The Mahatma Gandhi National Rural Employment Guarantee Act 2005, known everywhere as MGNREGA, rests on a reversal so simple that its consequences are still not fully absorbed by the administrative system built to serve it. Every public works programme before it began with an allocation: a sum of money was sanctioned, a set of works was approved, and employment existed for as long as the sanction and the works lasted. A worker in such a programme holds nothing. The worker waits to be selected, and a season in which no work is opened in the village is a season in which the programme has not failed anyone in particular, because it promised work to no one in particular. The employment guarantee statute begins at the opposite end. It begins with a household that asks. The asking triggers a duty, the duty carries a deadline, and the deadline carries a price for missing it, payable to the very household that was kept waiting. The volume of employment under the Act is therefore meant to be a function of registered demand rather than of budgetary allocation, and that single design choice explains the statute’s strength, its administrative difficulty, and the precise point at which it can be defeated without any official ever recording a refusal.

A reader who understands that reversal understands the Act. The hundred days figure, which dominates public discussion, is in truth the least distinctive part of the design, because employment programmes had announced day counts for decades. What no earlier national programme had done was to make the day count an entitlement of a named household, enforceable through a sequence of dated acts: registration, the issue of a job card, a written application, a dated receipt, an obligation to provide work within fifteen days, an unemployment allowance where the obligation is missed, wages payable within a fixed period after the work is done, and compensation where the wage itself is late. Each step leaves a document, and each document converts an administrative intention into a fact that can be proved. The statute is best read as a machine for producing such facts, because a guarantee that generates no evidence of its own breach is a guarantee that cannot be enforced by the person holding it.

The field within which the Act sits, the wider body of welfare legislation in which some benefits rest on statutes and others on executive schemes, is mapped in the complete guide to welfare laws in India, which owns the classification of the field as a whole. The present article owns the statute itself: its text, its clocks, its money, its machinery, and the reasoning by which its contested choices should be judged. The evidence of how the guarantee has performed on the ground, district by district and year by year, belongs to the implementation record and is deliberately kept at the level the statutory text supports in the pages that follow. A pillar article that tried to settle the performance question would cease to be a pillar, and a reader who wants the verdict on delivery should follow the pointer given in the assessment section below rather than expect the record to be re-argued here.

What existed before the guarantee, and why it failed the worker

Rural public works in India are older than the Republic. The famine codes of the colonial period treated relief work as the test of destitution and the instrument of relief in a single device: a person willing to labour on a public work at a subsistence wage had demonstrated need, and the work itself delivered the relief. Independent India inherited the device and multiplied it into a sequence of national programmes. The National Rural Employment Programme of 1980, the Rural Landless Employment Guarantee Programme of 1983, the Jawahar Rozgar Yojana of 1989, the Employment Assurance Scheme of 1993 and the Sampoorna Grameen Rozgar Yojana of 2001 each carried employment among their objects, and each was built as a budgetary programme governed by guidelines. Targets were expressed as person-days to be generated. Funds flowed down through administrative tiers against approved plans. Whether work existed in a given village in a given month depended on whether a work had been sanctioned, whether funds had reached the executing agency, and whether the season and the local administration aligned.

The defect of that architecture was not corruption, though corruption flourished inside it, and it was not underfunding, though funding was chronically thin. The defect was juridical. No worker under any of those programmes held a claim. A household that presented itself for work and was turned away had been disappointed, not wronged, because the programme documents promised employment in aggregate and to no one in particular. There was no application to file, no receipt to demand, no period within which work was owed, and no consequence priced into the system for the work not existing. Accountability ran upward through reports of expenditure and person-days, and a district that generated few person-days had performed poorly in a statistical sense without breaching any duty owed to any identifiable person. The programmes could therefore contract silently in exactly the places and seasons where distress was deepest, because distress generated no legal signal inside them.

Maharashtra broke that pattern a generation earlier. The Maharashtra Employment Guarantee Act of 1977 placed a state level guarantee on a statutory footing, funded in part through dedicated taxation, and its operation over more than two decades supplied the national debate with the one thing argument alone could not supply: a working demonstration that a government could bind itself by law to provide work on demand and administer the obligation at scale. When the national legislation was framed, the Maharashtra experience stood behind it as proof of administrability, and the older national programmes stood behind it as proof of the alternative. The Act of 2005 is best understood as the generalisation of the Maharashtra experiment and the repudiation of the programme sequence. Parliament examined four decades of schemes and enacted the thing each scheme had stopped short of being.

The political route mattered to the text. A guarantee of employment figured in the Common Minimum Programme on which the governing coalition of 2004 rested, and the drafting drew on a wider right to work movement that had spent years specifying, in unusual detail, what a statutory guarantee would have to contain in order to be real rather than rhetorical. That movement’s fingerprints are visible in the finished Act, in the dated receipt, in the allowance, in the social audit, and in the insistence that the demand side of the programme be documented as carefully as the expenditure side. Statutes drafted under campaign pressure often carry ornamental provisions that administrators quietly ignore. The provisions in this Act carry the opposite character. They are operational instructions aimed at the exact points where a guarantee dies in practice, written by people who had watched guarantees die.

The shape of the Act and where its weight sits

The statute was enacted as the National Rural Employment Guarantee Act, 2005, Act 42 of 2005. It received the assent of the President on 5 September 2005 and was published in the Gazette on 7 September 2005. In 2009 an amendment renamed it after Mahatma Gandhi, and the longer name, the Mahatma Gandhi National Rural Employment Guarantee Act, is the name the statute has carried since that renaming took effect on 2 October 2009. The renaming changed no obligation. The Act’s own commencement was staged by notification, and the staging is part of its legal identity. It was brought into force on 2 February 2006 in two hundred districts, chosen from among the most backward in the country. It was extended during the financial year 2007 to 2008 to a further one hundred and thirty districts, one hundred and thirteen of them from 1 April 2007 and seventeen districts in Uttar Pradesh from 15 May 2007. From 1 April 2008 the remaining rural districts were notified, and the guarantee has since covered the entire country apart from districts that are wholly urban in character.

The architecture of the Act follows the logic of the entitlement. The opening sections state the guarantee and the obligation to frame a scheme. A group of sections then constructs the councils at the Centre and in the states, the officers who coordinate the programme at district level and administer it at block level, and the panchayat institutions through which applications are received and works executed. The funding pattern and the employment guarantee funds are dealt with as legal provisions rather than budgetary conventions. Two Schedules complete the design, and they carry more of the statute’s character than Schedules usually do. Schedule I lists the permissible works, tying the wage to a defined class of rural assets. Schedule II sets the conditions of entitlement and of work, the application procedure, the distance rule, the worksite facilities, the wage and material ratio, the bar on contractors, the share of women among beneficiaries, and the compensation for delayed wages. A ministry circular cannot lawfully rewrite any of that, because Parliament put it in the statute and its Schedules rather than in guidelines.

Why is the hundred day figure not the heart of the Act?

Because employment programmes announced day counts for decades without creating any entitlement, and the hundred days become a right only through the machinery around them: the household unit, the application, the fifteen day duty, the allowance for failure and the wage clocks, which together convert a number into a claim a named family can enforce.

The guarantee itself, stated precisely, is that the state government shall provide not less than one hundred days of wage employment in a financial year to every rural household whose adult members volunteer to do unskilled manual work, in accordance with the scheme the state frames under the statute. That statement repays slow reading, because each element of it forecloses a common misunderstanding. The guarantee belongs to the household, not to the individual worker, so the hundred days are shared among the adult members of one household who volunteer. The work is unskilled manual work, which is the self selecting feature of the design: the character of the work and its wage are meant to draw those who need it and to leave those with better options to take them. The areas covered are rural areas as notified, which is why the Act is a rural guarantee and creates no urban entitlement. The financial year is the accounting frame, so the entitlement renews and expires with that frame. And the phrase in accordance with the scheme is not a dilution. Section 4 obliges every state government to frame a scheme for giving effect to the guarantee, and the scheme operates inside the Act’s conditions, not above them. A state scheme that contradicted Schedule II would contradict the statute itself.

Above the hundred days sits a further provision that public discussion routinely misses. Section 3 permits the Central Government, in areas affected by drought or other natural calamity, to notify additional days of employment beyond the hundred, and fifty additional days have been the notified measure in such areas. The hundred days are therefore a floor in the statute’s own terms, and the Supreme Court’s proceedings in drought litigation, examined later in this article, treated the enhanced entitlement in notified drought areas as part of the obligation rather than as an act of grace. The design assumes that distress is seasonal and geographic, and it builds the expansion valve into the section that creates the guarantee, so that deepening the guarantee in a drought district requires notification under the Act and not fresh legislation.

Registration, the job card and the application that starts every clock

The guarantee is set in motion by a sequence of administrative acts, and the sequence deserves to be stated as a sequence because the order is the substance. Nothing in the Act is self executing in favour of a household that has not entered the register. The first act is registration. Adult members of a rural household apply for registration to the Gram Panchayat of the village where they reside, giving names, ages and the particulars the scheme requires. The Gram Panchayat verifies the application, and registration is the entry of the household into the programme’s universe. The second act is the issue of the job card. The registered household receives a job card bearing the photographs of its adult members, and the card is at once an identity document, a record of employment provided and wages paid, and the household’s proof of standing inside the system. Schedule II obliges the issuing authority to provide the card free of cost, and the card is meant to remain in the custody of the household rather than of any official, a custody rule whose importance becomes plain wherever cards have been retained by intermediaries and the record of work has drifted out of the worker’s hands.

The third act is the application for work, and it is the hinge of the whole statute. A registered adult member may apply in writing to the Gram Panchayat or to the Programme Officer, stating the period and the duration for which employment is sought. Schedule II requires that an application seek a minimum block of work, fourteen days of continuous employment, so that the system is asked for a real engagement and not for a single day’s casual labour. Applications may be made in advance, naming a future date from which work is wanted, which is the provision that allows a household to plan around the agricultural calendar rather than present itself in desperation at the point of maximum distress. A group application may also be made. The receiving authority is bound to accept a valid application and to issue a dated receipt for it. Hold that receipt in mind. Every later entitlement in the Act is timed from the application it acknowledges, so the receipt is the document that converts a conversation at a panchayat office into the first day of a legal countdown.

The fourth act belongs to the administration. The applicant is to be intimated in writing at the address recorded on the job card, and public notice is displayed at the panchayat offices, when work is provided and where. The fifth act is the work itself, and the sixth is the payment, each governed by its own period and examined in the sections that follow. What matters at this stage is the character of the chain. Registration without an application generates no duty to provide work on any particular date. An application without a receipt generates a duty that cannot be timed and therefore cannot be proved to have been breached. The statute’s authors understood that a demand driven system is only as strong as its record of demand, and they wrote the receipt into the Schedule for exactly that reason. The point returns, with its full force, in the claim this article advances at its close.

How does a household move from registration to a legal claim for work?

A household registers with the Gram Panchayat and receives a job card, then an adult member applies in writing for a stated period of work, seeking at least fourteen continuous days, and receives a dated receipt, from which the fifteen day period for providing employment begins to run in favour of that household.

Registration carries its own disciplines, which protect the entitlement from the register itself becoming a source of exclusion. The job card remains valid for the period the scheme specifies and is renewable, and entries on it, of work demanded, work provided and wages paid, are to be updated as events occur. A household whose members migrate for part of the year does not lose its registration by absence, and the application mechanism is designed to be used on return, in advance where the return can be foreseen. Deletion of a registration is a governed act, not an informal one, because a deleted household has lost the standing from which every claim in the Act proceeds. The register is, in legal terms, the roll of rights holders, and its maintenance is therefore a legal function of the Gram Panchayat and not a clerical afterthought. Where the register is wrong, the guarantee is wrong at its root, and no later efficiency in works execution can cure the defect.

There is a further feature of the application design that marks the distance between this statute and an ordinary employment programme. There is no ceiling in Schedule II on the number of days for which a person may apply, subject always to the household’s aggregate entitlement of one hundred days in the financial year. A household may therefore apply in several blocks across the year, matching its demand to the lean seasons when other work fails, and the administration’s duty revives with each application. Demand under the Act is not a single annual event to be captured in a survey. It is a continuing legal act that households perform as their circumstances change, and a system that treats the labour budget prepared at the start of the year as the outer limit of demand has inverted the statute, because the budget is meant to be built from anticipated demand and revised against registered demand, not used as a reason for refusing to register it.

The four clocks of MGNREGA: the deadlines on which the guarantee stands

Strip the Act to its operative skeleton and four periods remain, and a worker who knows them holds the working content of the guarantee. The first clock governs the provision of work. Employment must be provided within fifteen days of the receipt of an application, or, where the application is made in advance, within fifteen days of the date from which employment is sought, whichever is later. The second clock governs the unemployment allowance. Where employment is not provided within that fifteen day period, the allowance entitlement accrues as a daily entitlement under Section 7, and the state’s liability continues until one of the events Section 7 specifies brings it to an end, the direction to report for work, the expiry of the period sought, the household reaching its hundred days, or the household’s combined wages and allowance reaching the wage value of one hundred days. The third clock governs wages. Section 3 requires disbursement of wages on a weekly basis and in any event within a fortnight of the date on which the work was done, and Schedule II frames the payment obligation around the closure of the muster roll, the attendance and measurement record for the work performed. The fourth clock governs compensation for the wage itself arriving late. Paragraph 29 of Schedule II entitles the wage seeker to compensation at the rate of 0.05 per cent of the unpaid wages for each day of delay beyond the fifteenth day from the closure of the muster roll, and it divides the payment process into stages, measurement, computerisation of muster rolls and of measurements, generation of wage lists and the uploading of fund transfer orders, so that responsibility for delay can be assigned to the stage and the functionary where the delay occurred.

Statutory clock	What starts it	Who must act	What is owed if it is missed	How the worker claims it
Provision of employment: fifteen days	Receipt of a written application, or the later date from which work is sought in an advance application	Gram Panchayat and Programme Officer	The daily unemployment allowance under Section 7	Produce the dated receipt and show that no work was allotted within the period
Unemployment allowance: daily from the sixteenth day	Expiry of the fifteen day period without work being provided	State government, which bears the cost	Not less than one fourth of the wage rate for the first thirty days and not less than one half for the remaining period of the financial year	Claim on the application record; liability ends only on a Section 7 event
Payment of wages: weekly, and within a fortnight of the work; fifteen days from muster closure in the Schedule II process	Performance of the work and closure of the muster roll	The implementing agency and the Programme Officer, through the stage wise payment process	The wage in full, and compensation for any further delay	Match the job card and muster entry against the wage credit
Delay compensation: 0.05 per cent of unpaid wages per day of delay	The sixteenth day after closure of the muster roll without payment	The state government under its compensation rules, with stage wise culpability fixed on functionaries	Compensation calculated on the unpaid amount for each day of delay, itself treated as delayed wages if paid late	Automatic calculation under the rules, claimed on the wage record where it is not paid

The table is the article’s findable artifact because it is the object a worker, a panchayat secretary and a programme officer can each use without translation. Notice its internal logic. The clocks are chained. The receipt starts the first clock, the failure of the first starts the second, the work performed starts the third, and the failure of the third starts the fourth. At no point does the chain depend on an official conceding that something has gone wrong. Each transition is produced by the calendar acting on a document. That is what it means for a guarantee to be drafted as law rather than as aspiration, and it is also why the system’s weakest point is at the very first link. A clock that was never started cannot be shown to have run out.

The comparison with a neighbouring entitlement sharpens the point. The food security statute builds its claim around identification and supply: a household is placed on a list, grain is allocated against the list, and the entitlement is breached when the supply fails a household that the state itself has recorded. The employment guarantee builds its claim around an act of the citizen. The state’s own record, in the food system, is the foundation of the claim, as the guide to the National Food Security Act 2013 sets out in its treatment of that statute’s identification machinery. Under the employment Act, the citizen’s application is the foundation, and the state’s receipt is merely its acknowledgement. Entitlements founded on the state’s list fail when the list is wrong. Entitlements founded on the citizen’s demand fail when the demand is not recorded. The two failure modes are different, and confusing them produces reform proposals aimed at the wrong link of the chain.

The unemployment allowance: the sanction that makes the duty real

A duty without a price for breach is an instruction. Section 7 prices the breach. If an applicant is not provided employment within fifteen days of receipt of the application, or of the date from which employment was sought in an advance application, whichever is later, the applicant is entitled to a daily unemployment allowance. The allowance is payable to the applicants of a household subject to the household’s entitlement, at a rate the state government specifies by notification in consultation with the State Employment Guarantee Council, and the Act fixes the floor beneath that rate. The allowance may not be less than one fourth of the wage rate for the first thirty days during the financial year, and not less than one half of the wage rate for the remaining period of the financial year. The escalation is deliberate drafting. A short failure to provide work is priced at a quarter of the wage. A long failure is priced at half, so that the longer the administration leaves a household without the work it demanded, the more expensive the failure becomes relative to the wage bill the administration avoided.

Section 7 then defines, with unusual care, when the liability stops. It ceases when the applicant is directed by the Gram Panchayat or the Programme Officer to report for work, either personally or by deputing an adult member of the household. It ceases when the period for which employment was sought comes to an end and no member of the household turned up for the employment. It ceases when the adult members of the household have received one hundred days of work in the financial year between them. And it ceases when the household’s wages and allowance taken together equal the wages for one hundred days of work. The allowance is payable jointly to the household’s applicants and is sanctioned and disbursed in the manner the scheme prescribes, and where allowance is paid or due, Schedule II requires the Programme Officer to inform the District Programme Coordinator in writing of the reasons why employment could not be provided. That reporting duty deserves attention, because it converts each allowance payment into an administrative admission, recorded in writing, that the guarantee failed at a named place for a stated reason. A well administered allowance would generate, as its byproduct, a complete register of the guarantee’s failures and their causes. The allowance is therefore both a compensation and an audit instrument, and its rarity in practice, a matter the implementation record owns, should be read against both functions at once.

Why does the allowance rate rise after thirty days?

Because the Act prices a long failure more heavily than a short one: the allowance begins at not less than one fourth of the wage rate for the first thirty days and rises to not less than one half of the wage rate for the remaining period, so delay grows costlier the longer the administration leaves the household without work.

Two features of the section are regularly misstated and should be fixed precisely. The allowance is charged to the state government, under the funding pattern examined below, and not to the Centre. That placement is a considered incentive. The Centre funds the wages of work actually provided, so a state that provides work draws central money into its villages, while a state that fails to provide work pays the allowance from its own resources. The fiscal architecture rewards delivery and taxes failure, and any account of the allowance that omits who pays it has omitted the mechanism by which it is meant to discipline state administrations. The second feature is the qualifier the section itself carries. The allowance is payable subject to the terms of eligibility the state prescribes and, in the section’s words, subject to the economic capacity of the state government. Critics of the drafting have long identified that phrase as a softness at the centre of the sanction, a textual handhold for a state pleading inability. Defenders answer that the rate floors are unqualified and that a state which has framed its scheme under the Act cannot plead its own budgetary choices as incapacity without inviting judicial scrutiny. The honest reading is that the entitlement is mandatory in form, floored in amount, and attended by one phrase whose weight a court would have to measure on facts, which is a stronger position than any programme guideline ever gave an Indian worker and a weaker one than the guarantee’s rhetoric suggests.

The wage: Section 6, the sixty rupee floor and the delinking question

Section 6 is the most contested single provision in the statute, and it is short. Notwithstanding anything contained in the Minimum Wages Act 1948, the Central Government may by notification specify the wage rate for the purposes of this Act, and different rates may be specified for different areas. A wage rate so specified may not be less than sixty rupees per day. Until a rate is fixed by the Centre for an area, the minimum wage fixed by the state government under the Minimum Wages Act for agricultural labourers is the wage rate applicable to that area. Three consequences follow from that text. The programme’s wage is a wage under this Act, set by notification, and it is not automatically the agricultural minimum wage of the state in which the work is done. The sixty rupee figure is a statutory floor of 2005 vintage, overtaken in practice by notified rates, whose legal function is to mark the boundary below which no notification may go. And the default rule runs in the opposite direction from the one most summaries state: the state minimum wage governs only in the interval before the Centre notifies a rate for the area, not as a permanent rule of the programme. The relationship between the notified programme wage and the agricultural wage it displaced is traced in the article on the Minimum Wages Act 1948, which owns the older statute’s own fixation machinery.

The delinking was a choice, and the statute should be judged with both cases at full strength. The case for a notified central rate begins with the funding pattern. The Centre bears the entire cost of unskilled wages under the Act. A wage defined as whatever each state fixes as its agricultural minimum would make the Centre’s largest welfare liability a variable set by twenty eight separate state decisions, taken for reasons internal to each state’s labour politics and revisable at any time. Fiscal planning for an open ended, demand driven entitlement is difficult in the best conditions, and tying its unit price to other governments’ notifications would add a layer of unpredictability that no finance ministry would accept in the design of a programme it must fund without a statutory ceiling. The supporting argument concerns labour markets. A programme wage set mechanically at or above the agricultural wage in states where that wage is high relative to prevailing earnings could pull labour out of agriculture at peak seasons and distort the very market on which the same households depend for the rest of the year, while a centrally notified rate, revised by reference to price movements, can be calibrated state by state with the programme’s self selecting design in view.

The case against the delinking is shorter and, on its own ground, harder to answer. The state, acting through the same legal order, requires every private employer of agricultural labour to pay not less than the minimum wage, and prosecutes or penalises payment below it. When the state then engages labour on its own works under a statutory programme at a notified rate below that minimum, it occupies both sides of the rule it enforces. The worker experiences the contradiction without any need for legal theory: the wage the law calls the minimum when a landlord pays it becomes optional when the government pays it. That argument has been litigated in the wider field of minimum wage law, where the Supreme Court’s older jurisprudence treated payment below the minimum wage, exacted under economic compulsion, as falling within the constitutional prohibition of forced labour, and it retains its force as a criticism of design even where the notified rates in a given state have in practice converged with or moved past the agricultural minimum. The balanced verdict the text supports is this. Section 6 is intelligible as fiscal engineering and defensible as such, and its defenders are entitled to insist that the Centre cannot write a blank cheque against other governments’ wage notifications. But a provision that permits the programme wage to sit below the state’s own agricultural minimum places a burden of justification on every notification that does so, and that burden is not discharged by pointing to the section that made it possible. The provision’s existence explains how a below minimum programme wage can be lawful. It does not establish that any particular instance of it is just.

Equal wages stand on different footing, and here the Act admits no qualification. Schedule II prohibits discrimination on grounds of gender in the payment of wages and requires that equal wages be paid to men and women for the same work, and the wage rate notified under Section 6 is a rate for the work, not for the worker. In a labour market where women’s agricultural wages have persistently run below men’s for comparable tasks, a statutory programme paying a single notified rate is itself an intervention, and the design choice interacts with the requirement, also in Schedule II, that at least one third of the beneficiaries to whom work is provided be women. The two provisions together moved the programme’s wage from a market reflection into an instrument of wage equality in the villages where it operates, which is one of the clearer instances of the Schedules doing substantive social legislation under the appearance of programme conditions.

Schedule I: the works the wage is allowed to build

A guarantee of employment could have been drafted as a guarantee of days alone, leaving the content of the work to administrators. Schedule I refuses that economy. It lists the permissible works, and the list is a development programme compressed into a Schedule: water conservation and water harvesting; drought proofing, including afforestation and tree plantation; irrigation canals, including micro and minor irrigation works; provision of irrigation facilities to land held by the households the Schedule names, including Scheduled Caste and Scheduled Tribe households, beneficiaries of land reform and of the rural housing programme; renovation of traditional water bodies, including desilting of tanks; land development; flood control and protection works, including drainage in waterlogged areas; rural connectivity to provide all weather access; and further works notified by the Central Government in consultation with the state governments. The centre of gravity is unmistakable. The Schedule directs the wage bill at water, land and the productive base of the village economy, so that the expenditure which relieves distress in a lean season is required to leave behind an asset addressed to the causes of the next one.

The list performs three legal functions beyond its development content. It limits discretion: a work outside the Schedule is not a work under the Act, and wages paid on it are not wages under the guarantee, which gives auditors and social audits a textual standard against which a muster roll can be checked. It shapes the labour intensity of the programme from the works side, since the named categories are works executable predominantly by manual labour with limited material input, which connects the Schedule to the wage and material ratio examined below. And it distributes the benefits of the assets in a deliberate direction, because land development and irrigation on the holdings of the households the Schedule names convert a wage programme into an instrument that raises the productive capacity of precisely the families most likely to appear on its muster rolls as workers. A worker may therefore meet the same statute twice in a single season, once as a seller of labour on a village work and once as the holder of land improved by such a work, and that doubling is a feature of the Schedule’s design rather than an accident of targeting.

Schedule II: the conditions that make the work a right rather than a favour

If Schedule I governs what is built, Schedule II governs how a human being is treated inside the programme, and its conditions should be read as rights of the worker stated in the grammar of programme administration. The application provisions have already been examined: written application to the Gram Panchayat or the Programme Officer, a minimum block of fourteen days of continuous work, advance applications permitted, group applications permitted, and a dated receipt owed on every valid application. The distance rule follows. As far as possible, employment is to be provided within a radius of five kilometres of the village where the applicant resides at the time of applying. Where employment is provided outside that radius, it must be provided within the block, and the workers must be paid ten per cent of the wage rate as extra wages to meet the additional transport and living costs. The rule prices distance. A distant work is not forbidden, because a village cannot always generate a permissible work within five kilometres on demand, but the administration that sends workers further away must pay for the displacement it imposes, and a worker offered distant work without the addition is being offered less than the statute promises.

A period of employment is ordinarily at least fourteen days continuously, with not more than six days of work in a week, a rhythm that inserts a weekly day of rest into the guarantee’s own text. The worksite facilities are stated as entitlements of presence at the site: safe drinking water, rest periods and shade, a first aid box with material adequate for emergency treatment of minor injuries and other health hazards connected with the work, a medical aid obligation where a worker meets with an accident in the course of employment, and compensation provisions that reach, in the gravest cases, the worker’s family. Where the number of children below the age of six accompanying women workers at a site is five or more, one of the women workers is to be deputed to look after the children and is paid wages for that duty at the applicable rate. That creche provision is small in word count and diagnostic in character. Its presence in the Schedule announces that the statute’s imagined worker is not an unencumbered individual but a parent, usually a mother, whose labour arrives at the worksite with her children in tow, and that a programme which made no provision for the children would exclude her as effectively as a refusal.

The structural conditions complete the Schedule. Contractors are barred from the execution of works under the programme, and the use of labour displacing machinery is prohibited, two prohibitions that defend the wage bill against the two channels through which public works money historically leaked away from workers: the intermediary who captured the margin between sanction and payment, and the machine that converted a labour budget into an equipment hire. The wage and material ratio gives the same defence a number. The cost of the material component, including the wages of skilled and semi skilled workers, is held to a proportion that preserves the programme’s labour intensity, expressed in the Schedule and the scheme framework as a sixty to forty balance between wages and material at the district level, so that a district cannot quietly convert an employment programme into a materials programme while continuing to report it as employment. Priority in the provision of work is to be arranged so that at least one third of the beneficiaries are women who have registered and requested work. And wages are to be disbursed through the accounts the framework prescribes, into bank or post office accounts of the workers, a channel condition whose object is the same as the contractor bar: to shorten the distance between the wage as sanctioned and the wage as received to a single auditable step.

Which conditions of work does the statute itself fix?

The statute fixes the five kilometre distance rule with ten per cent wages beyond it, a working week of not more than six days, drinking water, rest and first aid, a creche worker where five or more children below six attend, a bar on contractors and labour displacing machinery, and a wage and material balance keeping the programme labour intensive.

Read together, the two Schedules answer the objection that the guarantee is an open ended promise to pay for unspecified activity. The promise is specified at every joint: the works by category, the conditions by paragraph, the money by section, and the time by four clocks. What the statute leaves open is only the volume, and it leaves the volume open on purpose, because the volume is the demand.

The machinery: sabhas, panchayats, officers and councils

The Act distributes its functions across a vertical chain of institutions, and the distribution follows the entitlement’s logic: the acts closest to the worker are placed in the institution closest to the worker, and each higher tier exists to supply, supervise or review what the tier below it does. At the base stands the Gram Sabha, the assembly of the village’s voters, whose constitutional setting in the panchayat system is treated in the article on the 73rd Amendment and panchayati raj. Under this Act the Gram Sabha recommends the works to be taken up and determines the order of their priority, monitors the execution of works within the Gram Panchayat, and conducts the social audit. Section 17 gives the audit its statutory home: the Gram Sabha shall monitor the execution of works within the Gram Panchayat, and every Gram Panchayat is required to conduct regular social audits of all works executed within its area, in the manner the audit framework prescribes, so that the records of a work, its muster rolls, its measurements and its payments are read out and verified before the community in whose name the money was spent. The audit is therefore not an external inspection visited upon the village. It is a function of the village’s own assembly, and its findings carry into the programme’s record through the Gram Panchayat and the Programme Officer.

The Gram Panchayat is the programme’s working office at village level. It receives applications for registration, verifies them and registers households. It issues job cards. It receives applications for work, issues the dated receipts on which the guarantee’s clocks depend, and allots work within the statutory period. It identifies and plans works, building the shelf of projects from which allotments can be made without delay when demand arrives, and it executes works itself: the Programme Officer is required to allot at least one half of the works, in terms of cost, to the Gram Panchayats for implementation, so that the institution which registers the demand is also the institution that carries out at least half of the response to it. That coupling is the Act’s wager on local government. It concentrates registration, receipt, planning and execution in one elected body whose members face the applicants in the street, and it accepts the risks of that concentration in exchange for a proximity that no line department can reproduce.

Above the panchayat stand the programme’s officers. The Programme Officer, appointed at block level, coordinates the programme across the panchayats of the block, receives applications where the scheme so provides, ensures allotment of work within the period, consolidates plans, and supervises the payment process through its stages. The District Programme Coordinator, the office held by the district’s chief civil authority, coordinates the programme across the district, approves the district plan and labour budget, reviews the allowance cases reported upward, and answers for the district’s compliance with the Act’s periods. At state level, the State Employment Guarantee Council, constituted under Section 12, advises on the scheme, reviews its operation and connects the programme to the state government that bears the statutory duty. At the Centre, the Central Employment Guarantee Council, constituted under Section 10, performs the corresponding advisory and review functions for the country and for the government that bears the wage bill. The councils matter beyond their advisory form, because the Supreme Court’s directions in the drought litigation, examined below, treated the constitution of these councils as a statutory obligation whose neglect was itself a breach to be remedied on a timetable.

Who runs the programme at each level, and who watches them?

The Gram Sabha recommends works and conducts social audits, the Gram Panchayat registers households, issues job cards and receipts and executes at least half the works by cost, the Programme Officer administers the block, the District Programme Coordinator answers for the district, and the state and central councils review the whole under the Act.

Around this chain the Act builds its funds and its reporting. The money moves through employment guarantee funds at the Centre and in the states, created by the Act’s later sections, so that the programme’s finance has a statutory vessel rather than a place in a departmental budget head alone. The reports move upward through the same tiers that execute the work, and the social audit moves across them, which gives the system two different kinds of knowledge about itself: the administrative knowledge of expenditure and person-days, travelling up, and the audit knowledge of what the records look like when read against the ground, travelling from the village outward. A statute that contained only the first would know what it spent. This one is designed, at least in text, to know what it did.

The funding pattern: Section 22 and the price of failure

Section 22 writes the bargain between the Centre and the states into the statute, and the bargain deserves to be read as part of the guarantee rather than as its bookkeeping. The Central Government meets the entire amount required for payment of wages for unskilled manual work under the scheme. It meets up to three fourths of the material cost of the scheme, including the payment of wages to skilled and semi skilled workers, subject to the provisions of Schedule II. It also meets such percentage of the total cost towards administrative expenses as it determines, within which the salaries of Programme Officers and their staff, the expenses of the Central Council and the facilities required by Schedule II may fall. The state government meets the cost of the unemployment allowance payable under the scheme. It meets one fourth of the material cost, including skilled and semi skilled wages. And it meets the administrative expenses of the State Council.

Every element of that pattern is an incentive wearing the dress of an accounting rule. The Centre’s assumption of the whole unskilled wage bill removes the standard state objection to a demand driven programme, that meeting demand would break the state exchequer, because the marginal cost to a state of providing one more day of unskilled work is close to the material share and administration rather than the wage itself. The state’s assumption of the allowance does the opposite work, attaching the state’s own money to the single outcome the guarantee exists to prevent. Between the two, the material share, three fourths to one fourth, keeps both governments invested in the quality and honesty of the works: neither can treat the asset side of the programme as the other’s money. A funding pattern in a mere scheme could be revised by either government in a budget circular. Section 22 can be revised only by Parliament, which is why the welfare guide treats the statutory funding share as part of the entitlement’s legal design and why disputes about timely release of the Centre’s share are arguments about compliance with an Act and not only about the courtesies of fiscal federalism.

One tension inside the pattern should be stated honestly, because it is structural and no reading of the section dissolves it. The Centre’s liability is open ended in law, since it follows demand, yet it is discharged through annual appropriations, labour budgets approved in advance and releases made in tranches. The statute’s demand logic and the budget’s supply logic therefore coexist inside the same programme, and the friction between them is not an aberration but a permanent feature of the design, which the courts have had to police at exactly the seam where a state’s wage bill waited on a release the budget process had delayed. The Act gives the worker a right against the state government. It gives the state a funding claim against the Centre. The worker’s clock, however, runs against the state regardless of where the state’s own claim stands, and that ordering, awkward as it is for state administrations, is what keeps the entitlement pointed at the citizen rather than dissolving into an intergovernmental account.

Grievance, the ombudsman and the evidentiary life of the programme

A guarantee administered by the institutions it obliges needs a channel through which a failure can be raised without leaving the system that produced it. The Act requires the state scheme to carry a grievance redressal mechanism, and the framework built around the Act added the district ombudsman, an officer provided for under the scheme and the Union’s instructions, to receive complaints from workers and the public, to inquire into them and to require the programme authorities to act. The ombudsman’s position in the design is a frank admission written into the programme’s own documents. The officers who administer the guarantee are the first respondents to every complaint about it, and an appeal that runs only to the author of the act complained of is structurally weak. The ombudsman stands outside the executing line while remaining inside the programme, which is also why careful accounts of the field describe the office accurately as an executive graft on a statutory body rather than as a creature of the Act itself. Its authority rests on instructions. Its usefulness rests on the same quality the receipt has: it produces a written record, by an officer with no works of his own to defend, of what was demanded and what was done.

The grievance channel completes a pattern that runs through the whole statute, which is that the Act trusts paper more than it trusts assurances. The job card records demand, work and wages in the worker’s own custody. The dated receipt fixes the start of the fifteen day clock. The muster roll records attendance at the work and is read out at the social audit. The wage list and the fund transfer order record the payment process stage by stage, so that delay compensation can be calculated and culpability assigned. A worker pursuing a claim inside this system is, at each step, matching one document against another, and a worker whose documents are missing discovers that the claim has lost its handles. That is why the discipline of keeping the programme’s paper is not an administrative nicety, and why a reader assembling a household’s file for an allowance claim, a compensation claim or a grievance will keep the job card, every dated receipt, and the correspondence together in one place, for instance by working through the presentation of the household’s claim with a free online PowerPoint viewer for opening the training and claim slides in the browser before carrying the file to the panchayat office. The general account in this article is legal information about the statute’s design and does not substitute for advice on an individual claim, whose outcome will turn on the documents that claim can produce.

There is a lesson in the pattern for the wider series of which this article forms part. Again and again, across very different fields of Indian legislation, a statutory right has been found to depend on an administrative act that the duty holder controls: a certificate, an entry in a register, an acknowledgement, a receipt. The employment guarantee states the pattern in its purest form, because the controlling act, the acknowledgement of an application, is also the act that starts every clock the right depends on. Where that act is withheld, nothing else in the statute is even engaged. No allowance accrues, because the fifteen days cannot be shown to have run. No failure is recorded, because the statistics of the programme are built from applications received. No grievance has a foundation, because the grievant cannot prove the demand was made. The withholding leaves no trace, which distinguishes it from every cruder method of defeating a right. A refusal can be appealed. A silence cannot.

The courts: Swaraj Abhiyan and the enforceability of the clocks

The Act’s judicial history is dominated by proceedings that tested whether its periods and payments would be treated by a court as obligations or as targets. In Swaraj Abhiyan v Union of India, decided in 2016 in litigation arising out of drought across several states, the Supreme Court described the Act as social welfare and social justice legislation and directed that its provisions be faithfully implemented by all concerned. The directions repay attention because of their choice of targets. The Court directed the states to present realistic labour budgets and the Government of India to release adequate funds in a timely manner, fastening on the seam between the demand logic and the budget logic identified earlier in this article. It directed that compensation for delayed payment be made over to workers whose wages had been delayed beyond fifteen days, as paragraph 29 of Schedule II postulates, treating the compensation clause as a working entitlement and not as a statement of penal intent against defaulting offices. It directed the constitution of the Central Employment Guarantee Council under Section 10 within a fixed period, and pressed the states to establish their State Employment Guarantee Councils under Section 12 within a shorter one, on the reasoning that a statute’s monitoring authorities cannot review a programme from which their own existence is missing. And it recorded that the enhanced guarantee in drought affected areas, one hundred and fifty days where the additional fifty had been notified, formed part of the obligation to be funded and delivered.

In a later phase of the same litigation, a bench of the Court returned to the wage clock and stated the position in terms that admit no administrative gloss. A worker is entitled to payment of wages within a fortnight of the work being done, failing which the compensation follows, and the burden of compliance rests on the state governments and the Centre together, with neither permitted to pass the burden to the other. Delays, the Court said, are simply not acceptable, and bureaucratic red tape cannot be offered as an excuse for denying wages to workers who have expended unskilled manual labour. The reasoning connects the statute to the Constitution at the point where the Act’s own jurisprudence had always implied the connection ran: work extracted and left unpaid under economic compulsion implicates the prohibition of forced labour, so that a delayed wage under this Act is not merely a breached programme rule but a payment owed to a worker whose labour the state has already taken.

For the pillar’s purposes, three propositions can be taken from the judicial record without overstating it. A statutory entitlement of this kind is enforceable, and its timelines are justiciable facts rather than administrative aspirations. The compensation and allowance clauses are operative parts of the entitlement, which a court will direct governments to honour rather than treat as decorative. And the institutional provisions of the Act, the councils and the budget process among them, are themselves obligations whose neglect a court can remedy on a timetable. What the courts have not done, and what no judgment cited in this article should be read as doing, is to administer the programme themselves or to convert its every guideline into a decree. The litigation polices the Act’s own promises. It does not write new ones, and its discipline in that respect mirrors the discipline this article has applied to the performance record, which is owned elsewhere.

The amendment arc in brief

The statute’s text has been adjusted at its edges rather than rebuilt. The renaming of 2009, effected by amendment with effect from 2 October of that year, attached the name of Mahatma Gandhi to a guarantee whose intellectual ancestry in the dignity of manual labour made the association intelligible, and changed nothing else. Schedule II grew the compensation paragraph whose staged account of the payment process, measurement, computerisation, wage list generation and fund transfer, reflects hard administrative learning about where wage delay actually accumulates: not in a single act of refusal but in a relay of small delays, each too small to discipline and together amounting to months. The wage notification power in Section 6 has been exercised and revised over the years, with notified rates differentiated by state and revised by reference to movements in the prices agricultural labourers face, a development conducted entirely within the four corners of the section’s text. The permissible works in Schedule I have been enlarged by notification in the direction of the assets of individual households and of convergence with other rural programmes, again using the door the Schedule itself leaves open for works notified in consultation with the states. The arc is worth recording for what it shows about the drafting. The Act’s centre has not needed amendment, because its centre is a mechanism rather than a settlement of amounts, and mechanisms age more slowly than the figures written into statutes that fix them.

The labour budget: where the demand logic meets the budget logic

No section of the Act has attracted more misunderstanding than the planning process, because the planning process is where the statute’s two logics are forced to share a room. A demand driven entitlement cannot be planned in the manner of a road programme, by deciding in advance what will be built and then staffing it. Yet a programme funded through public money cannot be run without estimates, sanctions and a calendar. The Act’s answer is the labour budget: an advance estimation of the demand for work, prepared from the panchayat upward, aggregated through the block and the district, and considered at the Centre, against which funds are released and the shelf of works is sized. Prepared honestly, the labour budget is a forecast, revised as registered demand declares itself. Prepared as a ceiling, it becomes the instrument by which a demand driven statute is quietly reconverted into an allocation, because a panchayat told that its budget is exhausted has been given a reason, intelligible inside the administration, to stop registering the applications the statute commands it to accept.

The statutory text gives no authority for the second use. The duty to accept applications and to provide work within fifteen days is not expressed to be subject to the state of the labour budget, and the allowance that prices a failure to provide work is not expressed to be excused by it. The Supreme Court’s insistence, in the drought proceedings, on realistic labour budgets presented by states and pragmatic consideration by the Centre is best read against this background: a forecast honestly made and honestly funded is the mechanism that keeps the two logics from colliding at the worker’s expense, whereas a forecast trimmed to fit an appropriation merely relocates the collision to the panchayat counter where the application is handed in. The planning documents of the programme, the district plan, the shelf of projects and the perspective plan that panchayats prepare for their works, belong to the same analysis. A shelf of works is not bureaucratic decoration. It is the physical precondition of the fifteen day clock, because work cannot be allotted within fifteen days of an application unless a sanctioned work exists, or can be opened at once, within reach of the applicant’s village. A panchayat with an empty shelf will breach the first clock through no failure of willingness, which is why the Act loads the Gram Panchayat and the Gram Sabha with planning functions in advance of demand rather than permitting the system to improvise after demand arrives.

There is a discipline of timing inside the planning cycle that rewards attention. Demand for the guarantee peaks in the agricultural lean season and after failed rains, which is also when panchayat institutions are at their busiest with the consequences of the same distress. The application mechanism, with its advance applications and its fourteen day blocks, is the statute’s device for letting households declare demand ahead of the peak, so that works can be opened against declared demand rather than against distress already visible in migration. A system that waits for distress to announce itself in departures from the village has misunderstood its own statute, which provides the instruments for demand to be registered while the household is still present to register it.

A household’s year under the Act, worked through the clocks

Abstractions about entitlement become testable when they are run against a single household’s year. Take a household of two adult members in a notified rural district, registered with its Gram Panchayat and holding a job card in the adults’ own custody, with photographs affixed and entries blank at the year’s opening. In the lean weeks before the monsoon, one adult applies in writing to the Gram Panchayat for thirty days of work beginning on a named date a week ahead, and receives a dated receipt. The advance character of the application matters: the fifteen day period runs from the later of receipt and the date from which work is sought, so the administration holds the full period measured against the start the household actually wants. Work is allotted on a water conservation work within the village boundary, intimated in writing at the job card address and by public notice at the panchayat office. The first clock has been honoured, and the household’s card receives its first entries.

The work proceeds in blocks measured by the muster roll. At the closure of each muster, the payment process begins its staged journey: measurement of the work done, computerisation of the muster and of the measurements, generation of the wage list, and the fund transfer order that moves the wage toward the worker’s bank or post office account. The wage for the first block reaches the account inside the fortnight the Act allows. In the second block, the wage list waits at one stage of the process and the fortnight passes. From the sixteenth day after the muster closed, compensation at 0.05 per cent of the unpaid wages accrues for each day of delay, and the staged structure of paragraph 29 identifies the stage at which the delay occurred, which is also the stage whose functionary bears the culpability the compensation rules assign. The household does not have to prove that anyone intended the delay. The calendar and the muster closure date do the proving.

Later in the year the household applies again, in the slack after the harvest, and this time no work is allotted inside fifteen days. On the sixteenth day the allowance entitlement begins, at not less than a quarter of the wage rate, payable by the state government. The Programme Officer must record in writing, for the District Programme Coordinator, why the work could not be provided, and the household’s claim to the allowance rests on the receipt from the second application and on the absence of any direction to report for work. If work is then directed on the twentieth day and a member reports, the allowance liability ends under Section 7, the wages begin, and both streams count toward the household’s hundred day aggregate and toward the combined ceiling at which wages and allowance together equal the wage value of one hundred days. Nothing in the sequence required the household to persuade anyone of its poverty, to obtain a certificate of need, or to wait for a sanction in its favour. It required the household to perform two written acts and to keep two pieces of paper. The entire jurisprudence of the guarantee is contained in that ratio between what the citizen must do and what the citizen must keep.

Run the same year with one variation, and the design’s weakness appears in its exact location. In the variation, the second application is made orally at the panchayat office, is received with a verbal assurance, and is never entered or receipted. No work follows. No allowance follows either, and when the household complains, the record shows no application, no running clock and no breach. The programme’s own statistics for the village will report demand fully met, because the demand that statistics count is the demand that was recorded. The household has been refused nothing, in the only sense the system can perceive. That is the failure mode the brief for this article names as the guarantee’s central vulnerability, and it is not a failure of funding, of works or of wages. It is a failure at the single point where the citizen’s act must be converted into the state’s record, a conversion the duty holder itself performs.

Measurement, piece rates and the wage that work actually earns

The notified wage rate is a daily rate, but the wage a worker takes home under the programme is mediated by measurement, and a pillar account of the statute that stopped at the rate would misdescribe the entitlement. Work under the programme is organised by tasks, and the earnings of a worker or a group are assessed against the schedule of rates applicable to the work, which translates measured output, earth excavated, length of channel dug, area treated, into wages. The design inherits the older public works tradition in which payment by measured output protected the wage bill against payment for attendance alone. It also creates the programme’s most persistent tension at the level of the individual worker, because a daily rate filtered through a task norm yields different earnings for workers of different strength, health and experience on the same muster, and an older worker, a worker in poor health or a worker on hard soil may complete a day of genuine labour and be measured into earnings below the daily rate the public discussion treats as guaranteed.

The statute’s own text frames the answer at the level of principle rather than of centimetres. The entitlement Section 3 creates is to wage employment at the wage rate, and Schedule II’s conditions assume a working day and a working week as the units in which employment is provided and recorded, with muster rolls that record attendance day by day. The measurement system is an administrative instrument inside that frame, legitimate as a check on the wage bill and corrosive only where its norms are set so that a full day of ordinary labour cannot reach the daily rate. Setting the norms is technical work, done through time and motion study of the tasks concerned, and its quality decides whether the notified rate is a fact or a fiction for the median worker. A reader assessing any account of the programme’s wages should therefore ask two questions rather than one: what rate has been notified under Section 6, and what does a measured day of work in fact earn against that rate under the schedule of rates in force. The first question is legal. The second is the same question continued by other means.

The social audit in depth: the community reading its own record

Section 17 deserves a slower treatment than its single section usually receives, because the social audit is the point at which the Act’s theory of accountability becomes a practice. An ordinary audit, of the kind the Republic’s auditor performs, examines the records of an administration from outside and reports to a legislature. A social audit examines the same records from inside the community the records describe, in a public sitting of the Gram Sabha, with the workers whose names appear on the muster rolls present to confirm or deny what the rolls assert. The records are read out. The work is physically verified against the ground it claims to have changed. Payments recorded are matched against payments received, by asking the persons recorded as having received them. The findings are recorded in the Sabha’s proceedings and transmitted into the programme’s administrative record, where the duty to act on them rests on the Programme Officer and the district authorities.

The mechanism answers a specific epistemic problem that every other instrument in the programme handles badly. A muster roll bearing the name of a person who never worked, the classic fraud of public works, is invisible to a financial audit, because the expenditure it records is internally consistent: the money was drawn against a name, a thumb impression and a measurement. It is visible at once to the village, because the village knows who worked. The social audit is the statutory device for admitting the village’s knowledge into the official record with legal consequences attached, and its placement in the Gram Sabha rather than in an inspectorate is a judgment about where the relevant knowledge lives. The audit framework built under the Act, including the auditing standards framed with the Comptroller and Auditor General and the social audit units states maintain to support the process, exists to give that judgment an organised form, so that the Sabha’s verification is documented in a manner the administration above it cannot wave away as anecdote.

Two misreadings of the audit should be corrected with the text in hand. The audit is not a public hearing bolted onto the programme by activist pressure after the event; it is in the Act of 2005 itself, in Section 17, as one of the Gram Sabha’s statutory functions in relation to the works in its own panchayat. And the audit’s object is not expenditure in the abstract. It is all the works executed within the Gram Panchayat, which includes the decisions that precede expenditure: whether applications were received and receipted, whether work was allotted within the period, whether the allowance was paid where allotment failed. A social audit conducted at full statutory width is therefore the one forum in the system where the invisible failure, the unrecorded application, can be surfaced, because the Sabha can compare the register of applications against the village’s own memory of who asked for work. That comparison is available nowhere else in the architecture, which is the strongest textual reason for treating the audit as a load bearing provision rather than a transparency ornament.

Women inside the design: one third, equal wages and the creche

The programme’s treatment of women is often summarised as a reservation, which understates what the text does. Schedule II requires that priority be given in the allocation of work so that at least one third of the beneficiaries to whom employment is provided are women who have registered and requested work. The provision operates as a floor on the composition of the workforce the programme actually fields, and in a demand driven design its function is subtle: it disciplines the allotment stage, so that where applications exceed the work immediately available, the pattern of allotment cannot reproduce the labour market’s ordinary exclusion of women from public works employment. Around that floor the Schedule builds the equal wage rule, under which the notified rate is paid without discrimination by gender, and the creche rule, under which the presence of five or more children below six at a worksite obliges the deputation of a woman worker, paid at the wage rate, to care for them. The worksite facilities of drinking water, rest and first aid apply to all, and the accident and medical provisions of the Schedule protect the worker in the course of employment.

The conjunction of these provisions is the design working as a whole. A programme that fixed a one third floor while paying women less than men would have reproduced the market inside the guarantee. A programme that paid equal wages while making no provision for children would have admitted the unencumbered and excluded the mothers whose need is typically greatest. A programme that did both while leaving allotment wholly to local discretion would have watched the floor dissolve in practice. The Schedule addresses all three joints, and the result, in statutory terms, is that the rural woman worker is the person this Act imagines most concretely, with her wage, her children and her share of allotments each the subject of an express provision. Analysts of the programme’s social effects have argued for years about how far the text’s intention has been realised in earnings, in control over the wage once paid and in the character of the works allotted. Those are record questions, and they belong to the evidence articles. The pillar’s narrower point is that the text gives those inquiries something precise to measure realisation against, which is more than the programme guidelines of the predecessor schemes ever provided.

The federal dimension: one Act, many state schemes

The guarantee is national in its floor and state level in its operation, and the seam between the two is a deliberate allocation of competences that the Act’s text maps carefully. Parliament enacted the entitlement, the clocks, the funding pattern and the Schedules. Each state government frames the scheme that gives effect to the guarantee within its territory, appoints the officers, constitutes its council, specifies the allowance rate above the statutory floor, and administers every stage from registration to payment. The Central Government notifies the wage rate, bears the wage bill and the larger share of material cost, issues the framework instructions that harmonise administration, and answers for the programme’s national architecture. Either level, acting alone, can break the guarantee for the worker at the end of the chain: the Centre by starving or delaying the wage funds, the state by failing to register demand, open works or pay the allowance. Neither level can deliver the guarantee without the other, which is the federal form of the proposition the Supreme Court stated in the wage litigation when it refused to let either government pass the burden of compliance to the other.

State variation inside the frame is real and lawful within limits. The scheme a state frames may organise registration, application and payment through its own administrative arrangements, may set the allowance above the floor, and may add its own instruments of transparency and grievance handling around the Act’s skeleton. What a state scheme may not do is subtract from the Act: it cannot shorten the hundred days, lengthen the fifteen, lower the allowance floors, narrow Schedule I, or dissolve Schedule II’s conditions, because the scheme is made under the Act and derives its authority from it. Disputes at this seam are therefore disputes about the location of a provision in the hierarchy of norms, and they are resolved by asking the pillar’s habitual question: where does the rule live, in the statute, in a Schedule, in the state scheme, or in a guideline, and what is it permitted to override from that position. Half the public confusion about the programme consists of guidelines being reported as law and statutory entitlements being reported as if they were guidelines, and the confusion is consequential in both directions, since it inflates expectations the first way and deflates enforceable rights the second.

Employment, it should be recorded, is a field in which the states carry long independent experience, and the national Act did not write on a blank slate. Maharashtra’s statutory guarantee preceded it by nearly three decades. Other states had run their own employment and relief works under their own arrangements, and the administrative cadres that operate the national programme were formed inside those state traditions. The Act’s achievement was not to invent rural public works but to fix a common legal floor beneath them, with common clocks and a common funding bargain, while leaving the machinery in the hands of the governments that would have to work it. Judged against the series’ wider map of welfare legislation, that allocation is the standard Indian settlement for a Concurrent field: a Union statute carrying the entitlement, state schemes carrying the administration, and the courts policing the floor.

The honest assessment: the case for demand, the case against its record keepers

The design debate about this statute is genuine, and both sides of it are entitled to their strongest statement. The case for the demand driven design is a case about information and about dignity. An allocation based programme must guess where distress will fall, village by village and month by month, and its guesses are made a year in advance by officials who cannot see the season coming. Demand registers distress where and when it occurs, declared by the only persons who know its timing exactly, and it scales the programme automatically: a good monsoon year draws fewer applications, a failed one draws more, without any official having to declare an emergency or revise a target. The design also reverses the posture of the citizen. A beneficiary of an allocation waits to be chosen and is grateful to be chosen. An applicant under the Act demands what the law owes, on a form, against a receipt, and the difference between those postures is the difference between a welfare programme and a right, which is not a small difference to the person inside it. Supporters of the design add that self selection through unskilled manual work at a notified wage targets the programme without any means test bureaucracy: those who have better options take them, and the cost of identifying the poor, with all the exclusion errors identification systems generate, is designed out.

The case against is not a case against the poor having work. It is a case about who keeps the gate. Demand under the Act is registered by the same authorities who must meet it and whose performance is judged by their success in meeting it, an arrangement that pays the gatekeeper, in reputation and in quiet, for every application that never enters the register. Fiscal critics add a distinct objection: an open ended entitlement funded predominantly by the Centre sits uneasily inside annual budgeting, and the practical result, visible in the programme’s administrative history, is rationing by the release of funds, which reproduces allocation under a demand label and does so without the honesty of an allocation announced as such. Administrative critics point to capacity: the Act loads its most demanding functions, registration, planning, measurement, audit support and payment discipline, onto the tier of government least staffed to perform them, and a design that is elegant at the level of rights may be unkind at the level of the panchayat secretary’s working week. Each of these criticisms identifies something real. None of them identifies a defect that returning to allocation would cure, because allocation suffered from every one of the same diseases, capacity, fiscal pressure and local capture, while adding the incurable one that nobody under it held a claim.

The verdict this pillar defends follows from the text rather than from the programme’s reputation. As a piece of statutory design, the Act is the most carefully engineered welfare statute of its generation in India, and its engineering is concentrated at exactly the joints where welfare programmes fail: the trigger, the deadlines, the price of breach, the funding incentives, the audit by the governed, and the custody of the record. Its central weakness is inscribed in the same design and cannot be praised away: the guarantee runs on documents the duty holder issues, so its enforceability is hostage to the integrity of its first administrative act. That weakness is a reason to defend the receipt, the register and the social audit with more energy, not a reason to prefer a design in which the worker holds no documents at all. The detailed evidence of how the guarantee has in fact been met, evaded, funded and delayed, and the district level record on which any final judgment of the programme must rest, belongs to the article on the MGNREGA implementation record, which owns that evidence in this series and to which the reader who wants performance rather than design should turn.

The namable claim: the application is the guarantee

Every entitlement in this statute is downstream of one act. Work is owed because an application was made. The fifteen days run because an application was received. The allowance accrues because the application was not answered. Compensation for delayed wages presupposes a worker inside the system the application entered. The register of demand, from which labour budgets are built and against which performance is reported, is an aggregate of applications and of nothing else. The claim this article leaves with the reader can therefore be stated in one sentence: the application is the guarantee, and the dated receipt is the single most important document in the entire scheme, because it is the one document that proves the guarantee was ever engaged. A system that issues receipts faithfully has made its own failures visible and therefore actionable, which is the most that law can do for an entitlement of this scale. A system that does not has converted a legal right into a discretionary programme by the simple method of declining to record the demand, and it has done so without refusing a single person anything that any record will ever show.

That claim reorganises how every other question about the statute should be asked. Debates about funding adequacy, about wage rates, about the quality of assets and the depth of audit all concern stages of a machine whose intake is the application. Where the intake is honest, those debates are the ordinary politics of a large programme, to be settled with evidence. Where the intake is not, the debates are theatre performed over a register that does not contain the demand it claims to describe. The reader who assists workers, the official who administers a block, and the researcher who studies the programme are therefore pointed at the same first question, from their three different directions: when a household in this village asks for work, what paper changes hands, and can the household prove, a month later, that it asked. The statute’s answer is written in Schedule II, in the obligation to accept the application and to issue the dated receipt. Everything else in this long Act is commentary on that obligation and machinery for its consequences.

Who counts: household, adult and the edges of eligibility

The entitlement’s unit is the household, and the Act’s definitions do quiet but decisive work at its edges. An adult, for the purposes of the Act, is a person who has completed eighteen years of age, which ties the guarantee to the population for whom manual wage employment is lawful ordinary work and separates the programme from the schooling obligations that protect younger persons. A household is defined through the family as it lives and eats together in the village, and the definition reaches the single person living alone as a household of one, the separated or deserted member maintaining a separate hearth, and the group of relatives living under a common roof and sharing a common kitchen, so that the unit of entitlement follows the reality of rural living arrangements rather than the revenue record’s idea of a family. The drafting choice matters in predictable places. A widow living alone is a household and holds the full hundred days in her own right, not a fraction of a joint family’s entitlement controlled by others. Adult children who have separated their hearth from the parental household form their own unit and register separately. A household that splits during the year presents the register with the administrative problem the definition anticipates, of entitlement following residence and commensality rather than surname.

Eligibility is otherwise deliberately bare. There is no income test, no land ceiling and no requirement to prove poverty, because the work itself, unskilled manual labour at the notified wage, is the test the design applies, and a household that does not need the wage at the cost of that labour will not apply for it. Registration does require residence in the rural area of the panchayat concerned, since the guarantee is territorial and the machinery that must answer the application is local. The contrast with the certificate based design of the disability legislation examined elsewhere in this series is instructive. Under that statute, the gate stands in a medical authority and the argument is about a percentage. Under this one, the gate stands in the panchayat register and the argument is about a receipt. Each design places its point of maximum contest where its theory of the beneficiary says the risk of wrongful exclusion lies, and the employment guarantee’s theory, that the poor are excluded by procedure more often than by definition, is written into its choice of gate.

The constitutional setting: a directive the legislature chose to enact

The right to work enters the Constitution as a directive principle, in the article that directs the state, within the limits of its economic capacity and development, to make effective provision for securing the right to work, to education and to public assistance in cases of unemployment, old age, sickness and disablement. Directive principles are not enforceable by any court, and for half a century the right to work in India lived in that unenforceable register, available as rhetoric to manifestos and as an interpretive aid to judges, owed by no one in particular. Alongside it, the courts developed the right to livelihood inside the right to life, holding in the pavement dwellers’ litigation of the 1980s that a person deprived of livelihood is deprived of life in the sense the Constitution protects, and the prohibition of forced labour acquired, through the minimum wage cases, a reach that covered labour exacted by economic compulsion at less than the statutory minimum. The constitutional materials for an enforceable employment right therefore existed in fragments: a directive naming the object, a fundamental right protecting livelihood against arbitrary deprivation, and a prohibition policing the terms on which labour may be taken.

What the fragments could not do was create the entitlement itself, with a quantum, a procedure and a budget. Only a legislature could fix one hundred days rather than fifty or two hundred, could choose the household rather than the individual as the unit, could set fifteen days rather than thirty as the period for providing work, and could divide the bill between the Centre and the states. The Act of 2005 is the directive principle of the right to work translated by Parliament into a statute, and the translation is the event the constitutional lawyer should fasten on, because it changed the character of every subsequent argument. After 2005, a worker denied the guarantee does not ask a court to invent an obligation from a directive and a value. The worker pleads a section. The litigation examined in this article shows the difference in practice: the courts enforced labour budgets, councils, compensation and timelines, all of which are statutory artefacts, and the Constitution supplied the background principles, livelihood and freedom from forced labour, in whose light those artefacts were read. The lesson generalises across the welfare field. Statutes are how directive principles acquire clocks, and clocks are how rights acquire remedies that do not depend on the continuing attention of a constitutional bench.

The assets question: whether the works justify the wage bill twice over

A programme that pays wages for work invites a double audit, and the second audit, of the works themselves, has generated its own controversy which a pillar must state fairly. The sympathetic account begins with Schedule I. The permissible works are weighted toward water: conservation, harvesting, renovation of tanks, drought proofing, irrigation and drainage, with land development and rural connectivity around them. In the semi arid districts where the guarantee began, water is the binding constraint on the agricultural economy from which the workers come, and a wage bill that deepens the water table and desilts the tank is an investment whose return appears in the same households’ other income, in a second crop, in wells that survive the summer, in livestock kept alive. The individual beneficiary works named in the Schedule extend the logic to private land of the named households, where a dug well or a levelled field converts one season’s wages into a durable change in a family’s productive position. On this account the programme is not a transfer with a work requirement attached. It is a rural capital programme that pays its builders a guaranteed wage and lets them, as residents, inherit the capital.

The critical account answers that the audit of assets tells a less comfortable story in many districts: works abandoned half finished when the season turned, assets recorded as complete that verification found incomplete, expenditure concentrated in the material share in ways the ratio was meant to prevent, and a shelf of works chosen for ease of measurement rather than for value to the village. The criticism has force wherever it is documented, and the design contains its own reply, which is not a denial but an allocation of responsibility. The choice of works belongs to the Gram Sabha, the execution substantially to the Gram Panchayat, the verification to the social audit and the technical sanction to the programme’s engineering line, so an asset failure under this Act is traceable to named institutions in a way that the failures of the old departmental programmes were not. The honest pillar verdict is that the statute builds the best available machinery for making the works matter and cannot, by text alone, make a village choose its tank over its road or finish in one season what its labour budget sized for two. Schedule I settles what may be built. It deliberately leaves what should be built to the assembly of the people who will live with it, and that deference is either the design’s wisdom or its evasion depending on the capacity of the assembly in question, which is a fact that varies by village and cannot be legislated from Delhi.

Reading the programme’s numbers without being read by them

Because the implementation record is owned elsewhere in this series, the pillar’s duty toward the programme’s statistics is limited to teaching the reader how any figure in this field must be interrogated before it is believed, whoever publishes it. Three questions discipline the reading. The first concerns the denominator of demand: person-days generated, the statistic headlines prefer, measures what the system delivered and is silent about what was asked, so a year of rising person-days may record rising distress, rising administrative capacity or rising registration, and only the applications register can separate the three. The second concerns the wage: a notified rate is a statement of law, an average earning per day is a statement about measurement and attendance together, and the distance between them is the schedule of rates and the muster roll, not necessarily anyone’s dishonesty. The third concerns delay: the payment statistics count delay to the stages the system defines, and the definition of the endpoint, the generation of a fund transfer order as against the credit reaching a worker’s account in a village served by a distant bank, decides what the figures can see. The Supreme Court’s staged analysis in paragraph 29 of Schedule II, assigning each segment of the payment journey to a functionary, is the statute’s own instruction in how to read its numbers: by stage, against a clock, with a name attached to the segment that failed. A reader who applies those three questions to any annual account of the programme, favourable or hostile, will extract from it roughly what it contains and no more, which is the skill the evidence articles in this series then put to work.

The Maharashtra lineage and what the national Act learned from it

The state experiment that stands behind the national statute deserves more than the passing acknowledgement it usually receives, because the national text is legible as a set of answers to problems Maharashtra’s guarantee had already exposed. Maharashtra built its employment guarantee out of the drought years of the early 1970s, first as a programme and then, in the Maharashtra Employment Guarantee Act of 1977, as a statute, funded in part by taxes levied for the purpose on professions, trades and employments, so that the guarantee carried its own revenue line rather than depending wholly on annual generosity. Three lessons travelled from that experiment into the Act of 2005. The first was that a statutory guarantee is administrable: registration, application, works opened on demand and wages paid at scale, across decades and changes of government, in a state of ordinary administrative capacity. Every later claim that a national guarantee was a romantic impossibility had to argue against a functioning counterexample two states away. The second lesson concerned funding honesty. A guarantee financed by a visible, dedicated stream sets the argument about its cost in the open, and the national Act’s answer, a statutory funding pattern dividing the bill between the Centre and the states, is the federal translation of the same insistence that the money for a right be arranged by law and not by hope. The third lesson was about the guarantee’s effect on the rural labour market itself. A standing offer of work at a known wage alters the bargaining position of every agricultural labourer in the district, including those who never appear on a muster roll, because the employer who will not match the floor the guarantee sets must watch the workforce exercise the option the statute gives it. Economists describe this as the reservation wage effect, and it is one of the guarantee’s largest consequences, achieved by workers who, in the programme’s own statistics, do not exist.

The national Act also departed from the state model where scale required it. Its entitlement is framed by household and financial year with a precision the earlier statute’s practice had to work out administratively. Its clocks are tighter and its compensation clauses more elaborated, reflecting the national debate’s long study of the points at which the Maharashtra system strained, in delayed payments, in works that existed on paper and in the sheer administrative weight of a guarantee carried by a single state’s machinery. Reading the two statutes together is the fastest route to understanding what the drafters of 2005 thought they were doing: not inventing a right, but nationalising one, and hardening its weak joints with the evidence of twenty eight years.

Transparency as a working part: muster rolls, records and public information

The Act’s transparency provisions are sometimes described as governance garnish applied to a welfare programme. They are better understood as part of the entitlement’s machinery, because every claim the worker holds is proved from the programme’s records, and the worker’s access to those records is therefore a component of the right itself. The muster roll is the foundational document: it records who was allotted work, who attended, what was measured and what wages were calculated, and it is maintained at the worksite and open to scrutiny rather than filed in an office the worker cannot enter. The job card carries the household’s copy of the same history home. The wage list, the measurement record and the payment order extend the chain from the worksite to the account, and the social audit reads the whole chain aloud in the village. Around these paper instruments the programme’s information systems were later built, digitising the registers and publishing the programme’s records for public inspection, so that a worker’s claim, a researcher’s dataset and an auditor’s sample are all drawn from the same underlying record the panchayat created. The statutory significance of the digitisation should not be overstated, since the Act’s obligations were complete before it, and should not be understated either, since a record that can be inspected from outside the office that made it is a record whose custodian knows it will be.

It is in this setting that the compensation paragraph’s stage wise division of the payment process shows its full purpose. By requiring the process to be broken into measurement, computerisation of musters and measurements, generation of wage lists and uploading of transfer orders, with maximum periods specified for each stage and the responsible functionary identified, the Schedule converts a grievance about delay from a complaint against the system in general into a question with an address: at which stage, in whose custody, for how many days beyond the stage limit, did this wage wait. Accountability in a programme of this size is impossible at the level of outcomes alone, because outcomes have too many authors. It becomes possible at the level of stages, and the statute, unusually for welfare legislation of its period, contains the analytical tool of its own supervision inside its text.

Seasonality, migration and the guarantee as insurance

The guarantee’s deepest economic function is insurance, and its text is shaped by the risks it insures against. Rural household income in the districts where the Act began is hostage to the monsoon in a way that salaried India rarely experiences: a failed rain is simultaneously a crop failure, an employment failure and a food price event, arriving together in the same season. Private insurance against that compound risk barely exists for the households concerned, and credit against it is priced by moneylenders at rates that convert one bad season into a decade of debt. A statutory offer of one hundred days of work, expandable by notification in drought, is the state’s assumption of a share of that risk, and its insurance character explains features of the design that puzzle readers who approach it as ordinary employment policy. The premium is zero because the insured are those least able to pay one. The claim is made by presenting oneself for labour, which is the form the deductible takes in a work based design. The payout is wages, deliberately unglamorous, and the insurer’s obligation is triggered by the insured event’s personal form, the household’s decision that it needs the work, rather than by any official declaration that a drought exists. The additional fifty days available on notification in calamity are the policy’s catastrophe clause, and the Supreme Court’s drought proceedings, which treated the enhanced entitlement as obligatory in the notified districts, read the clause exactly as an insurance term would be read: against the drafter who delayed honouring it.

Migration is the household’s older insurance, and the guarantee’s relationship to it is more respectful than the rhetoric of stopping migration suggests. The Act does not forbid, discourage or penalise movement for work, and a registered household that migrates loses nothing but the days it does not demand. What the guarantee changes is the terms on which migration is chosen. A household with a statutory fallback at home migrates toward better wages because they are better, rather than toward any wages because the alternative is hunger, and it can time its return against the agricultural calendar and the advance application mechanism the Schedule provides. The programme’s wage, its local availability and its five kilometre rule all belong to this analysis: they set the value of staying, and the worker’s freedom consists in the ability to compare that value, guaranteed by law, against the uncertain offers of the road. A statute that began as famine relief machinery ends, in this reading, as an instrument of bargaining power, and that is the reading its authors, who came out of the movements that organised rural labour, would have recognised as their own.

What the Act does not do

A pillar earns trust by marking the boundary of its statute as carefully as its content, and this Act’s silences are as instructive as its provisions. It does not guarantee work to any individual as such: the hundred days belong to the household, and two adults of one household share one entitlement, a choice that spreads a limited fiscal commitment across the rural population by co-residence and that households experience as a ration whenever more than one member needs the work in the same season. It does not apply in urban areas, however urban in character a large panchayat may have become, because coverage follows the notification of rural areas and the urban poor hold no claim under it. It does not guarantee employment of any particular kind beyond unskilled manual work on the Schedule I works: the engineer, the teacher and the clerk of the village hold the same entitlement as the labourer, to the same ditch and the same wage, which is the self selection principle operating exactly as designed and the reason the programme cannot be criticised coherently for failing to provide skilled employment it never promised. It does not set its own wage at the agricultural minimum, for the Section 6 reasons already weighed, and it does not index its sixty rupee statutory floor, which stands in the text as a monument to the price level of 2005 while the operative rates live in notifications.

The Act also does not appropriate its own money. Its funding pattern binds the two governments in law, but the annual supply of funds passes through the ordinary budget processes of Parliament and the state legislatures, and the distance between a statutory share and a released tranche is the distance the courts were asked to police in the drought litigation. It does not create a specialised tribunal for the worker’s claims: the grievance route runs through the programme’s own officers, the ombudsman of the scheme framework and, beyond them, the ordinary constitutional remedies of the writ courts, which is a ladder whose top rung is powerful and whose bottom rungs are the very offices complained against. It does not penalise the failure to provide work with any consequence beyond the allowance itself and the reporting duty it triggers, so the sanction for defeating the guarantee is priced in money rather than in discipline. And it does not protect its own register against the quietest of its enemies, neglect: a household wrongly deleted, a card wrongly withheld, an application wrongly unrecorded has remedies in the grievance channel and the audit, but no damages claim of the kind a private statute of obligations might have supplied. These silences define the space in which the programme’s politics occur, and a reader who knows them will not mistake a grievance that belongs to the budget process for a breach that belongs to a court, or a breach that belongs to a court for a design choice that belongs to Parliament.

Accountability beyond compensation: offences, recovery and the audit trail

The Act’s disciplinary architecture runs on money recovered and records corrected more than on prosecution, but it is not toothless. Contravention of the Act’s provisions attracts the penal clause the statute itself carries, under which a person contravening the Act is punishable on conviction with a fine that may extend to one thousand rupees, a figure whose modesty states plainly that Parliament expected enforcement to travel by other roads. The operative roads are administrative and financial. Amounts misappropriated in the programme are recoverable through the audit and inspection machinery, the social audit’s findings feed disciplinary proceedings against the officials whose records failed verification, and the staged payment framework assigns culpability for delay to identified functionaries, so that compensation paid to workers can be traced to the desk where the delay occurred and visited on it through the state’s own service discipline. The Gram Sabha’s audit findings, the ombudsman’s inquiries and the programme’s inspections converge on the same design principle observed throughout this article: consequences under this statute attach where a document can be produced, a stage can be timed or a name can be matched against a roll.

For the worker, the practical hierarchy of remedies follows the same principle. The allowance and the compensation are automatic in design, calculable from dates and amounts without discretion, and are therefore the first remedies to pursue. The grievance channel and the ombudsman address the failures that are not automatic: the unrecorded application, the unopened work, the muster that does not match the labour performed. The social audit addresses the collective failures, the pattern across a panchayat’s works in a season. The writ jurisdiction stands behind all three for the case where the machinery itself will not move, and the litigation history shows the constitutional courts using that jurisdiction at the level of systems, budgets, councils and timelines, rather than case by case for individual musters. A worker’s adviser reading the statute as a whole will notice that each remedy is strongest exactly where the worker’s paper is strongest, which returns the analysis, for the last time, to the receipt. The Act gives with one hand a complete system of timed entitlements and takes nothing away with the other, but it delivers the system into the custody of the record, and those who keep the record keep the right.

Three distinctions that organise the whole statute

Readers who carry away three distinctions will misread very little of what is written about this programme. The first is between the entitlement and the scheme. The entitlement lives in the Act: one hundred days, the fifteen day duty, the allowance floors, the funding pattern and the Schedules. The scheme is the state government’s instrument under Section 4 for giving effect to the entitlement, and it organises registration centres, application forms, payment channels and administrative detail. Confusing the two produces the two standard errors of the field: treating a scheme condition, a form demanded or a channel prescribed, as if it could qualify the entitlement itself, and treating an entitlement, the allowance for example, as if it were a discretionary feature of a scheme that a state might omit. The hierarchy runs one way. The scheme that contradicts the Act fails, and the entitlement that the scheme omits survives its omission.

The second distinction is between demand registered and demand felt. The statute can only answer the first, because a legal system acts on acts it can prove, and the application is the form in which felt demand becomes registered demand. Public argument slides constantly between the two: a district’s low person-days are read as low need, or a village’s distress is read as proof that applications must exist in proportion. Both readings skip the register, which is the only place the statute lives. The programme’s integrity problem, its reform problem and its research problem are all, at root, the problem of the distance between the two kinds of demand and of the institutions that stand at the point of conversion between them.

The third distinction is between the wage as a price and the wage as a right’s measure. As a price, the notified rate is an instrument of labour market and fiscal policy, to be argued about with the tools of economics. As the measure of the entitlement, the same figure is the unit in which the allowance floors are fractions, the distance addition is a percentage and the delay compensation accrues, so that a change in the notified rate silently revalues every other promise in the statute. Debates that treat the rate as merely a price miss its constitutional position inside the Act’s arithmetic, and debates that treat it as beyond argument because it measures a right miss its character as a notification revisable by the executive that issues it. The statute holds both characters in one number, and the Section 6 controversy, examined at length earlier in this article, is what happens when a polity argues about the number in one character while its opponents answer in the other.

Why the guarantee endures: the politics of a statutory floor

The durability of the Act, across changes of government and repeated fiscal argument, is itself evidence about the difference between a statute and a scheme, and a pillar should record the mechanism rather than treat survival as luck. A scheme can be ended by the executive that created it, in a budget speech, without a vote on its merits and without any beneficiary holding a document that says otherwise. A statutory entitlement can be ended only by Parliament, in public, by a repeal that must name what it is taking away and from whom. Between those two procedures lies an enormous difference in political cost, and governments of every complexion have responded to it rationally: the guarantee has been argued about, funded generously in some years and tightly in others, administered with varying conviction and renamed once, but the entitlement itself has remained the fixed point around which the argument moves. The funding argument, which is where the contest actually occurs, is the safety valve the design provides: a government that cannot repeal the right can still ration its exercise through appropriations, labour budgets and release schedules, and a citizenry that cannot prevent the rationing can still measure it against a text and call the distance by its legal name, a breach. That is not a small residue. It is the whole difference between a promise and a law, stated in institutional terms.

The entitlement also manufactures its own constituency in a form schemes cannot replicate. Every registered household holds a card, every applicant a receipt, every worker a wage record, and those documents give millions of rural families a specific, personal and provable relationship with the statute that no speech about welfare creates. A programme’s beneficiaries are a statistic. This Act’s beneficiaries are a register, and a register can be organised, consulted and mobilised in a way a statistic cannot. Political scientists who study the programme’s survival point to exactly this conversion of beneficiaries into rights holders as the source of its unusual protection, and the conversion is not a side effect of the drafting. It is the drafting, working as its authors, who emerged from the movements that organised rural labour, intended it to work. The pillar’s final analytical point is therefore the same as its first, seen from the opposite end. The Act reversed the direction of public works, from allocation downward to demand upward, and the reversal, once enacted, created the political facts that have kept the reversal in force.

The pillar’s place in the cluster

A statute pillar in this series is a map of the territory its specialist articles then occupy in depth, and the employment guarantee cluster is arranged so that each large question has exactly one home. The pillar owns the text: the guarantee, the sequence, the clocks, the money and the machinery, stated once and stated completely. The implementation record owns the evidence of delivery, the figures and the failures, interrogated with the reading discipline this article has taught. The history of the demand that produced the Act, the campaigns and the drafting, belongs to the legislative record of the period rather than to a reading of the finished text. The wage question opens outward into the older minimum wage statute, whose fixation machinery predates the guarantee by more than half a century and whose agricultural rates supplied the guarantee’s first defaults. The panchayat institutions through which the Act works open upward into the constitutional architecture of local government. A reader who begins here and follows those doors in order acquires the statute the way its own Schedules acquire a worker: by registration first, and then by stages, each stage documented, until the whole record can be produced on demand.

Frequently Asked Questions

Q: What does MGNREGA guarantee?

MGNREGA guarantees that the state government will provide not less than one hundred days of wage employment in a financial year to every rural household whose adult members volunteer to do unskilled manual work, in rural areas notified under the Act. The guarantee is demand driven: it is triggered by a written application for work rather than by an allocation of funds to a village. If work is not provided within fifteen days, an unemployment allowance becomes payable, and fifty additional days may be notified in areas affected by drought or natural calamity.

Q: Who is entitled to MGNREGA work and how do they apply?

Every adult member of a rural household in a notified area is entitled, where adult means a person who has completed eighteen years of age and the hundred days belong to the household as a unit. The household first registers with the Gram Panchayat and receives a job card. An adult member then applies in writing to the Gram Panchayat or the Programme Officer, stating the period and duration for which work is sought, seeking at least fourteen days of continuous work, and is owed a dated receipt for the application.

Q: What happens if MGNREGA work is not provided within fifteen days?

The applicant becomes entitled to a daily unemployment allowance under Section 7 of the Act, payable by the state government. The rate is specified by the state in consultation with the State Employment Guarantee Council and may not be less than one fourth of the wage rate for the first thirty days of the financial year and not less than one half of the wage rate for the remaining period. The liability ends when work is directed, the period sought expires, or the household reaches its hundred day entitlement in wages, work or a combination.

Q: How is the MGNREGA wage rate fixed?

Section 6 of the Act empowers the Central Government, notwithstanding the Minimum Wages Act 1948, to specify the wage rate for the programme by notification, with different rates permitted for different areas and a statutory floor of sixty rupees per day. Until the Centre notifies a rate for an area, the minimum wage fixed by the state government for agricultural labourers under the Minimum Wages Act applies as the wage rate for that area. Notified rates are revised by notification over time, and equal wages are payable to men and women.

Q: What worksite facilities must MGNREGA provide?

Schedule II of the Act requires safe drinking water at the worksite, periods of rest and shade, and a first aid box with material adequate for emergency treatment of minor injuries and other health hazards connected with the work, with medical aid where a worker meets with an accident in the course of employment. Where five or more children below the age of six accompany women workers at a site, one woman worker must be deputed to look after the children and is paid wages at the applicable rate for that duty.

Q: What does a social audit under MGNREGA examine?

A social audit under Section 17 of the Act is conducted by the Gram Sabha and examines all works executed within the Gram Panchayat. It verifies the programme’s records against the ground: applications and receipts, job card entries, muster rolls and attendance, measurement of work, wage calculation and payment, and the physical existence and quality of the asset created. Workers and residents confirm or contest the record in a public sitting, and the findings enter the programme’s administrative record for action by the Programme Officer and the district authorities.

Q: Who bears the cost of MGNREGA?

Section 22 of the Act fixes the funding pattern. The Central Government bears the entire cost of wages for unskilled manual work and up to three fourths of the material cost, including wages of skilled and semi skilled workers, together with a determined share of administrative expenses. The state government bears the whole cost of the unemployment allowance, one fourth of the material cost including skilled and semi skilled wages, and the administrative expenses of the State Employment Guarantee Council. The pattern is statutory and can be altered only by Parliament.

Q: What role does the Gram Panchayat play in MGNREGA?

The Gram Panchayat is the programme’s working office at village level. It receives and verifies applications for registration, registers households and issues job cards. It receives applications for work, issues the dated receipts from which the fifteen day period runs, and allots work within that period. It identifies and plans works, building the shelf of projects the Gram Sabha recommends, and it executes at least one half of the works in terms of cost, which the Programme Officer is required to allot to Gram Panchayats for implementation.

Q: How does a household register and obtain a job card under MGNREGA?

The adult members of a rural household apply for registration to the Gram Panchayat of their village, giving their names, ages and the particulars the scheme requires. The Gram Panchayat verifies the application and registers the household, then issues a job card bearing the photographs of the adult members, free of cost. The card records employment demanded, employment provided and wages paid, and it is meant to remain in the custody of the household rather than of any official, so that the family’s own evidence of its history inside the programme stays in its hands.

Q: Why does the dated receipt for a work application matter under MGNREGA?

Every clock in the Act runs from the application the receipt acknowledges. The fifteen day period for providing work is measured from receipt of the application or from the later date from which work is sought, the unemployment allowance accrues only when that period expires unanswered, and the programme’s register of demand is built from recorded applications. A dated receipt is therefore the document that proves the guarantee was engaged on a particular day. Without it, an unanswered demand leaves no trace in any record, and neither the allowance nor a grievance has a date to stand on.

Q: For how many days must an application for MGNREGA work be made?

Schedule II requires an application for work to seek a minimum block of fourteen days of continuous employment, and a period of employment is ordinarily at least fourteen days continuously, with not more than six days of work in a week. Applications may be made in advance, naming a future date from which work is wanted, and there is no limit on the number of days for which a person may apply across the year, subject to the household’s aggregate entitlement of one hundred days in the financial year.

Q: What extra wages are paid when MGNREGA work lies beyond five kilometres?

Employment is to be provided, as far as possible, within a radius of five kilometres of the village where the applicant resides at the time of applying. Where employment is provided outside that radius, it must still be provided within the block, and the workers must be paid ten per cent of the wage rate as extra wages to meet the additional transport and living expenses. The rule does not forbid distant work, since a permissible work cannot always be opened near every village on demand, but it prices the displacement the administration imposes.

Q: When must MGNREGA wages be paid after the work is done?

Section 3 of the Act requires wages to be disbursed on a weekly basis and in any event within a fortnight of the date on which the work was done. Schedule II frames the payment process around the closure of the muster roll for the work performed: measurement of work, computerisation of the muster roll and the measurements, generation of the wage list and uploading of the fund transfer order, with wages to reach the worker within fifteen days from the closure of the muster roll.

Q: What compensation follows a delay in MGNREGA wage payment?

Paragraph 29 of Schedule II entitles the wage seeker to compensation for delay at the rate of 0.05 per cent of the unpaid wages for each day of delay beyond the fifteenth day from the closure of the muster roll. The payment process is divided into stages, measurement, computerisation of muster rolls and measurements, generation of wage lists and uploading of fund transfer orders, with stage wise time limits and identified functionaries, so that responsibility for the delay can be fixed. Delay in paying the compensation itself is treated in the same manner as delay in paying wages.

Q: Which works are permissible under Schedule I of MGNREGA?

Schedule I lists water conservation and water harvesting; drought proofing, including afforestation and tree plantation; irrigation canals and micro and minor irrigation works; irrigation facilities on land held by the households the Schedule names; renovation of traditional water bodies, including desilting of tanks; land development; flood control and protection works, including drainage in waterlogged areas; and rural connectivity for all weather access. The Central Government may notify further works in consultation with state governments. A work outside the Schedule is not a work under the Act, and wages on it are not wages under the guarantee.

Q: Why does MGNREGA bar contractors and labour displacing machinery?

Schedule II bars contractors from executing works under the programme and prohibits labour displacing machinery, because the programme’s wage bill is its object and not merely its cost. A contractor captures the margin between the sanctioned amount and the wage the worker receives, and machinery converts a labour budget into equipment hire, so both channels historically drained public works expenditure away from the workers the expenditure existed to reach. The bar operates together with the wage and material ratio, which holds the material component to forty per cent at district level so the programme remains labour intensive.

Q: How are fifty additional days of MGNREGA work provided in a drought?

Section 3 of the Act permits the Central Government to provide additional days of employment beyond the hundred days in areas affected by drought or other natural calamity, when so notified, and fifty additional days have been the notified measure, taking the entitlement in those areas to one hundred and fifty days in the financial year. In the drought litigation in Swaraj Abhiyan v Union of India, the Supreme Court treated the enhanced entitlement in notified drought areas as part of the obligation to be funded and delivered, rather than as an act of administrative grace.

Q: What did the Supreme Court direct in Swaraj Abhiyan on MGNREGA wages?

In Swaraj Abhiyan v Union of India, decided in 2016 in proceedings arising out of drought in several states, the Supreme Court directed the Government of India to ensure that compensation for delayed payment is made over to workers whose wages were delayed beyond fifteen days, as paragraph 29 of Schedule II postulates, and directed timely release of adequate funds against realistic state labour budgets. In a later phase of the litigation the Court held that a worker is entitled to payment within a fortnight of the work, that delays are not acceptable, and that neither government may pass the burden of compliance to the other.

Q: How does the household basis of the MGNREGA guarantee work?

The hundred days are an entitlement of the household, shared among its adult members, not a separate hundred days for each worker. Members may divide the days between them across the year as the family’s needs dictate, applying in separate blocks and at different seasons. The ceiling also operates in combination: liability for the unemployment allowance ends, among other events, when the household’s wages and allowance taken together equal the wages for one hundred days of work. A single person living alone counts as a household and holds the full entitlement in that person’s own right.

Q: What is the role of the district ombudsman under the MGNREGA framework?

The district ombudsman is an officer provided for under the state scheme and the Union’s instructions, rather than by the Act itself, to receive complaints from workers and the public about the programme’s administration, to inquire into them and to require the programme authorities to act on the findings. The office exists because the officers who administer the guarantee are the first respondents to complaints about it, and an appeal that runs only to the author of the act complained of is structurally weak. The ombudsman stands outside the executing line while remaining inside the programme.