Faceless assessment, given statutory form in section 144B operating from 1 April 2021, removes the territorial Assessing Officer and the personal interface from scrutiny assessment. The case is allocated through national and regional faceless centres without regard to where the taxpayer is located, the officers who examine the return, issue notices and prepare the draft order do not know the assessee’s identity, and the assessee never meets them. All communication travels through the electronic portal, and draft orders pass through verification, technical and review units before being finalised. The design aims to eliminate discretion and the rent-seeking that personal contact made possible. Its critics answer that a taxpayer’s ability to explain a complex case to a single accountable reader is lost in exchange, and both effects are part of the recorded experience of the scheme.

Q: Is the Income Tax Act 1961 still the governing law?

For income of periods up to 31 March 2026, yes. The Income-tax Act 2025, Act 30 of 2025, operates from 1 April 2026 and governs the tax years beginning on or after that date. The repeal of the 1961 Act is accompanied by savings under which the charge, assessment, appeal, penalty and prosecution for every earlier period continue under the 1961 Act, in the version that stood for the year in issue. The 1961 Act therefore remains the governing law for a large and slowly diminishing stock of past-year matters, and it will be applied by officers, tribunals and courts for many years. For any period beginning on or after 1 April 2026, the governing statute is the 2025 Act, which preserves the substance of the older law in reorganised and renumbered form.

Q: What happens to assessments pending under the Income Tax Act 1961?

They continue and conclude under the Income-tax Act 1961. The repeal and savings provisions of the Income-tax Act 2025 preserve rights, liabilities and proceedings under the repealed Act for periods governed by it, following the pattern of the savings that attended the repeal of the 1922 Act in 1962 and the general doctrine of the General Clauses Act. An assessment pending for a year up to the financial year 2025-26 is completed under the 1961 Act, and appeals, revisions, reassessments and recovery arising from it travel through the 1961 Act’s machinery and appellate ladder, applying the text of that Act as it stood for the relevant year. Nothing is transferred to the new statute mid-proceeding, and no pending matter has to be begun again under the 2025 Act.

Q: Why did the Income Tax Act 1961 become so difficult to use?

Because it was amended in substance almost every year for over six decades, and amendment proceeded by insertion rather than renumbering. The Act began with 298 numbered sections in twenty-three chapters. Successive Finance Acts inserted lettered sections and whole lettered chapters, added provisos to provisos and explanations to explanations, and left expired incentives and transitional provisions in the text because removing dead law required legislative effort nobody was rewarded for. The Finance Ministry’s own count at the replacement put the closing text at 819 effective sections across forty-seven chapters and about 5.12 lakh words. The government’s stated reason for replacing the Act was this difficulty of use rather than disagreement with its policy, which is why the successor statute simplifies language and structure while preserving the substance of the tax.

Q: What is the difference between the previous year and the assessment year?

The Income-tax Act 1961 runs on two linked years. The previous year is the financial year, 1 April to 31 March, in which income is earned, and it is the income of that year that is charged. The assessment year is the year immediately following, in which that income is assessed, the return for it is filed and the tax on it is finally determined. Income earned in the financial year 2024-25 is therefore assessed in assessment year 2025-26. The charging section ties the two together by charging tax for each assessment year on the total income of the previous year. The Income-tax Act 2025 abolished the pair and substitutes a single tax year, which is the year of earning itself, but the dual vocabulary governs every matter arising under the 1961 Act.

Q: Why does it matter which head of income a receipt falls under?

Because the heads are computation codes, not labels. Each of the five heads in section 14 carries its own charging section, its own definition of the receipts it captures, its own list of allowable deductions and its own treatment of losses, and several carry special rates. A gain on shares assessed as a capital gain is computed after deducting acquisition cost and taxed at the capital gains rates; the same gain assessed as business income is taxed as a profit at ordinary rates with business deductions available and business loss rules applying. Interest taken under house property, business or other sources is computed three different ways. Classification is therefore decided first in most serious assessments, the department and the taxpayer frequently argue for different heads, and the argument is worth money in both directions.

Q: What is the difference between an exemption and a deduction under the Income-tax Act?

An exemption keeps a receipt out of the computation entirely, while a deduction allows a payment or outlay to be subtracted after total income is computed. Exempt receipts, principally those listed in section 10, never form part of total income: agricultural income is the leading example. Deductions, principally those in Chapter VI-A, sections 80C to 80U, are allowed from total income up to prescribed ceilings and on prescribed conditions, such as specified savings, insurance premia, medical insurance and donations. The difference has practical consequences. An exemption is not reported as income and attracts no ceiling, a deduction must be claimed, proved and fitted within its limit, and the alternative tax regime introduced in 2020 defined itself by requiring the taxpayer to surrender most deductions and several exemptions in exchange for lower rates.

Q: How does a person’s residence status change the income India can tax?

Residence is the switch that sets the territorial reach of the charge. Under section 5, a person resident in India is taxed on worldwide income: income received in India, income accruing in India and income accruing anywhere else in the world. A person who is not ordinarily resident is taxed on foreign income only where it comes from a business controlled in India or a profession set up in India. A non-resident is taxed only on income received in India or accruing or deemed to accrue in India under section 9, which covers Indian business connections, Indian assets, salaries for services rendered in India and similar sources. Residence itself is determined under section 6 mainly by counting days of physical presence in India across the relevant years, with separate rules for companies, so the same person can move between the three reaches of the charge as his movements change.

Q: What is tax deducted at source and why does the Act depend on it?

Tax deducted at source, governed by Chapter XVII of the Act, obliges the payer of specified income, including salary, interest, dividends, rent, professional fees and commission, to deduct tax at prescribed rates at the time of payment or credit and to deposit it with the government. The payee then claims credit for the deduction against his final liability in assessment. The Act depends on the mechanism for three reasons. It collects most individual revenue in advance, spread across the year rather than after it. It creates a reporting trail, since every deduction is recorded against the payee’s permanent account number and visible to the department. And it enforces itself, because a payer who fails to deduct faces interest, fees, penalty and disallowance of the expense itself, which makes payers diligent collectors on the state’s behalf.

Q: When can the Income Tax Department reopen an assessment that is already complete?

Under sections 147 to 151, where the Assessing Officer has reason to believe that income chargeable to tax has escaped assessment for a year, the completed position may be reopened by notice within the limitation periods the Act prescribes: three years from the end of the relevant assessment year in the ordinary case, and a longer period extending to ten years where the escaped income is substantial and represented in the form of an asset, expenditure or entries in books. Since the recast by the Finance Act 2021, the officer must first conduct an enquiry where required and pass an order under section 148A after giving the assessee a show-cause opportunity, deciding whether the case is fit for reopening before any notice issues. Information suggesting escapement, such as unexplained deposits or mismatches with reported transactions, is the usual trigger, and the courts supervise the sufficiency of the material and the fairness of the procedure.

Q: What is transfer pricing and why does the Income-tax Act regulate it?

Transfer pricing is the pricing of transactions between associated enterprises of the same group, and the Act regulates it in Chapter X, sections 92 to 92F, inserted by the Finance Act 2001, because internal prices are set by the group itself rather than by a market. A multinational can shift profit out of India by overcharging its Indian company for imports, services or royalties, or undercharging it for exports, leaving the Indian tax base thinner than the Indian activity warrants. The chapter therefore requires income from an international transaction to be computed at the arm’s length price, the price unrelated parties would have agreed, prescribes methods for finding that price, and imposes documentation, reporting and certification duties. Disputes about comparable companies and adjustments under this chapter form one of the largest and most technical bodies of litigation under the Act.

Q: What is the difference between tax evasion, tax avoidance and tax planning?

The three terms mark points on a spectrum the courts and the statute treat very differently. Tax evasion is illegal conduct: concealing income, fabricating deductions, keeping false books, and it attracts reassessment, penalty and prosecution. Tax planning is the lawful arrangement of affairs to use reliefs Parliament intended to be used, such as claiming deductions, choosing a business form or timing a transaction, and it is not actionable. Tax avoidance in the contested middle sense means arrangements that comply with the letter of provisions while defeating their purpose to obtain a tax benefit. Indian courts moved from condemning such arrangements in the McDowell decision of 1985 toward respecting lawful arrangements in the Azadi Bachao Andolan and Vodafone decisions, and Parliament then answered with a statutory line in the general anti-avoidance rule of 2012, which targets arrangements whose main purpose is a tax benefit and which bear tainted elements such as a lack of commercial substance.

Q: Who must file a return of income under the Income-tax Act?

Section 139 obliges every company and every firm to file a return regardless of income, and obliges other persons, including individuals, Hindu undivided families, associations and bodies, to file where their total income, computed before allowing specified exemptions and deductions, exceeds the basic exemption limit for the year. Filing is also compulsory in defined situations regardless of income, and the Act and rules have extended mandatory filing to specified high-value transactions and circumstances, so that a person below the income threshold may still be required to file. Many others file voluntarily to claim refunds of tax withheld, to carry forward losses, which generally requires a timely return, or to maintain a compliance record for loans, visas and tenders. Failure to file attracts a late fee, interest, loss of carry-forward rights and, in cases of wilful default involving substantial tax, exposure to prosecution.

Q: How does the Income-tax Act tax agricultural income?

Agricultural income is excluded from the Union’s charge. The Constitution assigns taxes on agricultural income to the states, and Parliament’s power extends to income other than agricultural income, so section 10(1) of the Income-tax Act exempts agricultural income, which section 2 defines around rent and revenue from land used for agricultural operations and the processes and sales connected with them. The exclusion is not complete in its effects for individuals, Hindu undivided families and certain other assessees, because since the 1970s the Act applies partial integration: agricultural income is added to other income for the sole purpose of determining the rate at which the other income is taxed, then the tax attributable to the agricultural slice at the threshold is deducted. The result is that a farmer’s agricultural income bears no central tax, while a taxpayer with both agricultural and other income pays a higher rate on the other income than he would without it.

Q: What happens to court decisions on the 1961 Act after the 2025 Act replaces it?

They continue to govern the 1961 Act, and they remain the starting point for the 2025 Act wherever the underlying concept survived. Judgments interpreting the repealed Act bind all proceedings for periods under it, which will continue for years, because precedent attaches to the statute being applied. For the successor statute, the position follows the principle the courts used in 1962: where the new Act reproduces the substance of an old provision, the settled construction of that provision is persuasive and ordinarily travels with the concept, and the 2025 Act preserves most concepts, including the five heads, the residence scope and the assessment machinery. Where the new Act deliberately changes the words in a simplification exercise that disavows policy change, courts will have to decide case by case whether the change is one of expression or of meaning, and litigants will cite the 1961 case law on both sides of that question.

Q: What does the Central Board of Direct Taxes do under the Income-tax Act?

The Central Board of Direct Taxes, working under the Ministry of Finance, is the apex administrative authority for direct taxes, constituted under separate legislation of 1963, and the Income-tax Act works through it. The Board directs and superintends the Income Tax Department whose officers make assessments, collect tax and conduct searches and surveys under the Act. It issues circulars and instructions binding on the department, a power the courts recognise while holding that circulars cannot override the Act itself, and it frames the schemes and notifications through which much of the Act’s machinery, including the faceless assessment scheme, is operated. It also advises the Finance Ministry on the annual amendments that arrive through the Finance Bill. The Board is thus the hinge between the statute Parliament enacts and the administration taxpayers experience, which is why its circulars are themselves frequently litigated.