The Modi government’s replacement of the MGNREGA with the newly introduced VB-G Ram G Bill comes at a deeply fragile moment in India’s labour market landscape. The new Bill, which is going to affect millions of lives and livelihoods, almost without any parliamentary consultation, has been railroaded through Lok Sabha. It reflects a lack of will and intent on part of the government to show any care for the actual respondents of the Bill, i.e. the rural workforce seeking employment guarantee in absence of alternative employment. When the Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) was passed in 2005, it marked a conscious effort to rethink the Indian state’s relationship with rural workers at that time. In doing so, India took a rare step: it turned employment into a legally enforceable right – and set a critical example for the developing (and developed) world. Rural households were no longer passive recipients of state largesse; they could demand work from the state, and the state was legally obliged to provide it within 15 days or pay an unemployment allowance.Close to 20 years later, the VB-G Ram G Bill now promises to fortify this architecture by raising guaranteed workdays to 125. A common misread is to treat MGNREGA as a programme whose success hinges on the number of days sworn into law. But that’s never the real constraint. The scheme’s own data show the core issues have always been fiscal and institutional, centred around who pays when demand spikes, how quickly funds flow, and whether states can enforce a legally binding right.To answer that, one must read the data slowly, year by year.Also read: Workers and Economists Protest as Government Sounds MGNREGA’s Death KnellThe early years: fiscal seriousness, uneven delivery (2009-2013)In the initial 10 years of operation, MGNREGA had excellent political and fiscal backing from the Union government. In 2012-13, allocations for the Ministry of Rural Development had nearly tripled from 2006-07; the MGNREGA itself accounted for about 36% of that ministry’s budget. Unskilled wages were entirely paid for by the Union government and funds were released based on state labour budgets that were supposed to reflect expected demand.Employment generation was sizeable during this period. In the year 2009-10, MGNREGA generated 284 crore person-days of work throughout the country. This decreased in 2010-11 to 257 crore person-days even while expenditure increased marginally. This early divergence between expenditure and employment indicated administrative and implementation caps.For most of its life, MGNREGA never really delivered on the promise of 100 days of work. The average rural household got much less. In 2010-11, Tamil Nadu managed about 54 days per household. Punjab barely crossed 27. Most states stayed far below the legal ceiling.Where work was generated mattered as much as how much. States with stronger administrative systems did better, year after year. Tamil Nadu and Andhra Pradesh stood out. Poorer states did not. Uttar Pradesh, Bihar, West Bengal, and Madhya Pradesh together account for close to 60% of India’s rural Below Poverty Level population. Yet they generated only about a third of total MGNREGA employment. Andhra Pradesh and Tamil Nadu, with far fewer poor households, accounted for almost a quarter of all person-days.We can gather here that from its inception, MGNREGA operated less as a pure needs-based entitlement and more as a scheme whose results were contingent upon state capacity.A plateau sets in: stable budgets, declining intensity In the years to follow, MGNREGA entered a phase of apparent stability. Allocations stopped growing rapidly but did not collapse either. By 2018-19, the Union government was allocating Rs 55,000 crore to MGNREGA. That was higher than earlier budget estimates, but it was also the same as the previous year’s revised figure. On paper, nothing had changed. On the ground, things were thinning out.Total person-days of work fell from 236 crore in 2016-17 to 176 crore by January the following year. Households were getting less work too. The average dropped from 49 days in 2015-16 to just 39 two years later.This wasn’t because rural distress had eased. It hadn’t. What had changed was the room states had to respond. Budgets were tighter. Releases were slower. Once funds ran out, work stopped.What changed instead was the fiscal environment. States increasingly found themselves operating close to their budget ceilings. Once allocations were exhausted, additional demand translated into delayed payments, pending liabilities, or outright rationing of work.Also read: ‘Replacing Father of Nation with G Ram G’: Govt Pushes Ahead to Table VB-G Ram G Bill in Lok Sabha Amid UproarWages compounded these problems. Although the Centre notified minimum wage rates every year, a majority of states invariably paid less than those rates in practice. In FY 2016-17, Tamil Nadu’s notified wage was Rs 203 per day, though the average wage paid was only Rs 140. Rajasthan paid an average of Rs 126 against a notified rate of Rs 181. For many households, MGNREGA wages no longer provided a credible income floor.By the late 2010s, the scheme was still large – but thinner, slower and less reliable.The pandemic shock: what MGNREGA could still do (2020-2022)The pandemic made this clearer than anything else. When urban jobs disappeared and migrant workers came back to villages, demand for MGNREGA shot up. In 2020-21, 8.6 crore households asked for work. About 7.6 crore got it. No other policy instrument absorbed rural distress on this scale.At the same time, the limits of the system were exposed. Even at peak expansion, 12% of households who demanded work were denied employment – a direct violation of the Act. Nearly one in eight households that demanded work were not being provided it, despite the legal guarantee. States spent far more than they were allocated. By 2022–23, expenditure had crossed 117% of available funds. Pending payments climbed past Rs 8,400 crore. States were paying wages first and hoping the Union government would settle the bill later.The MGNREGA worked during the crisis precisely because this reimbursement eventually came. Without that implicit central backstop, the scheme would have collapsed under its own demand.The present moment: demand persists, guarantees weakenRecent data indicates that high demand has not receded. By January 2022-23, 6.3 crore households had demanded work. Only 5.6 crore got it. Nearly 80 lakh households were left out, even though the law said otherwise.Timing is everything. Unmet demand peaked in the months when agricultural employment is usually low. This, of course, is when an employment guarantee is supposed to matter most.Payment delays further hollowed out the scheme. Although digital fund transfers improved some aspects of processing, only about 30% of wages were credited within the legally mandated 15 days. Compensation for delays another legal entitlement was increasingly not paid. In most large states, less than 70% of approved delay compensation actually reached workers.At this stage, MGNREGA in practice, is transformed into a rationed programme.Reading the VB-G Ram G Bill through this critical history of MGNREGAIt is against this empirical record that the VB-G Ram G Bill has to be judged. The headline promise of 125 days of work sounds expansive, but history shows that statutory ceilings were never the binding constraint. At no point did the average household even approach 100 days of employment.It was fiscal willingness and administrative capacity that constrained employment.The most consequential change in the new framework is the shift to a 60:40 cost-sharing arrangement between the Centre and states. Excess expenditure during periods of high demand under MGNREGA was eventually borne by the Union government. States will face a hard budget constraint under the new framework. Once their share is exhausted, employment will have to be rationed.This is a change that will hurt the states unevenly. States that are financially and administratively stronger will manage. The poorer states, which have the largest employment demands, will suffer first.The new framework also links employment more closely to centrally determined asset priorities and introduces breaks during peak agricultural seasons. Taken together, these changes in approach relocate the purpose of rural employment programmes from social protection towards labour-market management and infrastructure delivery.That shift – from rights-based employment to budget-managed provisioning – is the real story. And it is a story the data has been telling for years. Geetaali Malhotra, Aditi Lazarus and Arun Nagappan contributed to this article.