For a finance director or a company secretary, CSR under Section 135 of the Companies Act, 2013 presents itself as a deceptively simple instruction: spend two per cent of average net profits on a list of approved activities. Almost everything about that sentence is more complicated than it looks. The two per cent is computed on a profit figure that the Act defines for itself and that matches neither the tax computation nor the number at the foot of the profit and loss account. The approved activities are broad in words and narrow in exclusions. The obligation is triggered by any one of three financial tests, so a modest business with a single profitable year can be caught while a far larger business trading at a loss escapes. Above all, the consequence of failing to spend has been rewritten twice in quick succession, and a large share of the guidance in circulation still describes a legal position that either never operated or has since been replaced.

This article is written for the reader who has to operate the provision rather than merely describe it. It works through the questions in the order a compliance team meets them. Does the obligation attach to this company at all? On what profit figure is the amount calculated, and who inside the company owns each step? What may the money lawfully buy, and what does it stop being once it is spent through an agency or locked into an asset? What must happen, by what date, to any part of the amount that remains unspent when the financial year closes? And what precisely follows if the company gets any of this wrong? The answers changed materially between 2019 and 2021, so the article treats the amendment record as part of the law itself, because a provision that has moved from explanation to escrow to penalty inside seven years cannot be understood from its text alone.

What Section 135 built, and why it had no predecessor

Section 135 sits inside the Companies Act, 2013, the statute that rebuilt Indian company law around disclosure, board accountability and class-wide obligations rather than government permission. Within that architecture the CSR provision was the outlier. The rest of the Act largely tells companies how to govern themselves and what to reveal. Section 135 tells a defined class of companies how much of their profit to spend, and on what. Nothing in the Companies Act, 1956 had attempted anything of the kind, and the contrast between the two statutes is sharpest here: where the older law left corporate giving to board discretion and shareholder tolerance, the new one converted a slice of it into a statutory charge on profits, which is why the provision is best approached through the comparison of the Companies Act 1956 and the Companies Act 2013 rather than as a footnote to either.

The claim most often made about the section is that India became the first country to make corporate social responsibility spending a legal obligation rather than a disclosure duty or a voluntary code. The claim is widely reported and is repeated by the Ministry of Corporate Affairs itself, and it captures something real: no earlier Indian law, and few laws anywhere, had fixed a percentage of profits and attached it to a schedule of permitted social purposes. It should still be stated with care. Mauritius had operated a levy-based CSR charge on companies several years before Section 135 commenced, collected through its revenue authority, and several jurisdictions had mandatory disclosure regimes that stopped short of mandatory spending. India’s distinction lies in the combination of a profit-linked quantum, a statutory list of purposes and enforcement through company law. That combination was new, and it is the combination, rather than an unqualified claim of world priority, that the evidence supports.

The section was brought into force on 1 April 2014, together with the Companies (Corporate Social Responsibility Policy) Rules, 2014, which supply much of the operating detail. Its design at birth was deliberately soft at the point of enforcement. A covered company had to constitute a committee, adopt a policy, and spend the prescribed amount, but if it failed to spend, the statute asked only that the board explain the shortfall in its report to shareholders. That comply-or-explain architecture reflected the hesitation of the drafting moment: Parliament was willing to name a figure and a list of purposes, but not yet willing to treat an unspent balance as a default. The history of the provision since then is the history of that hesitation being withdrawn in stages, first by requiring the money to be moved into statutory custody when it was not spent, and then by attaching a money penalty to the failure to move it.

From voluntary guidelines to a statutory charge

The section did not arrive unheralded, and its pre-history explains its original softness. In 2009 the Ministry of Corporate Affairs issued Corporate Social Responsibility Voluntary Guidelines, inviting companies to adopt CSR policies and report against them, and in 2011 the ministry followed with the National Voluntary Guidelines on Social, Environmental and Economic Responsibilities of Business, a principles framework addressed to Indian business as a whole. Both instruments were exhortatory. They assumed that disclosure and peer practice would move boards, and for a small number of large business houses with established foundations, they described what was already happening. For the wider corporate sector they changed little, and the legislative judgment taken in the Companies Bill was that exhortation had had its trial.

The Bill that became the 2013 Act carried the two per cent proposition through a long parliamentary passage, and the design that emerged bears the marks of the argument it survived. Industry representations resisted a hard spending mandate as a tax on profits administered by boards; supporters of the provision answered that a purely voluntary regime had been measured and found thin. The compromise written into the enacted section was to fix the quantum and the purposes in law while leaving the immediate consequence of a shortfall at the level of explanation to shareholders. The parliamentary standing committee that examined the Bill pressed on the reality of the obligation rather than its symbolism, and the enacted text’s insistence that the board report state reasons for any failure to spend was the residue of that pressure: the explanation was meant to be an uncomfortable public act, not a filing formality.

Within eighteen months of commencement the ministry had appointed a High Level Committee under Anil Baijal, a former Home Secretary, to review how the framework was working and to suggest measures for monitoring. The committee’s existence so early in the regime’s life is itself a finding: the state that had legislated the obligation was unsure how to watch it. The questions left open at enactment, whether explanations would discipline boards, whether the money would reach beyond company towns, and whether anyone in government owned the data, ran through the regime’s first five years and were taken up again, with sharper instruments, by the second High Level Committee of 2018 whose recommendations are traced in the discussion of penalties below. A compliance reader should hold on to this arc, because it explains why the provision’s enforcement history is one of successive tightening drafted in committee rooms rather than in courtrooms: there is almost no case law on Section 135, and the operative interpretation of the section sits in ministry circulars, frequently asked questions and committee reports, administrative materials that bind practice without ever having been tested as precedent.

The first years in practice: administration by clarification

Because the section arrived with a short text and a long tail of practical questions, its early law was made by the ministry’s circulars and frequently asked questions rather than by courts, and a reader who wants to understand the operative regime has to read that administrative layer as law in action. The clarifications issued in the first phase settled the questions that boards actually asked. One-off events, award functions, sponsorship-linked marathons and similar episodic activities were stated not to be CSR, however public-spirited the occasion, because the obligation is to undertake projects and programmes, not to buy presence at events. Activities that a company performs for its own workforce were confirmed outside the definition. The treatment of salaries paid to staff engaged on CSR work, the eligibility of contributions to state-level funds, the position of companies in their first years, the permissibility of building capacities within agencies: each was answered administratively, in instruments that companies cite in board papers and auditors test against, and which have never been replaced by a body of precedent. The absence of litigation is itself a fact about the provision’s design. A spending duty with an explanation as its only sanction generated few disputes worth a writ petition, and the interpretive tradition of the section is accordingly a file of circulars rather than a shelf of reports.

The administrative layer also reveals the ministry’s recurring anxiety, which is about conversion: the conversion of business expenditure into compliance, of employee welfare into social spending, of marketing into philanthropy, of a statutory duty under another law into a second discharge under this one. Nearly every clarification of the first phase is an anti-conversion rule, and the 2021 revision of the rules wrote most of them into the definition of CSR itself, which is why the exclusions in the rules read like a catalogue of attempts already made. A practitioner advising on a doubtful project can use that history as a test: if the structure of the proposal is that expenditure the company has commercial reasons to incur anyway will be counted towards the two per cent, the proposal is walking into the exclusions, whatever Schedule VII head is written on the file.

How Section 135 decides which companies must spend on CSR

The gate is Section 135(1). Every company that had, during the immediately preceding financial year, a net worth of Rs 500 crore or more, or a turnover of Rs 1,000 crore or more, or a net profit of Rs 5 crore or more, must constitute a Corporate Social Responsibility Committee and enter the regime. The three tests are alternatives. Meeting any one of them is enough, and failing two of them is no protection. This is the point on which the largest number of published summaries, and a surprising number of board papers, go wrong, because the eye reads a row of thresholds as cumulative when the statute makes them disjunctive.

The profit limb deserves particular attention because it operates independently of size. A company with a turnover of Rs 60 crore and modest assets crosses the gate in a year when its net profit reaches Rs 5 crore, and it then carries the full apparatus of committee, policy, spending and reporting on the strength of that single figure. The converse also holds. A company with a turnover far above Rs 1,000 crore is inside the regime even in a year of losses, because the turnover limb does not ask about profitability at all. Applicability is therefore a three-question inquiry, not a general impression of whether a business is large, and each question is asked about the immediately preceding financial year, so the duty for a given year is fixed by figures that are already audited and closed before the year begins.

The thresholds are stated in the section itself and have not been revised since commencement, which is itself a fact of consequence. A net profit figure of Rs 5 crore captured a certain population of companies in 2014, and the same unindexed figure captures a larger population with every year of nominal growth in profits. Parliament has adjusted many monetary limits in the Companies Act by notification over the years, but the three figures in Section 135(1) stand where they were enacted, so the reach of the CSR obligation widens silently through inflation rather than through any deliberate legislative decision to widen it.

The obligation also follows the corporate group in ways that catch people out. The rules apply the section to holding companies and subsidiaries that meet the tests on their own figures, and a foreign company with a branch office or a project office in India is brought within the rules in respect of its Indian business. What matters in each case is the individual company’s own net worth, turnover and net profit, not the consolidated position of the group, so a group can contain companies on both sides of the line, each with a different compliance position and, where both are covered, each with its own committee, policy and computation.

The corporate form gives no shelter. A company incorporated for charitable objects under Section 8 of the Act is still a company for this purpose, and if its own figures cross a threshold it must constitute the committee and spend like any other covered company, a position that surprises non-profit companies with large grant income until their first computation. Government companies and public sector banks and enterprises organised as companies are inside the section on the same terms, which is why public sector CSR is a visible part of the national total. What stands outside is the limited liability partnership, which is not a company under the Act and carries no Section 135 obligation however large it grows, and the unincorporated firm or trust trading on its own account. The boundary therefore follows incorporation, not economic substance, and two businesses of identical size can face entirely different positions because one is a company and the other is not.

Two of the three gate figures are defined terms of the Act, and the definitions matter at the margin. Net worth carries the meaning in Section 2(57): paid-up share capital, reserves created out of profits, the securities premium account and the balance of the profit and loss account, after deducting accumulated losses, deferred expenditure and miscellaneous expenditure not written off, and excluding reserves created out of revaluation of assets or write-back of depreciation. Turnover carries the meaning in Section 2(91): the gross amount of revenue recognised in the profit and loss account from the sale, supply or distribution of goods and from services rendered, together with other operating revenues and income. A company near either line must therefore test itself against the statutory definitions rather than against the labels in its own management accounts, because items that a finance team casually calls reserves, including a revaluation reserve, do not count towards net worth, and income that swells a management turnover figure may sit outside the Section 2(91) description. The third figure, net profit, is tested on the same reconstructed basis used for the computation of the amount, so the gate and the charge are measured with the same ruler.

The trigger year deserves one clarification that board papers frequently muddle. The tests are applied to the immediately preceding financial year, and they switch the obligation on for the year that follows; a company that first crosses a threshold in the financial year ending 31 March of one year organises its committee, adopts its policy and spends during the year that begins the next day. There is no pro-rating for the trigger year and no grace period while the machinery is built, which is why the first annual action plan of a newly covered company is often adopted some months into its first spending year, regularising in the minutes a programme that had to begin before the paperwork existed.

The same backward-looking structure governed the regime’s own birth. When the section commenced on 1 April 2014, the first spending year was 2014-15, and the average that fixed the first obligations was drawn from the financial years 2011-12, 2012-13 and 2013-14, years in which the duty did not exist and in which no board had kept its profit records with Section 198 exclusions or CSR in mind. The first computations were therefore reconstructions of closed years under a definition those years had never heard of, and the applicability tests for the first year were applied to the figures of 2013-14, a year that had ended the day before the section began. Retrospectivity of that mild, arithmetic kind is built into the formula permanently: every year’s obligation is the creature of three closed years, which is why a company can see its duty coming a full year ahead and why a plea of surprise at the size of an obligation is never available to a board that reads its own audited figures.

Does a company that made a loss still have CSR duties?

A loss in the preceding financial year does not take a company out of Section 135, because two of the three gates ignore profit. A company with a net worth of Rs 500 crore or more, or a turnover of Rs 1,000 crore or more, remains covered after a loss-making year. The loss affects the amount, not the applicability.

The spending figure is two per cent of average net profits across the three preceding years, and a loss year pulls that average down, sometimes to a figure so small that the practical obligation shrinks towards nothing. The governance duties, including the committee and the policy, persist while the gate is met.

Release from the regime is correspondingly slow. The rules provide that a company which ceases to meet the three criteria for three consecutive financial years is not required to comply with the section until it meets the criteria again. A single bad year therefore changes very little. A company that slips below every threshold must stay below them for three years in succession before the committee can be stood down and the reporting stopped, and one strong year inside that window restarts the count. The design is asymmetric in a way compliance teams should note: entry into the regime takes one year, exit takes three.

The two per cent: how the amount is actually computed

Section 135(5) obliges the board to ensure that the company spends, in every financial year, at least two per cent of the average net profits made during the three immediately preceding financial years. Where a company has not completed three financial years since incorporation, the average is taken over the financial years it has completed. Three features of that sentence carry the practical weight. The base is an average, so a single exceptional year is diluted by its neighbours. The period is the three years immediately before the spending year, so the obligation for a year is knowable in full on the day the year opens. And the profit figure is a statutory construct, because the section directs that net profit be calculated in accordance with Section 198 of the Act.

Section 198 is the provision governing the calculation of profits for managerial remuneration, and borrowing it for CSR imports its whole apparatus of credits and exclusions. Capital profits, profits from the sale of an undertaking, revaluation surpluses and similar items that pass through the accounts are stripped out, because Section 198 is designed to approximate the profit earned from operations rather than the profit produced by transactions in the company’s own assets. The CSR rules then add exclusions of their own to the definition of net profit: any profit arising from an overseas branch of the company is left out, and any dividend received from another Indian company that is itself covered by and compliant with Section 135 is left out, so that the same profit is not charged twice as it moves up a group chain. The consequence is that the figure on which two per cent is levied is routinely different from the profit after tax in the financial statements, and a computation that starts from the wrong base infects everything built on it.

A worked example makes the mechanics concrete. Take a company whose net profits, computed under Section 198 and after the rule-based exclusions, were Rs 120 crore, Rs 80 crore and Rs 100 crore in the three preceding financial years. The average is Rs 100 crore, and the obligation for the spending year is Rs 2 crore. If the middle year had instead been a loss of Rs 20 crore, the average would fall to Rs 66.67 crore and the obligation to about Rs 1.33 crore. The example also shows why the applicability question and the computation question must never be merged: the company in the second version is plainly inside the regime, because its profits in the trigger year crossed Rs 5 crore, but the amount it must spend is set by the three-year average, not by the year that brought it through the gate.

The Section 198 adjustments are where most computation errors live, because the exercise looks like copying a number and is actually a reconstruction. The starting point is the net profit shown by the financial statements prepared under the Act. Onto that figure the computation applies the discipline of Section 198: sums that represent gains on capital account, including profits on the sale of an undertaking or of fixed assets of a capital nature and premiums on shares, are excluded from credit, while the ordinary charges of running the business are recognised. The practical consequence is that a company which has sold a division at a large book profit in one of the three years will find its CSR average far below its reported average, and a company whose reported profit is depressed by items that Section 198 treats differently will find the reverse. Professional guidance issued for the regime reads the rules’ definition together with Section 198 in exactly this way, and the computation is conventionally presented to the board as a reconciliation from the audited profit, item by item, rather than as a single unexplained figure.

Three recurring situations test the reconciliation. A company with carry-forward losses may show a small accounting profit while its Section 198 figure is smaller still or negative, and the obligation follows the statutory figure, not the sentiment of the year. A company that has received a large dividend from a subsidiary that is itself CSR compliant excludes that dividend, but a dividend from a subsidiary below the thresholds, or from a foreign subsidiary, stays in the base, so the same group can have very different effective rates at different levels. And a company whose profits arrive through an overseas branch excludes the branch profit altogether, even though the branch’s turnover counts in the company’s accounts for other purposes. Each of these is a designed outcome of the rule-based exclusions, and each is invisible to a reader who begins from the headline profit after tax.

Two further points of computation recur in practice. The two per cent is a floor for spending, not a budget ceiling, and amounts spent beyond it are not wasted effort in law, because a later amendment allows an excess to be carried forward in defined circumstances. And the amount is an obligation of the company for the financial year, so it must be provided for in the accounts of that year whether or not it has been disbursed by the balance sheet date, which is why the treatment of unspent balances, taken up below, is an accounting event as much as a legal one.

The committee, the board and the annual action plan

The institutional centre of the regime is the Corporate Social Responsibility Committee of the board. Section 135(1) requires a committee of three or more directors, of whom at least one must be an independent director. A company that is not required to appoint an independent director under the Act may constitute the committee without one, and the rules make parallel relaxations for classes of companies whose boards are small. The committee’s statutory functions under Section 135(3) are to formulate and recommend a CSR policy indicating the activities to be undertaken, to recommend the amount of expenditure to be incurred, and to monitor the policy from time to time. Section 135(4) then places the decision with the board: the board considers the committee’s recommendations, approves the policy, discloses its contents in the board’s report and places it on the company’s website.

Since the 2021 revision of the rules, the committee’s work has a defined annual product. The rules require the committee to formulate an annual action plan and recommend it to the board, covering the list of projects approved for the year, the manner of execution, the modalities of utilisation of funds, the implementation schedules, the monitoring and reporting mechanism, and the details of any need and impact assessment for the projects undertaken. The board approves the plan and may alter it during the year on the committee’s recommendation. This converted CSR from a policy statement refreshed occasionally into a budgeted programme settled in advance, and it is the document against which the year’s spending, transfers and disclosures are later audited.

Can the board change the annual action plan during the year?

Yes. The CSR rules expressly allow the board to alter the annual action plan at any time during the financial year, on the recommendation of the CSR Committee and on the basis of a reasonable justification recorded for the change. A project can be dropped, substituted or rescaled, and money reallocated between approved projects.

The alteration is a board decision on advice, not an informal management adjustment, so the recommendation and the justification belong in the committee’s and the board’s minutes. What the board cannot do by altering the plan is reduce the total two per cent obligation itself, which is fixed by Section 135(5) whatever the plan says.

One relaxation matters to smaller covered companies. Section 135(9), inserted by the Companies (Amendment) Act, 2020 and operative from 22 January 2021, provides that where the amount to be spent by a company under Section 135(5) does not exceed Rs 50 lakh, the requirement to constitute a CSR Committee does not apply, and the committee’s functions are discharged by the board itself. The figure is the spending obligation, not the turnover or the profit that triggered coverage, so the test is applied to the computed two per cent amount for the year. A company caught by the profit limb with a small average profit will often fall inside this relaxation, which is a deliberate legislative judgment that the full committee apparatus should not be imposed where the money at stake is modest.

Composition has its own set of working rules that company secretaries apply and summaries often skip. The committee is a committee of the board, so its members are directors, and the independent director requirement attaches to the committee because the section wants at least one voice in the room that owes nothing to the promoters or the executive. A company that is not required to appoint an independent director under Section 149 may constitute the committee without one, and the rules carry the logic down to small boards: a private company with only two directors constitutes its committee with those two directors. The committee has no statutory meeting rhythm of its own, unlike the audit committee, and its calendar is set by the compliance clock: it must meet, in substance, when the policy is framed, when the action plan is recommended, when alterations are proposed, when utilisation is reviewed and when the year’s shortfall is classified. A committee whose minutes show a single meeting adopting everything in March is a committee whose paperwork will read badly against the year’s events, because the regime’s later steps, the certification, the classification, the set-off resolution, all presuppose deliberation at the points where the decisions were actually open.

The record-keeping burden around all of this is heavier than the statutory text suggests, because nearly every later step in the regime depends on documents created here: the policy, the action plan, the minutes recording an ongoing project’s approval, the board resolution that elects to set off an excess. A company that keeps its CSR file in the same disciplined way that a litigant keeps a matter file, with the papers ordered, dated and retrievable, is the company that can answer a registrar’s notice without reconstruction; it is the same instinct that leads journalists to hold sensitive material in a private encrypted workspace, a practice described in this account of source protection through VaultBook. The comparison is not decorative. CSR default is usually proved or disproved from the file, not from the bank statement.

The policy itself: what the document must contain

The CSR policy is the constitution of the company’s regime, and the rules prescribe its contents closely enough that a policy can be audited against them. Under the rules, the policy takes into account the committee’s recommendations and states the approach and direction given by the board. It contains the guiding principles for the selection, implementation and monitoring of activities, and the list of activities and projects the company plans to undertake within the heads of Schedule VII. Since the 2021 revision, the definition of the policy also embraces the formulation of the annual action plan, so the standing policy and the year’s programme are formally one instrument at two levels of generality, the first stating the company’s chosen fields and methods and the second committing the year’s money to named projects.

The disclosure of the policy is part of its legal operation. Section 135(4) requires the board to disclose the contents of the policy in its report and to place it on the company’s website, if any, in the manner prescribed, and the rules add that the composition of the CSR Committee and the projects approved by the board must also be disclosed on the website for public access. A policy that exists only in a minute book is therefore an incomplete compliance, and the website disclosure has a real function in the regime’s design: it is the point at which shareholders, agencies seeking support and the ministry itself can see what a company has committed to before the money moves, rather than reconstructing the commitment from the annual report after it has.

Monitoring closes the loop that the policy opens. The committee’s statutory duty to monitor the policy from time to time is discharged in practice through the monitoring and reporting mechanism that the annual action plan must itself describe, and through the board’s obligation to satisfy itself that disbursed funds were used as approved. For listed companies and the larger unlisted classes, CSR compliance also enters the perimeter of the secretarial audit and the statutory audit’s review of the board’s report, because the annual report on CSR is an annexure to that report and errors in it are errors in a document the auditors read. The consequence is that the CSR file is examined by more eyes, under more mandates, than the section’s brevity suggests, and inconsistencies between the policy on the website, the annexure in the board’s report and the figures in Form CSR-2 are the kind of discrepancy that draws a registrar’s query without any complaint ever being made.

Schedule VII: what the money may be spent on

The spending must fall within Schedule VII of the Act, which lists twelve heads of activity. Read as a whole rather than as a catalogue, the schedule groups into a small number of families. The first family is subsistence and health: eradicating hunger, poverty and malnutrition, promoting health care including preventive health care and sanitation, contributions to the Swachh Bharat Kosh, and making available safe drinking water. The second is capability: promoting education, including special education and employment-enhancing vocational skills, especially among children, women, the elderly and the differently abled, and livelihood enhancement projects. The third is equality: promoting gender equality, empowering women, setting up homes and hostels for women and orphans, setting up old age homes and day care centres, and measures for reducing inequalities faced by socially and economically backward groups. The fourth is environment and inheritance: ensuring environmental sustainability, ecological balance, protection of flora and fauna, animal welfare, agroforestry, conservation of natural resources and maintaining quality of soil, air and water, including contributions to the Clean Ganga Fund; and protection of national heritage, art and culture, including restoration of buildings and sites of historical importance and public libraries, and the promotion and development of traditional arts and handicrafts. The fifth is the state’s own funds and institutions: contributions to the Prime Minister’s National Relief Fund, the Prime Minister’s Citizen Assistance and Relief in Emergency Situations Fund, and other funds set up by the Central Government for socio-economic development and relief and for the welfare of the Scheduled Castes, the Scheduled Tribes, other backward classes, minorities and women; contributions to incubators and to research and development projects in science, technology, engineering and medicine funded by governments, public sector undertakings and specified agencies; and contributions to specified public-funded universities and national laboratories engaged in research aimed at the sustainable development goals. The sixth is place and emergency: rural development projects, slum area development, disaster management including relief, rehabilitation and reconstruction, measures for the benefit of armed forces veterans, war widows and their dependents and the families of personnel of the central armed police forces, and training to promote rural sports, nationally recognised sports, Paralympic sports and Olympic sports.

The schedule has not stood still, and its amendments are a small history of national emergencies registering themselves in company law. The Swachh Bharat Kosh and the Clean Ganga Fund were recognised within the sanitation and environment heads by ministerial clarification as those national missions were launched. The Prime Minister’s Citizen Assistance and Relief in Emergency Situations Fund, created during the COVID-19 emergency, was expressly brought within the funds head of the schedule in 2020, and contributions to it became one of the largest single destinations for CSR money in the pandemic years. The disaster management head, under which relief, rehabilitation and reconstruction sit, acquired its prominence through the same pressures. A compliance team reading the schedule should therefore read it with its notifications, because the operative list at any moment is the enacted twelve heads as glossed and extended by the ministry’s instruments, and a project framed against a superseded version of the list can fail at the reporting stage even though the spending was real.

The Ministry of Corporate Affairs stated early in the regime’s life, in a clarificatory circular issued within months of commencement, that the entries in Schedule VII must be interpreted liberally so as to capture the subject matter they describe, and that position has governed administrative practice since. Liberal interpretation, however, operates on the heads of activity. It does not dissolve the exclusions, which are set out not in the schedule but in the definition of CSR in the rules, and which do more practical work than the schedule itself. An activity undertaken in the normal course of the company’s business is not CSR, subject to a spent exception that allowed companies engaged in research and development of new vaccines, drugs and medical devices to treat COVID-19 related research in that line as CSR for the three financial years beginning 2020-21. A contribution of any amount, directly or indirectly, to a political party under Section 182 of the Act is not CSR. Activities that benefit only the employees of the company and their families are not CSR. Activities supported on a sponsorship basis for deriving marketing benefits for the company’s products or services are not CSR. Activities carried out in fulfilment of any other statutory obligation are not CSR, so a company cannot discharge a duty imposed by an environmental clearance or a labour statute and count the same rupee again under Section 135. Each exclusion answers a specific attempted conversion of ordinary business expenditure into compliance, and together they mark the boundary the statute actually polices.

Does a company have to spend CSR money near its own operations?

No, although the statute leans in that direction. A proviso to Section 135(5) directs the company to give preference to the local area and the areas around it where it operates, when choosing where to spend. Preference is not confinement. Spending outside the local area is lawful, and large national programmes run by covered companies are built on that footing.

The preference does real work in two ways: it is the first question a reader of the annual report asks when a company with plants in one state spends exclusively elsewhere, and it explains much of the geographic pattern in the spending data discussed below, because companies that follow the statutory nudge concentrate their projects around their own facilities.

Can CSR money be spent outside India?

Almost never. The definition of CSR in the rules excludes any activity undertaken by the company outside India, with a single exception for training of Indian sports personnel representing a state or a union territory at national level or India at international level. A covered company with operations, suppliers or charities abroad cannot route its two per cent through them.

The spending must be on projects in India, which is why the transfer destinations for unspent money are Indian funds and an Indian bank account, and why multinational groups run their Indian CSR through the Indian company’s own policy rather than through a global foundation’s programme.

Read together, the exclusions disclose the theory of the provision more honestly than the schedule does. Schedule VII describes the destinations the state wishes to subsidise with compelled money. The exclusions describe the state’s distrust of the spender. A company is not trusted to count what it would have spent anyway, because then the mandate adds nothing; it is not trusted to count what benefits its own workforce, because then the mandate is a wage supplement; it is not trusted to count what markets its products, because then the mandate is an advertising subsidy; it is not trusted to count what another law already compels, because then the mandate double-counts a single act of obedience; and it is not trusted to spend abroad, because the mandate is an instrument of domestic policy. The additionality principle that runs through all five, that compelled money must buy something the world would not otherwise have had, is nowhere stated in the Act, and it is the most useful single idea a board can carry into a doubtful case, because it predicts the administrative answer in almost every instance the circulars have addressed.

Choosing the head for a project is an interpretive act with reporting consequences, and boards underestimate it. A single field programme can plausibly sit under several heads: a women’s livelihood collective reads as gender equality and empowerment, as vocational skills within education, and, if it works through a self-help group that holds a capital asset, as rural development. The head chosen at approval governs how the spending is classified in the annual report, how it aggregates into the sectoral statistics on which the national debate runs, and how an impact assessment later frames its questions. The ministry’s direction that the schedule be interpreted liberally protects the choice from pedantic challenge, but it does not make the choice free of consequence, because a company that classifies opportunistically, writing every project against the head that flatters its preferred narrative, produces an annual report that cannot be compared with its own earlier years. The disciplined practice is to fix the head at the annual action plan stage, record the reasoning where a project spans heads, and classify the same kind of project the same way year after year, so that the record accumulates meaning instead of noise.

Two boundary disputes deserve mention because they recur every year. The first concerns relief funds run by states. Item (viii) of Schedule VII speaks of the Prime Minister’s funds and other funds set up by the Central Government, and the ministry has had to clarify more than once that a contribution to a Chief Minister’s relief fund does not, without more, answer the description, so companies that gave to state relief funds in an emergency have found the spending disallowed for CSR purposes. The second concerns the company’s own foundation. Spending through a foundation the company has established is a recognised route, examined below, but the foundation’s ordinary running costs, its brand promotion and its activities for the company’s employees do not become CSR by passing through it.

Spending routes: agencies, assets and overheads

Section 135 does not require a company to deliver projects with its own staff. The rules permit implementation through several channels: the company itself; a Section 8 company, a registered public trust or a registered society established by the company, singly or with others; an entity established by the Central Government or a state government; an entity established under an Act of Parliament or a state legislature; or an independent Section 8 company, registered public trust or registered society with an established track record of at least three years in similar activities. Implementing agencies in the covered categories must be registered under Sections 12A and 80G of the Income Tax Act, 1961, and since 1 April 2021 every agency through which a company routes CSR money must register with the Ministry of Corporate Affairs in Form CSR-1 and obtain a CSR registration number before funds move. The registration requirement was one of the principal recommendations of the High Level Committee on CSR of 2018, and it changed the practical relationship between companies and agencies: an unregistered agency, however reputable, is not a lawful conduit.

The board’s responsibility does not travel with the money. The rules require the board to satisfy itself that funds disbursed have been utilised for the purposes and in the manner approved by it, and the chief financial officer of the company, or the person responsible for financial management, must certify to that effect. That certification duty, introduced in the 2021 rules, is the provision that moved CSR assurance from the programme team to the finance function, and it is why utilisation certificates, milestone reports and site evidence sit in the CSR file alongside the board papers.

Money spent on a capital asset raises a distinct question, because the company has spent its two per cent but something durable remains. The rules answer it by taking the asset out of the company’s hands: a capital asset created or acquired with CSR funds must be held by a Section 8 company, a registered public trust or a registered society with charitable objects and a CSR registration number, or by the beneficiaries of the project in the form of self-help groups, collectives or entities, or by a public authority. A school building funded under CSR cannot sit on the sponsoring company’s balance sheet as its property. The rule prevents the obligation from being discharged by what is, in substance, investment in the company’s own estate.

Collaboration between companies is permitted and common. The rules allow companies to undertake activities jointly or through pooled arrangements, and group companies routinely run a single large project across several covered subsidiaries, provided each participating company is in a position to report its own share of the spending in its own board report. Pooling solves a real problem, because the two per cent of a mid-sized subsidiary may be too small to fund a durable project alone, while the combined obligations of a group can build and run something of consequence. The discipline is in the paper: the collaboration must be structured so that utilisation, milestones and outcomes can be attributed company by company, since the obligation, the certification and the penalty all attach to the individual company and never to the group.

Two quantitative disciplines govern the spend itself. Administrative overheads, meaning the expenses of general management and administration of the company’s CSR functions, excluding expenses directly incurred in designing, implementing, monitoring and evaluating a particular project, are capped at five per cent of the company’s total CSR expenditure for the financial year. And any surplus arising out of CSR activities is not business profit of the company: it must be ploughed back into the same project, or transferred to the Unspent CSR Account, or to a fund specified in Schedule VII, and it cannot be netted off silently against the next year’s spend except through the set-off mechanism described in the next section.

The money in the accounts

The obligation is a creature of company law, but it is discharged or defaulted inside the financial statements, and the accounting treatment follows from the legal structure with unusual directness. CSR expenditure incurred on projects is an expense of the year, charged in arriving at the profit of the company; it is not an appropriation of profit, because the duty is imposed by statute on the company rather than declared by shareholders out of profit, and the distinction matters for how the charge is presented and analysed. Where the year ends with a shortfall, the legal duty to transfer converts the unspent amount into a liability of the company at the balance sheet date: the sum destined for the Unspent CSR Account or for a Schedule VII fund is recognised as payable, and professional guidance issued for the regime, including the frequently asked questions published by the Accounting Standards Board of the Institute of Chartered Accountants of India, works through the recognition of that liability under the accounting standards on provisions. Money standing in the Unspent CSR Account remains the company’s money in a restricted sense: it is held in a designated account for a designated project, disclosed accordingly, and is not available for the company’s general purposes, while the interest it earns belongs to the project and not to the company’s income. Excess spending that the board has resolved to set off is tracked as a carried-forward entitlement against future obligations rather than recognised as an asset in the ordinary way, and the disclosure of the movement, prescribed amount, amount spent, amount transferred and amount carried, is what allows a reader of the statements to reconcile the legal position with the financial one. The finance team’s year-end checklist therefore runs in a fixed order: compute the obligation, agree the spend to evidence, classify the shortfall, pass the entries that recognise the transfer liability, and only then close the books, because each later document, the annexure, Form CSR-2 and the chief financial officer’s certificate, is built from these entries.

Impact assessment: the regime learns to ask what changed

For the regime’s first seven years, no provision asked whether a completed project had achieved anything. The annual report recorded money in and money out; a project that spent its budget in full and changed nothing stood in the record identically with a project that transformed its village. The Companies (CSR Policy) Amendment Rules, 2021 inserted the first outcome instrument. Every company with an average CSR obligation of Rs 10 crore or more across the three immediately preceding financial years must undertake an impact assessment, through an independent agency, of its CSR projects with outlays of Rs 1 crore or more that have been completed not less than one year before the assessment. The reports must be placed before the board and annexed to the annual report on CSR, and the rules place a ceiling on the share of the year’s CSR expenditure that a company may book towards the exercise, so that assessment cannot itself become a way of spending the obligation on paper.

The design choices inside that obligation are instructive. The thresholds confine the duty to the largest spenders and their largest projects, which is where the money is but also where assessment is easiest to buy. The one-year interval between completion and assessment acknowledges that outcomes in health, education or livelihoods do not declare themselves at the inauguration. The independence requirement addresses the obvious capture problem of a company grading its own programme. And the annexure requirement puts the findings, favourable or not, into the same public document as the spending claims, where a reader can hold the two against each other. What the provision does not do is attach any consequence to a bad finding. A company whose flagship project is assessed as ineffective has committed no default; it has merely published that fact, and the statute leaves the response to the board’s judgment and the shareholders’ attention. Impact assessment is therefore best understood as the regime’s first evidentiary instrument rather than as an enforcement one: it creates the record on which an argument about quality can at last be conducted, without itself conducting the argument.

Its arrival also changed the market underneath the regime. Independent assessment created a professional discipline between the implementing agency and the auditor, with its own methods, baselines and site protocols, and it gave boards a document to demand before renewing a large project for a further cycle. The annual action plan and the impact assessment face each other across the life of a major project: the plan records what the board intended, in terms specific enough to be checked, and the assessment records what an outside agency found. For the aspirant, the pair is the cleanest illustration of the direction of travel of the whole section, from disclosure of intention, through custody of money, towards evidence of effect, with each stage arriving years after the stage before it.

Unspent money: from explanation to escrow

The treatment of a shortfall is where the statute changed most, and it is the part of the regime that most published guidance still states in its superseded form. In the original design, which operated from 2014 until the amendments took effect, a company that failed to spend the prescribed amount had to do one thing: the board’s report had to specify the reasons for not spending. Nothing had to be transferred, nothing had to be escrowed, and no penalty attached to the shortfall itself. The design assumed that disclosure before shareholders and the public would discipline boards more cheaply than compulsion. The evidence of the early years, examined below, was that a substantial body of companies spent less than the prescribed amount and treated the explanation as the compliance.

The Companies (Amendment) Act, 2019 rewrote that position, and the rewritten provisions were brought into force on 22 January 2021. The architecture that results divides every shortfall into two kinds, and the classification is the single most consequential decision in the CSR calendar. If the unspent amount relates to an ongoing project, Section 135(6) requires the company to transfer it, within thirty days from the end of the financial year, to a special account opened in a scheduled bank for that financial year, called the Unspent Corporate Social Responsibility Account. The money in that account must be spent on the project within three financial years from the date of transfer. Anything still unspent at the end of that period must be transferred to a fund specified in Schedule VII within thirty days from the date of completion of the third financial year. If the unspent amount does not relate to an ongoing project, the second limb of Section 135(5) applies instead: the board must transfer the shortfall to a fund specified in Schedule VII within six months of the expiry of the financial year. Under either route, the money leaves the company’s free disposal. The shortfall is no longer explained; it is surrendered.

The category of ongoing project therefore carries money with it, and it is defined with some precision in the rules. An ongoing project is a multi-year project undertaken in fulfilment of the CSR obligation, having a timeline not exceeding three years excluding the financial year in which it commenced, and the category also includes a project that was initially approved for one year but whose duration has been extended by the board beyond that year on reasonable justification. A project that a company wishes to treat as ongoing must have been approved as such, with its timeline and year-wise allocation recorded, before the year closes. A board cannot look at an underspend in April, describe the programme it belonged to as multi-year, and move the money into the account on that basis alone; the classification must exist in the approvals made while the year was running. This is the point at which weak minute-keeping converts directly into a transfer to a government fund.

The account itself has a discipline that repays attention. A separate Unspent Corporate Social Responsibility Account is opened for each financial year, so balances of different years, each running on its own three-year clock, are not merged into a single pool whose deadlines blur. The money in the account is not working capital: it may be applied only to the ongoing project for which it was set aside, and its presence is disclosed in the reporting of each year until it is exhausted or surrendered. Spending out of the account in a later year is reported in that later year’s annual report on CSR as spending against the earlier obligation, which is why a company’s annexures form a chain across years and why a break in the chain, an account balance that appears in one year’s report and vanishes without a corresponding spend or transfer in the next, is the sort of discrepancy that draws regulatory attention. The account is, in substance, a statutory escrow operated by the company itself, and its annual, segregated structure is the mechanism by which a three-year promise is kept visible to auditors and registrars who were not parties to the original approval.

One distinction runs through the whole custody architecture and decides hard cases at the margin: the difference between money disbursed and money utilised. A company discharges its programme through implementing agencies, and money leaves the company’s bank account in tranches against milestones. At the year end, a tranche may have been disbursed to an agency and yet be unutilised in the field, sitting in the agency’s account against work scheduled for the following season. The regime’s instruments look at the two events differently. The spending obligation is concerned with the company’s application of its two per cent to approved projects, and disbursement against an approved project is the act the company controls; the certification duty, by contrast, is concerned with utilisation, and it obliges the board to satisfy itself, through the chief financial officer’s certificate, that what was disbursed was in fact used for the approved purposes in the approved manner. The gap between the two is where agency management earns its keep: milestone schedules set so that tranches track the implementing calendar, utilisation certificates collected before the next tranche moves, and balances with agencies reconciled at the year end so that the board’s report, the certification and the agency’s own accounts tell one story. A board that treats disbursement as the end of its responsibility will find the certification conversation uncomfortable; a board that treats utilisation as the only event that counts will misstate its spending position. The statute requires both, at different points, for different purposes.

The mirror image of the shortfall rules is the set-off for excess spending. A proviso inserted into Section 135(5) by the Companies (Amendment) Act, 2020 permits a company that spends more than its obligation in a financial year to set off the excess against the amount it must spend in the three immediately succeeding financial years, provided the board has passed a resolution to that effect and the excess does not include any surplus arising out of the CSR activities themselves. Set-off rewards a company that front-loads a large project, and it is the only lawful form of carry-forward in the regime: unspent money cannot be carried forward as spending capacity at all, because the statute sends it to the account or to a fund, while overspent money can be recognised, once, through a board resolution taken in the year of the excess.

The transition into the custody regime was harder than the destination. The amendments commenced on 22 January 2021, in the final quarter of the financial year 2020-21, and the ministry’s position was that the transfer duties applied to the unspent amounts of that same financial year. Companies that had run the year on the comply-or-explain understanding, with no ongoing project approvals framed for the purpose and no Unspent CSR Account in existence, had roughly two months to classify their shortfalls, obtain the approvals the classification required and open the accounts. The episode is worth recording in a compliance article because it set the pattern of behaviour the regime assumes: boards approve project classifications and timelines during the year precisely because the year-end classification cannot safely be improvised, and the annual action plan has become the instrument by which the classification is fixed in advance.

The consequence of default: an offence that never operated

At enactment, default in spending carried no penalty under Section 135 at all. The only money consequences in the neighbourhood were the general penalties for a defective board’s report, which attach to the disclosure rather than to the shortfall. The Companies (Amendment) Act, 2019, which received assent on 31 July 2019, changed that abruptly. It attached to a failure to transfer an unspent amount the consequence that the company would be punishable with a fine of not less than Rs 50,000 and up to Rs 25 lakh, and every officer of the company in default would be punishable with imprisonment for a term extending to three years, or a fine of not less than Rs 50,000 and up to Rs 5 lakh, or both. For a provision that had begun life as an obligation to explain, this was a startling escalation, and industry reaction was immediate and hostile.

The reversal was already in preparation while the ink dried. The High Level Committee on CSR, constituted in October 2018 under the chairmanship of Injeti Srinivas, the Secretary for Corporate Affairs, presented its report to the Finance Minister, Nirmala Sitharaman, on 13 August 2019. The committee recommended that a violation of the CSR provisions be treated as a civil wrong under a penalty regime rather than as a criminal offence, with a penalty of two to three times the amount in default and no imprisonment, and in August 2019 the Finance Minister announced publicly that CSR violations would be treated as a civil liability and the provisions reviewed accordingly. The Companies (Amendment) Act, 2020 then substituted Section 135(7). Default in spending or in transferring exposes the company to a penalty of twice the amount required to be transferred to a Schedule VII fund or to the Unspent CSR Account, or Rs 1 crore, whichever is less, and every officer in default to a penalty of one-tenth of that amount, or Rs 2 lakh, whichever is less.

The sequence has one consequence that matters for anyone reading older commentary. The 2019 and 2020 provisions were commenced together, on 22 January 2021, by which date the civil penalty had already replaced the criminal offence in the text. The imprisonment provision was therefore law on paper for a period in which it was never brought into force, and no prosecution under it was ever possible in practice; when a criminal complaint founded on the short-lived offence reached the Gujarat High Court in 2023, the court quashed it, applying the decriminalising amendment to the company’s benefit. Any source that describes CSR default as punishable with imprisonment is describing a provision that was enacted, announced against, replaced and only then commenced in its replacement form. The accurate one-line statement is that CSR non-compliance is a civil default visited with a money penalty, adjudicated by the registrar through the penalty machinery of the Act rather than prosecuted before a criminal court.

The penalty’s structure repays a moment’s attention, because it is a penalty on the failure to move money, calibrated to the amount that should have moved. For a large default the cap of Rs 1 crore will often be far below twice the untransferred amount, which means the marginal cost of default stops rising once the shortfall passes Rs 50 lakh, while the underlying obligation to transfer survives the penalty and remains enforceable. A worked example fixes the arithmetic. A company that should have transferred Rs 80 lakh of unspent amount and did not faces a penalty measured at twice that sum, Rs 1.6 crore, capped at Rs 1 crore; each officer in default faces one-tenth of the amount that should have been transferred, Rs 8 lakh, capped at Rs 2 lakh. A smaller default of Rs 10 lakh produces a company penalty of Rs 20 lakh and an officer penalty of Rs 1 lakh, below both caps, which is why the regime bites proportionately hardest in the middle range of companies, large enough to have real shortfalls and small enough that the caps do not rescue them.

The penalty is imposed through adjudication. Section 454 of the Act gives the Registrar of Companies, as adjudicating officer, the power to impose penalties for defaults of this character after a hearing, with an appeal lying to the Regional Director, and Registrars have issued adjudication orders against defaulting companies in the years since the regime commenced, treating CSR defaults as routine adjudication work alongside other filing penalties. Alongside the transfer penalty sit the older consequences for the paperwork: a board’s report that omits or falsifies the CSR disclosures attracts the penalty under Section 134(8), which stands at Rs 3 lakh for the company and Rs 50,000 for every officer in default, and it attaches to the disclosure failure whether or not the spending itself was made. Default is therefore neither theoretical nor catastrophic: it is a priced, recorded and publishable failure, sitting on the company’s record at the ministry, and for most boards the reputational entry will cost more than the penalty.

Where companies go wrong

The defaults that reach adjudication are rarely exotic. They cluster in a small number of recurring errors, and a compliance team that clears these has cleared most of the regime’s practical risk. The first is the thresholds error: reading the three tests in Section 135(1) as cumulative, and concluding that a company with a net profit above Rs 5 crore but a small turnover is outside the section. It is not, and the years of missed compliance that follow from the misreading are not cured by the company’s good faith when the registrar computes them. The second is the base error: computing two per cent on the profit after tax in the financial statements, without the Section 198 reconstruction and the rule-based exclusions, which overstates the obligation for a company with capital profits and understates it for a company with particular charges, and in the understating case manufactures a shortfall the company does not know it has.

The third cluster concerns the character of the spending. Welfare expenditure on a company’s own employees, a scholarship scheme confined to employees’ children, a hospital wing reserved for the workforce, all fail the exclusion of activities benefiting only employees, however worthy they are as human resources policy. Sponsorship of an event, a team or a property from which the company draws marketing benefit fails the sponsorship exclusion even where the event has a social theme and a Schedule VII head can be named for it. A contribution routed to a political party fails outright. And expenditure the company was already obliged to incur under another law, resurfacing a factory effluent plant required by its environmental clearance, fails the statutory-obligations exclusion. Each of these has the same anatomy: the spending is real, the money is gone, and the two per cent remains undischarged because the activity never counted.

The fourth cluster is administrative, and it is the cluster the 2021 rules created. Money sent to an implementing agency that had not obtained its CSR registration number in Form CSR-1 is money sent outside the permitted channels. A capital asset built with CSR funds and retained on the company’s own books breaches the holding rules even though the asset serves the public. Surplus earned on a project and taken into income in the profit and loss account misstates both the accounts and the CSR position. Administrative overheads booked without watching the five per cent cap quietly consume obligation money on programme salaries. None of these involves a decision to default; all of them are discovered, if they are discovered, by the finance team’s own reconciliation or by an auditor’s query, and the correction window is always the year in which the spending occurred.

The fifth and most expensive cluster is the year-end failure described earlier in this article: a shortfall discovered after 31 March, an ongoing project classification that was never approved while the year ran, a transfer made late or to the wrong destination, a Schedule VII fund chosen without checking that it is a fund the schedule actually specifies. The penalty attaches to the transfer default, not to the spending decision, so a company that spent ninety-five per cent of its obligation on impeccable projects and then missed the 30 April transfer of the balance is in default in exactly the same provision as a company that spent nothing. The lesson the regime teaches, and teaches expensively, is that CSR compliance is a calendar discipline laid over a spending programme, and that the calendar is the part that carries the penalty.

A sixth cluster lives in the disclosure layer and is the easiest of all to avoid. Companies omit the CSR policy or the committee’s composition from the website, or leave a superseded policy displayed after the board has approved a revised one. The figures in Form CSR-2 are filed from a working sheet that was not reconciled with the annexure in the board’s report, so the registry data and the signed report disagree. The annual action plan is altered in practice, projects substituted and budgets moved by management decision, without the committee’s recommendation and the board’s approval that the rules require for an alteration, so the spending is real but unauthorised in form. None of these defaults involves money lost or projects failed; each is a failure of the record, and each is discoverable from the public documents alone, which is why disclosure discipline belongs on the same checklist as the transfer deadlines rather than in a separate, lesser category of housekeeping.

Who answers for a default: the officer in default

The penalty provisions speak of the company and of every officer in default, and the second limb deserves the same precision as the first, because it is the limb that concentrates minds in boardrooms. Officer in default is a defined expression in Section 2(60) of the Act, and it reaches beyond the signatories of the CSR documents: it covers whole-time directors and key managerial personnel, directors who were aware of the default through board proceedings in which they participated without objection, and persons charged with responsibility for the relevant compliance. In the CSR setting the practical exposure settles on the directors who approved the year’s accounts and report without the transfer being made, the chief financial officer whose certification and payment systems are the mechanism by which a transfer would have happened, and the company secretary who owns the compliance calendar. The exposure is a money penalty, capped at Rs 2 lakh or one-tenth of the untransferred amount, and it is personal: the company cannot lawfully indemnify an officer against a penalty imposed on the officer, and the adjudication order names individuals. This personal, if modest, exposure explains a feature of CSR practice that puzzles outside observers, which is the seniority of the attention the subject receives in the final quarter. The transfer deadlines sit in the same weeks as the closing of accounts, and the officers who would be named in an adjudication order are the officers in the room when the accounts are closed, which aligns the incentive to classify and transfer correctly with the people who have the authority to do it.

There is a corresponding point about what the penalty does not reach. It attaches to the defaults specified in Section 135(5) and (6), the spending and transfer duties. It does not attach to the quality of projects, to the composition of the portfolio, or to a failure to achieve outcomes, none of which is a default known to the section. An officer in default is therefore an officer who failed at the calendar and the custody, not an officer whose company’s projects disappointed, and the distinction preserves the boundary the whole regime observes between enforcing movement of money and judging its use.

A year in the life of the obligation: a worked passage

The provisions interlock most clearly when they are followed through a single company’s year, so take a components manufacturer with a factory in an industrial belt and offices in two cities. Its audited figures for the preceding financial year show a net worth of Rs 340 crore, a turnover of Rs 890 crore and a net profit, reconstructed under Section 198, of Rs 6.2 crore. The first two gates are missed; the third is crossed, and the company is inside Section 135 for the coming year on profit alone, which is the entry route the brief summaries most often miss. Its Section 198 profits for the three preceding years, after excluding a capital profit on the sale of a surplus plot in the earliest of them and a dividend from a compliant subsidiary in the second, are Rs 6.2 crore, Rs 4.8 crore and Rs 7 crore, an average of Rs 6 crore, so the year’s obligation is Rs 12 lakh. That figure is below Rs 50 lakh, so no CSR Committee is required; the board itself discharges the committee’s functions, adopts the policy, and approves an annual action plan committing the money to a sanitation programme run with the district administration near the factory and to scholarships administered by a registered trust with a CSR registration number.

Follow the same company forward a few years, grown by acquisition to a turnover of Rs 1,400 crore and average profits of Rs 210 crore, and the full machinery engages. The obligation is Rs 4.2 crore. A committee of three directors, one of them independent, frames the annual action plan: a three-year watershed project with an implementing agency, approved as an ongoing project with year-wise allocations; a health programme run by the company’s own foundation; and a contribution held ready for a Schedule VII fund. The board approves, the policy and the committee’s composition go on the website, and the chief financial officer’s office builds the utilisation file as disbursements are made against milestones. At the year end the company has spent Rs 3.5 crore. Of the Rs 70 lakh shortfall, Rs 55 lakh is the undisbursed tranche of the watershed project, classified as ongoing in the approvals made during the year, and Rs 15 lakh is an unspent balance on a programme that ended. By 30 April the Rs 55 lakh moves into the Unspent Corporate Social Responsibility Account opened for that year, where it must be spent on the watershed project within three financial years. By 30 September the Rs 15 lakh moves to a fund specified in Schedule VII. The annexure to the board’s report tells exactly this story with dates, Form CSR-2 repeats it to the registrar, and the watershed project’s later spending out of the account is reported year by year until the account is exhausted. Every element of that passage, the gate crossed on profit, the reconstructed base, the small-obligation relaxation, the committee at scale, the classification, the two transfer dates, is a provision examined in this article, and the passage shows why the order of operations matters: a single missed approval in the middle of the year would have sent the watershed balance to a government fund in September instead of to the project across three years.

The CSR compliance clock

The regime runs on the financial year, and its deadlines interlock tightly enough that a failure at an early stage manufactures a default at a later one. The clock below states the sequence for a company with a 31 March year end, which is the standard Indian financial year, and gives the consequence that attaches to missing each step. Everything in it is drawn from the provisions worked through above; the table is a finding aid for the prose, not a substitute for it. The general position set out in this article is information about how Section 135 works, not advice on any individual company’s obligation, and because the transfer deadlines admit no condonation, a company working to the last dates in the table should confirm its own position on its own figures before the board acts.

Stage	When	What must happen	Consequence of missing the step
Close of the preceding financial year	31 March of the year before the spending year	The audited net worth, turnover and net profit figures fix whether Section 135 applies and supply the profit base	A wrong applicability call infects every later step; the duty, once triggered, cannot be undone by the spending year’s results
Computation of the obligation	Before the spending year opens	Average the Section 198 net profits of the three immediately preceding financial years and take two per cent	An understated obligation produces an unplanned shortfall at year end, classified and transferred on the wrong figure
Committee recommendation and annual action plan	Early in the spending year	The CSR Committee formulates the policy recommendations and the annual action plan; the board approves both	Spending outside an approved plan lacks the approvals the rules require and weakens every later certification
Board approval and disclosure	With the plan	The board approves the CSR policy, discloses it in the board’s report and places it on the company website	Defective disclosure attracts the penalties for a defective board’s report, separate from any spending default
Spending through the year	Across the financial year	Projects are executed directly or through registered implementing agencies; utilisation is tracked against milestones	Unmonitored agency spending cannot support the chief financial officer’s utilisation certificate
Year end classification	At 31 March	Every rupee of shortfall is classified as relating to an ongoing project approved as such, or as other unspent amount	A shortfall without a prior ongoing-project approval cannot enter the Unspent CSR Account and must go to a Schedule VII fund
Transfer for ongoing projects	Within thirty days of year end, by 30 April	Transfer the ongoing-project shortfall to the Unspent Corporate Social Responsibility Account in a scheduled bank	Section 135(7) penalty: twice the amount that should have been transferred or Rs 1 crore, whichever is less, on the company, and one-tenth or Rs 2 lakh on each officer in default
Transfer of other unspent amounts	Within six months of year end, by 30 September	Transfer any shortfall not relating to an ongoing project to a fund specified in Schedule VII	The same Section 135(7) penalty, and the money is lost to the company’s programme in any event
Completion of the third financial year	Three financial years from the transfer into the account	Spend the account balance on the ongoing project; transfer any residue to a Schedule VII fund within thirty days of the third year’s completion	The residue transfer carries the same penalty if missed, closing the last route by which a shortfall could remain with the company
Annual reporting	With the board’s report and the annual filings	The annual report on CSR is annexed to the board’s report and Form CSR-2 is filed with the Registrar	Reporting defaults are separately penalised and place the company’s position on the public record

Reporting: the board’s report, Form CSR-2 and the public record

The reporting layer is where the regime becomes visible to everyone outside the company, and it has two components that serve different readers. The first is the annual report on CSR, annexed to the board’s report in the format the rules prescribe. The format is a structured account of the year: the committee’s composition, the computation of the average net profit and the prescribed amount, the amount spent on projects and on administrative overheads, the amounts transferred to the Unspent CSR Account and to Schedule VII funds with the dates of transfer, the details of ongoing projects carried across years, any set-off claimed, and the reasons where the company failed to spend the prescribed amount, closing with the responsibility statement of the committee and the board. Read across several years, the annexure is the biography of a company’s obligation, and it is the document from which journalists, researchers and rival boards draw their accounts of who spent what.

The second component is Form CSR-2, the report on CSR filed with the Registrar of Companies as part of the annual filing cycle, introduced for the financial year 2020-21 and filed as an addendum to the form in which the financial statements themselves are lodged. Form CSR-2 converts the annexure’s narrative into registry data, and it is that data which feeds the ministry’s compilations and the National CSR Portal, the public database through which company-wise and year-wise spending can be searched. Every aggregate figure quoted in the debate about this section, including the figures used in this article, descends from these filings, which is why the accuracy of the filing is a matter beyond the company’s own compliance: an error in Form CSR-2 is an error in the national evidence base, repeated by every analyst who builds on it.

The two components discipline each other. The annexure is signed off through the board’s report process and read by the auditors in that context; the form is a structured filing whose internal numbers must reconcile with the annexure and with the financial statements. A company whose three versions of the year, the audited accounts, the annexure and the form, do not agree has created a discrepancy that any reader of the public record can find, and registrars’ queries in the regime’s enforcement phase have repeatedly begun from exactly such mismatches rather than from any inspection of projects. This is the practical sense in which the regime polices itself: the spending happens in distant districts, but the record of it is centralised, standardised and public, and most defaults are visible from a desk.

The tax position: no deduction for the spend itself

The question finance teams ask second, immediately after the computation, is whether the two per cent can at least be deducted. The answer is that Parliament closed that door deliberately and early. An Explanation inserted into Section 37(1) of the Income Tax Act, 1961 by the Finance (No. 2) Act, 2014, effective from the assessment year 2015-16, provides that any expenditure incurred by an assessee on activities relating to corporate social responsibility referred to in Section 135 of the Companies Act shall not be deemed to be expenditure incurred for the purposes of business or profession. The reasoning recorded at the time was fiscal symmetry: the obligation exists to make companies share the burden of providing social services, and allowing the same expenditure as a business deduction would have the exchequer subsidise roughly a third of it through forgone tax, converting a two per cent obligation into a materially smaller net cost distributed unevenly by tax position. The detailed working of the disallowance belongs to the complete guide to the Income Tax Act, 1961, which owns the tax treatment; what a CSR reader needs is the boundary it draws.

The Central Board of Direct Taxes added a refinement in a circular issued in January 2015 that finance teams still rely on. The disallowance operates through Section 37(1), the general deduction provision; expenditure that is of a kind described in the specific provisions of Sections 30 to 36 of the Income Tax Act, rent, repairs, depreciation and the like, is not denied merely because the spending also answers a CSR description, provided the conditions of the specific provision are independently satisfied. The same circular confirmed that the Explanation’s bar is directed at the claim of CSR expenditure as a business deduction in general terms. The practical line that results is between the CSR outlay as such, which is never deductible under the general head, and the ordinary incidents of carrying on a project through assets and premises whose costs the Act treats under their own heads.

The boundary is narrower than the headline. The Explanation kills the deduction of CSR expenditure as business expenditure under the general provision; it does not repeal the specific deduction provisions elsewhere in the Income Tax Act. A contribution that independently satisfies the conditions of a specific provision, most often a donation to a fund or institution approved for the purposes of Section 80G, may still qualify for deduction under that provision on its own terms, even though the same contribution counts towards the two per cent under the Companies Act. The converse discipline also applies: expenditure claimed as CSR for company law purposes cannot be dressed as ordinary business expenditure of another character to escape the Explanation. The High Level Committee on CSR of 2018 recommended that CSR expenditure be made tax deductible, as part of a package intended to reward quality of spending rather than mere quantum; the recommendation was not accepted, and the disallowance stands, which is itself evidence of how the state understands the levy, as a charge on profits that is not to be softened through the tax system.

The line between CSR and improper payments

One boundary of the regime is policed from the other side by a different statute. CSR spending moves money from companies to projects, agencies and public funds, often in places where the same company seeks licences, land, clearances and contracts, and the overlap creates an obvious risk: a social project can be a genuine discharge of the two per cent and, at the same time, a benefit conferred on a decision-maker’s constituency with an expectation of return. The exclusion of political contributions from CSR removes the crudest version of that risk. It does not remove the subtler one, and this is where the CSR file meets the Prevention of Corruption Act, 1988, under which giving an undue advantage to a public servant is an offence and a commercial organisation can be liable where a person associated with it bribes to obtain business or a business advantage. A contribution to a Schedule VII fund specified by the Central Government is a statutory transfer and raises no corruption question. A discretionary project chosen because it pleases an official whose file is pending is a different matter, and the fact that the spending counts towards the two per cent is no answer to it. Boards that site projects through the annual action plan, on the committee’s recorded recommendation and against published criteria, are building the evidence that separates the two cases; that is a further reason the paperwork discipline of the earlier sections matters beyond mere filing.

The CSR record outside the Act: diligence, ratings and the BRSR

A company’s CSR file is read by more audiences than the registrar, and the collateral readings have become part of the compliance incentive. In a transaction, the target’s CSR record is a standard diligence item: a buyer examines whether the obligation was correctly triggered, whether the computations reconcile, whether transfers were made on time and whether any adjudication or penalty proceedings are pending, because an unremedied transfer default is a liability that survives the change of ownership and a misclassified shortfall distorts the profit record on which the price was built. Lenders and investors read the same record as a governance signal of a particular kind: not whether the company is generous, but whether a board-managed, calendar-driven obligation imposed by statute was discharged precisely, which is a proxy for the reliability of the compliance function generally.

For listed companies, the CSR record also feeds a second reporting regime. The Securities and Exchange Board of India’s Business Responsibility and Sustainability Report, required of the top one thousand listed entities by market capitalisation from the financial year 2022-23, draws on the same underlying data about social spending and community programmes, presented in the language of environmental, social and governance disclosure for investors rather than in the language of statutory compliance for the ministry. The two regimes ask different questions of the same money: the Companies Act asks whether the obligation was discharged, the securities framework asks what the spending says about the company’s management of its social footprint. A compliance team that keeps one clean record serves both; a team that treats the CSR-2 filing, the board report annexure and the sustainability disclosures as separate exercises discovers the discrepancies between them in public. This widening circle of readers is, in the end, the strongest enforcement mechanism the section possesses, stronger than the capped penalty: the registrar can fine a default once, but the record of it is republished to every future reader of the company’s filings.

The comparative check: how unusual the Indian design is

Set against other jurisdictions, the Indian provision is an outlier in kind rather than in degree, and the comparison clarifies which of its features are choices. The dominant model internationally is disclosure: the United Kingdom requires large companies to report on social and environmental matters in their strategic reports and obliges directors to have regard to community and environmental interests in the general duty to promote the company’s success, but it fixes no spending figure, and the European Union’s sustainability reporting architecture likewise compels standardised disclosure of impacts and policies without directing any quantum of expenditure. The disclosure model trusts investors and reputation to price what is revealed. Section 135 does not: it fixes the quantum and the permitted purposes, and trusts disclosure only to police the boundary.

The nearest relative is Mauritius, which since 2009 has required companies to contribute two per cent of their chargeable income to corporate social responsibility, collected substantially through the revenue system and channelled in significant part through a national foundation. The Mauritian design is more frankly fiscal than the Indian one: the levy is collected as revenue and allocated centrally, with a portion available for the company’s own approved projects. India considered the same fork in the road and chose differently, leaving the project choice with boards and using government funds only as the destination for default. South Africa’s broad-based black economic empowerment scorecard offers a third model, in which socio-economic development contributions earn points towards a rating that affects access to public procurement, so the inducement is commercial advantage rather than legal compulsion. Each model answers the same question, how to move private money towards public purposes, with a different theory of the firm: as taxpayer, as reporter, or as bidder. India’s answer, the firm as a statutory spender under board control, is the most ambitious and the most administratively demanding, which is why its history is a history of machinery being added after the fact: registration for agencies, certification by finance heads, custody accounts for residue, assessment for outcomes. The machinery is the price of keeping the choice of projects in the boardroom.

One further comparator marks the outer boundary of the family. Indonesia’s company law of 2007 obliges companies whose business is in, or related to, natural resources to carry out social and environmental responsibility, budgeted and accounted for as a cost of the company, with sanctions left to the sectoral laws. The Indonesian duty is sector-bound and unquantified: it names the industries whose footprint justifies the charge and leaves the amount to judgment. India made the opposite choice on both axes, applying the duty across every sector above a size line and fixing the amount by formula. Between the two sits the whole design space of mandatory CSR, and the Indian experiment is the one other legislatures study, because it is the only version that generates, year after year, a complete public dataset of what compulsion actually collects and where the money goes.

What the record shows, and what it cannot show

The state has tried to steer the flow without amending the section. The Aspirational Districts Programme, run by NITI Aayog across 112 districts chosen for weak development indicators, was opened to CSR participation, and the ministry has circulated the aspirational districts to companies as priority geographies, an attempt to use moral suasion and visibility to correct the concentration that the local area preference produces. Companies are also recognised through national CSR awards instituted by the ministry to honour exemplary spending, an instrument of the disclosure philosophy surviving inside a mandatory regime: the awards cannot compel a single rupee, but they publish a hierarchy of admired practice. These steering instruments matter to the analysis because they show the state reaching for tools beyond the statute when the statute’s own incentives point the wrong way, and because their limited effect on the aggregate distribution is part of the evidence about what mandates can and cannot do.

The developmental defence of Section 135 begins with quantum, and on quantum the record is genuinely large. Figures that the Ministry of Corporate Affairs placed before the Rajya Sabha, drawn from company filings on the MCA21 registry with data up to 31 March 2024, record reported CSR spending of Rs 10,065.93 crore in the first year, 2014-15, rising year by year to Rs 29,987.92 crore in 2022-23, and a total across the nine financial years from 2014-15 to 2022-23 of Rs 1,84,222.87 crore. Whatever view is taken of its effectiveness, a mechanism that moves sums of that order from corporate profits to Schedule VII purposes, through board-approved plans and audited disclosures, is not nothing, and the counterfactual matters: voluntary corporate giving in India before 2014 was real but unstructured, concentrated in a small number of business houses, and nowhere near this aggregate.

The trajectory inside those totals also records what the comply-or-explain design was worth. In the first years of the regime, the annual reports of covered companies carried explanations for shortfalls as a routine item, and the gap between the amount prescribed and the amount spent was a standing feature of the ministry’s own summaries of the programme; spending rose towards the prescribed figure as the regime matured, as boards internalised the obligation and as the data made laggards visible, but the decisive closure of the gap came only when the explanation route was abolished and the residue had to be surrendered. The pandemic years added a further pattern: with the Prime Minister’s Citizen Assistance and Relief in Emergency Situations Fund inside Schedule VII and disaster management an available head, a large share of corporate India directed its two per cent towards central emergency funds, which was lawful, swift and, for companies whose own projects had stopped, rational. It also demonstrated the escape-hatch character of fund contributions at scale, because a transfer to a central fund discharges the section completely while requiring no project, no agency and no local presence from the company at all.

The distribution of the money is where the defence weakens. An analysis by NITI Aayog of ministry data for the first five financial years, from 2014-15 to 2018-19, found spending of Rs 71,190 crore concentrated heavily by subject, with education taking about thirty per cent, health care about seventeen per cent, rural development about eleven per cent and environment about seven per cent, and concentrated heavily by place, with about forty-three per cent of the money received by seven states and Maharashtra alone receiving Rs 11,361 crore. Later analyses tell the same story on a larger canvas: an examination of the decade’s filings published in the Economic and Political Weekly by Atul Ghorpade and Sabuj Kumar Mandal of the Indian Institute of Technology Madras found Maharashtra accounting for about a quarter of state-level CSR spending between 2014 and 2024, with the five most industrialised states together taking over half. The local area preference in Section 135(5) predicts exactly this pattern, because money that follows factories and head offices follows the map of Indian industry, which is not the map of Indian need. Aspirational districts with the weakest services are, by and large, districts without large plants.

Against the developmental defence stands the corporate objection, which the strongest reading of the record supports more than polite commentary usually allows: as amended, the provision operates in substance as a hypothecated tax on the profits of larger companies, collected at two per cent, with the unusual feature that the taxpayer may choose the projects on which part of it is spent and must surrender the unspent residue to government funds. The objection that mandatory CSR crowds out genuine philanthropy is related but distinct: it holds that giving which a company would have done freely, through its own foundations and on its own judgment, is captured inside a compliance frame, reported in a statutory annexure and audited like a filing, with the predictable result that boards optimise for defensibility rather than for effect. There is evidence for the mechanism in the pattern of spending itself: the pull towards safe, countable, photographable projects, schools built and toilets enumerated, over long-horizon work whose outcomes are harder to certify to a registrar.

The statute’s own machinery for answering the effectiveness question is thin, and it is here that the 2021 reforms cut both ways. Mandatory impact assessment, confined to larger obligations and larger projects, is the first provision in the regime that asks what the money achieved rather than where it went, and the annual action plan at least forces intention to be recorded before the fact. But the obligation the statute enforces with penalties is the obligation to spend or transfer; no consequence attaches to a project that spent in full and achieved nothing. The recurring reform proposal, that reporting should shift from expenditure to outcomes, runs into the honest difficulty that outcomes in health or education are produced by many hands over many years and resist attribution to a single company’s outlay. The reader preparing for examinations will find these tensions, thresholds and dates set out in study form in the corporate law exam study guide; the practitioner should note that the argument about them is conducted in committee reports and consultation papers, and that the text of Section 135 has so far absorbed criticism mainly by tightening custody of the money rather than by measuring its results.

The reform agenda that surrounds the section shows what its operators think is missing, because most of it was written by the state’s own committee. The High Level Committee of 2018, beyond the recommendations that were enacted, proposed a set of changes that were not taken up, and the rejected list is as informative as the accepted one. Tax deductibility was recommended and refused, preserving the fiscal character of the levy. A carry-forward of unspent balances for three to five years was recommended; what arrived instead was the narrower custody account for ongoing projects and the set-off for excess spending, both of which keep the money inside the regime’s timetable rather than the company’s. Alignment of Schedule VII with the sustainable development goals was recommended and has been pursued only as a mapping exercise. A CSR exchange, a platform matching companies, agencies and beneficiaries, was recommended and not built, leaving the matching of money to projects to private networks. Registration of implementing agencies and mandatory impact assessment, the two recommendations with the clearest administrative form, were the two enacted in full. The pattern is consistent: proposals that strengthened custody, identity and evidence were adopted; proposals that softened the fiscal edge or built new public infrastructure around the section were not.

The standing proposal that has not been settled by either committee is the shift from expenditure reporting to outcome reporting: that companies should report, against stated baselines, what their completed projects changed, and that the annual report should carry outcomes with the same prominence as outlays. The proposal’s strength is that it aims the regime at its declared purpose. Its difficulty is measurement discipline across thousands of companies and projects of wildly different kinds, and the risk that outcome claims, unlike transfer receipts, cannot be verified from a desk, so the regime would trade a hard record of money for a soft record of effects. The likelier path, visible in the impact assessment rules, is incremental: outcomes are required first from the largest spenders for the largest projects, in annexures an outside agency signs, while the expenditure record continues underneath as the enforceable layer. Whether that layering eventually makes quality as examinable as quantum is the open question the next decade of the section will answer, and it is a question about administrative capacity as much as about law.

Verdict

Section 135 began as a disclosure obligation with a number attached and has become, by the amendments of 2019 and 2020 as commenced in 2021, a spending obligation with an escrow mechanism and a civil penalty. The namable claim of this article is that the statute compels the amount and stays almost silent on the effect: mandatory spending, voluntary quality. A board can discharge the section completely by spending badly on eligible projects or by transferring the shortfall to a Schedule VII fund, and the law designed to direct corporate resources towards social outcomes contains an escape hatch that converts it, at the point of failure, into a tax with extra paperwork. That is not a reason to dismiss the provision. The sums moved are large, the disclosure architecture is real, and the custody rules ended an era in which the explanation was the compliance. It is a reason to read the section as what it is: a fiscal instrument wearing the clothes of a governance reform, best operated with precision about its thresholds, its computation, its deadlines and its penalties, and best judged by a question its own reporting system is only beginning to ask, which is not how much was spent but what the spending changed.

For the two readers this article has kept in view, the practical yield is different in kind. The compliance reader leaves with a calendar and a set of classifications: test the three gates on the preceding year’s audited figures, reconstruct the profit base under Section 198, fix the ongoing project classifications inside the year, watch the two transfer dates in April and September, and treat the annexure, the form and the website as one record that must agree with itself. The critical reader leaves with the shape of an argument: that a state which distrusts both corporate voluntarism and its own spending machinery built a hybrid that compels the transfer and privatises the choice, and that every amendment since 2019 has strengthened the compulsion while leaving the choice, and its quality, where the original design put it. Both readings are true of the same text, and the text is short. What makes Section 135 a subject rather than a footnote is that the machinery around it, the accounts, the committees, the circulars, the assessments and the penalties that never were, records a decade of the Indian state learning, in public, how hard it is to make generosity compulsory.

Frequently Asked Questions

Q: Which companies must spend on CSR in India?

Every company that, during the immediately preceding financial year, had a net worth of Rs 500 crore or more, or a turnover of Rs 1,000 crore or more, or a net profit of Rs 5 crore or more must comply with Section 135 of the Companies Act, 2013. The three tests are alternatives, so meeting any one of them brings the company inside the regime. The duty covers holding and subsidiary companies on their own figures, and the CSR rules extend it to a foreign company with a branch office or project office in India in respect of its Indian business. Once inside, the company must constitute a CSR Committee, adopt a policy and spend the prescribed amount each year.

Q: How much must a company spend on CSR?

A covered company must spend, in every financial year, at least two per cent of its average net profits of the three immediately preceding financial years, with the profit figure computed under Section 198 of the Companies Act and adjusted by the exclusions in the CSR rules for overseas branch profits and for dividends from other Section 135 compliant companies. A company that has completed fewer than three financial years averages the years it has completed. There is no ceiling in the section, and an amount spent beyond the two per cent may be set off against the obligation of the three succeeding financial years if the board resolves accordingly.

Q: What happens to unspent CSR money?

Unspent money no longer stays with the company. If the shortfall relates to an ongoing project approved as such, it must be transferred within thirty days of the end of the financial year to a separate Unspent Corporate Social Responsibility Account in a scheduled bank, spent on that project within three financial years, and any residue then transferred to a fund specified in Schedule VII within thirty days of the third year’s completion. Any other unspent amount must be transferred to a fund specified in Schedule VII within six months of the expiry of the financial year. Failure to make either transfer attracts the penalty in Section 135(7).

Q: Which activities qualify as CSR under Schedule VII?

Schedule VII lists twelve heads: hunger, poverty, malnutrition, health care, sanitation and drinking water; education and vocational skills; gender equality, women’s empowerment and institutions for women, orphans, the elderly and backward groups; environmental sustainability and animal welfare; protection of national heritage, art and culture; benefits for armed forces veterans, war widows and families of central armed police personnel; sports training; contributions to the Prime Minister’s relief funds and other Central Government funds for socio-economic development and welfare; contributions to incubators and to publicly funded research in science, technology, engineering and medicine; rural development; slum area development; and disaster management. Activities in the normal course of business, political contributions, employee-only benefits, sponsorship for marketing and activities abroad are excluded.

Q: Is CSR non-compliance a criminal offence?

No. CSR non-compliance is a civil default that attracts a money penalty. The Companies (Amendment) Act, 2019 had enacted a provision punishing default with a fine on the company and imprisonment of up to three years for officers in default, but that provision was never brought into force: the Companies (Amendment) Act, 2020 replaced it with a civil penalty before the two sets of changes were commenced together on 22 January 2021. The penalty is twice the amount that should have been transferred to a Schedule VII fund or the Unspent CSR Account, or Rs 1 crore, whichever is less, on the company, and one-tenth of that amount or Rs 2 lakh, whichever is less, on each officer in default.

Q: Was CSR always mandatory in India?

No, in two senses. Before the Companies Act, 2013, corporate giving in India was voluntary, guided by board discretion and, from 2009, by voluntary guidelines issued by the Ministry of Corporate Affairs that carried no legal compulsion. Section 135 commenced on 1 April 2014 and made the spending figure statutory, but in its original form the only consequence of not spending was that the board had to explain the shortfall in its report. The obligation became mandatory in substance only when the amendments of 2019 and 2020, commenced on 22 January 2021, required unspent amounts to be transferred to a statutory account or fund and attached a penalty to a failure to transfer.

Q: Can CSR spending be claimed as a business expense?

No. An Explanation to Section 37(1) of the Income Tax Act, 1961, inserted by the Finance (No. 2) Act, 2014 with effect from the assessment year 2015-16, provides that expenditure on activities relating to corporate social responsibility under Section 135 of the Companies Act shall not be deemed to be expenditure incurred for the purposes of business or profession, so it cannot be deducted in computing business income. The disallowance is aimed at the general deduction provision. A contribution that independently meets the conditions of a specific provision, such as a donation qualifying under Section 80G, may still be deducted under that provision on its own terms.

Q: Who approves and monitors a company’s CSR policy?

The CSR Committee of the board formulates the policy and recommends it, together with the amount of expenditure and, under the rules, an annual action plan listing the projects, the manner of execution, the utilisation modalities, the implementation schedule and the monitoring mechanism. The board approves the policy and the plan, discloses the policy in the board’s report and on the company’s website, and remains responsible for ensuring that the money is spent as approved. The committee monitors implementation from time to time. Where the spending obligation does not exceed Rs 50 lakh, no committee is required and the board discharges these functions itself.

Q: How is net profit calculated for the CSR obligation under Section 135?

Net profit for Section 135 is calculated under Section 198 of the Companies Act, the provision that computes profits for managerial remuneration, and not taken directly from the profit and loss account. Section 198 strips out capital profits and similar non-operating credits, so gains on the sale of an undertaking or a revaluation surplus do not inflate the CSR base. The CSR rules then exclude two further items: profits arising from any overseas branch of the company, and dividends received from other companies in India that are covered by and comply with Section 135, which prevents the same profit being charged twice within a group. The two per cent is applied to the average of this figure across the three immediately preceding financial years.

Q: What is an ongoing project under the CSR rules?

An ongoing project is a multi-year project undertaken by a company in fulfilment of its CSR obligation, with a timeline not exceeding three years excluding the financial year in which it commenced. The definition also covers a project that was approved for a single year but whose duration the board later extended beyond that year on the basis of reasonable justification. The classification matters because only a shortfall relating to an ongoing project may be transferred to the Unspent Corporate Social Responsibility Account and spent over the following three financial years. A project must have been approved as ongoing, with its timeline and year-wise allocation recorded, before the year ends; the status cannot be created retrospectively to hold back an underspend.

Q: Can a company set off excess CSR spending against future years?

Yes. A proviso to Section 135(5), inserted by the Companies (Amendment) Act, 2020, allows a company that spends more than its two per cent obligation in a financial year to set off the excess against the amount it is required to spend in the three immediately succeeding financial years. Two conditions apply under the rules: the excess available for set-off must not include any surplus arising out of the CSR activities themselves, and the board must pass a resolution choosing the set-off. The election must be made in the year of the excess spending; a company that does not resolve to carry the excess forward cannot revive the claim in a later year. Set-off is the only carry-forward the regime permits, and it runs in one direction only.

Q: Does a company stop having CSR obligations if it falls below the thresholds?

Eventually, but not at once. The CSR rules provide that a company which ceases to meet the net worth, turnover and net profit criteria in Section 135(1) for three consecutive financial years is not required to comply with the section until it meets the criteria again. A single year below every threshold therefore leaves the committee, the policy, the spending obligation computed on earlier profits, and the reporting duties intact. The three-year rule cuts both ways: a company that re-enters by crossing any one threshold in a later year returns to the full regime for the following year, and the three-year exit clock starts again from that point if it later falls below the line.

Q: What is the annual action plan that a CSR committee must prepare?

The annual action plan is the committee’s year-specific programme, introduced by the Companies (CSR Policy) Amendment Rules, 2021 and recommended by the committee to the board for approval. It must cover the list of CSR projects or programmes approved for the year, the manner of execution, the modalities of utilisation of funds and the implementation schedules for the projects, the monitoring and reporting mechanism, and the details of the need and impact assessment, if any, for the projects undertaken. The board may alter the plan during the year on the committee’s recommendation where there is reasonable justification. The plan is the document against which the year’s spending, transfers and board report disclosures are later checked.

Q: Which companies must get an impact assessment of their CSR projects?

Every company with an average CSR obligation of Rs 10 crore or more over the three immediately preceding financial years must obtain an impact assessment, through an independent agency, of its CSR projects that have outlays of Rs 1 crore or more and that were completed not less than one year before the assessment is undertaken. The assessment reports must be placed before the board and annexed to the annual report on CSR in the board’s report. The rules cap the expenditure a company may book towards impact assessment, so the cost is itself a regulated part of the CSR outlay. Companies below the thresholds may assess voluntarily, and many do, but no obligation or penalty attaches to the choice.

Q: Can a company carry out CSR through an implementing agency or NGO?

Yes, through defined channels. A company may implement through a Section 8 company, a registered public trust or a registered society, whether established by the company itself or independent, through an entity established by the Central or a state government, or through an entity created by an Act of Parliament or a state legislature. An independent agency must have a track record of at least three years in similar activities, and the agency must be registered under Sections 12A and 80G of the Income Tax Act. Since 1 April 2021, every implementing agency must also register on the Ministry of Corporate Affairs portal in Form CSR-1 and obtain a CSR registration number before receiving funds.

Q: Who certifies that CSR funds have been used as the board approved?

The chief financial officer of the company, or the person responsible for financial management where there is no chief financial officer. The CSR rules make the board responsible for satisfying itself that the funds disbursed for CSR have been utilised for the purposes and in the manner the board approved, and the certification by the finance head is the mechanism by which that assurance is recorded. The duty was introduced by the 2021 amendment of the rules and moved CSR assurance into the finance function: utilisation certificates from implementing agencies, milestone reports and project evidence are collected so that the certificate can be given. The certificate supports the board’s own monitoring; it does not transfer the board’s responsibility to the officer who signs it.

Q: What are administrative overheads in CSR, and how much is allowed?

Administrative overheads are the expenses a company incurs for the general management and administration of its CSR functions, such as the salaries of staff who run the CSR programme as a whole and the office costs of the CSR department. They do not include expenses directly incurred in designing, implementing, monitoring or evaluating a particular project, which are project costs. The board must ensure that administrative overheads do not exceed five per cent of the company’s total CSR expenditure for the financial year. Spending above the cap does not count towards discharging the two per cent obligation, so an overhead-heavy programme has to find its project spending within the balance of the amount.

Q: What happens to surplus generated by a CSR project?

A surplus arising out of CSR activities is not business profit of the company and cannot be retained as income. The rules require it to be ploughed back into the same project, or transferred to the Unspent Corporate Social Responsibility Account, or transferred to a fund specified in Schedule VII, in accordance with the company’s CSR policy and the annual action plan. Interest earned on money sitting in the Unspent CSR Account is treated the same way, as money of the project rather than money of the company. The surplus also cannot be counted within the excess spending that a company sets off against later years; the set-off is confined to amounts actually spent beyond the two per cent obligation.

Q: Can CSR money create a capital asset, and who must own it?

Yes, CSR funds may be spent on creating or acquiring a capital asset, such as a school building, a water plant or medical equipment for a charitable hospital, but the company itself must not own it. The rules require the asset to be held by a Section 8 company, a registered public trust or a registered society with charitable objects and a CSR registration number, or by the beneficiaries of the project in the form of self-help groups, collectives or entities, or by a public authority. The rule stops a company from discharging its obligation by building an asset on its own balance sheet, and it means the choice of the holding entity must be settled before the asset is created, not after.

Q: What must the board’s report disclose about CSR?

The board’s report must include an annual report on CSR in the format prescribed by the rules, and the contents of the CSR policy must be disclosed there and placed on the company’s website. The annual report states the composition of the CSR Committee, the average net profit and the prescribed two per cent amount, the amount spent on projects and on administrative overheads, any amount set off, the details of ongoing and other projects, the amounts transferred to the Unspent CSR Account and to Schedule VII funds, the reasons for any failure to spend the prescribed amount, and the responsibility statement of the committee. A parallel report in Form CSR-2 is filed with the Registrar of Companies with the annual filings.