Early this month, the Reserve Bank of India published a study with a reassuring message. Weak monsoons, it said, no longer hurt the rural economy as much as they once did. Farm output still moves with the rains, but non-farm work keeps growing regardless, and for the smallest farming households, those with up to one acre, wages now make up more than half of income. This was read as proof that rural India has grown resilient to a bad monsoon.

I spent a few years at the Ministry of Rural Development before becoming Finance Secretary, and resilience there was never an abstraction. It meant whether a family living on daily wages could put together three meals after a failed sowing season. On that test, the RBI’s study says almost nothing, because it measures output, not what rural households actually earned. When I looked behind the headline figure, I found no resilience story, only a decade of wages going nowhere, a shrinking safety net, and, as of three months ago, a change in law that strips away the protection that has mattered most to rural bargaining power in a generation.

What the Wage Data Actually Shows

Start with wages. Independent researchers using Labour Bureau data find real rural wages, adjusted for inflation, grew by less than 1% a year for most of the past decade, and fell outright between 2019 and 2023. Wages in rural construction, the biggest source of non-farm work, barely kept pace with the cost of living, and pay for artisans, weavers, carpenters and blacksmiths has been falling in real terms since at least 2014.

This year brought its own complication. The Labour Bureau quietly changed how it measures rural wages by adding a few small, high-wage states to its sample. Reported wage growth promptly jumped to nearly 20% a year, a number that made headlines before independent economists showed the real, underlying rise was closer to 4%, the weakest in four years. A change in measurement is not a change in reality.

The Same Story, Group by Group

The picture does not change by group. Small and marginal farmers, who the RBI says earn more than half their income from wages, lived through a sharply uneven year. The country recorded its highest-ever kharif foodgrain output, even as floods and drought together hit nearly 200 districts, with crop insurance payouts nowhere near covering the damage. Herders and dairy farmers saw their sector grow strongly on paper, but rising feed and fodder costs ate into whatever margin that growth should have delivered, and milk prices for consumers rose far faster than overall inflation. For the rural non-farm workforce as a whole, government data on small enterprises show hired workers’ pay growth collapsing from 18% to just 2% in a single year, which, once inflation is accounted for, means a real pay cut.

Diversification away from the farm, which the RBI credits with building resilience, appears instead to have shifted income towards salaried jobs for those able to secure them, not the work most rural households actually do, leaving them no better off than before.

The Floor That Gave Labour Leverage

What matters more is what has happened to rural labour’s bargaining power, not just its pay. MGNREGA’s real contribution was never only the wages it paid out. It was the floor it set under every other wage negotiation in the village. A contractor or a large farmer offering less than the notified MGNREGA rate risked losing workers to the public works site down the road. Economists who studied the scheme’s early years found that it raised private rural wages, especially in the lean season, simply by giving landless workers somewhere else to go. For the first time, a landless labourer negotiating with a landowner had a real alternative, not just a government promise.

That alternative has now been narrowed exactly where it counted most. Scheme workdays had already fallen to a decade low, and its budget had shrunk steadily as a share of the economy, even before this change. From the first of July, MGNREGA has been replaced by a law that raises the headline guarantee from 100 days to 121 days, but it lets states declare up to 60 days a year with no scheme work at all, specifically during the peak sowing and harvest season, a carve-out the old law never had. The new law targets precisely the months when landless labour had the most leverage, because that is also when large farmers and contractors most need their labour and were most likely to be undercut by the public works wage. If that leverage did not matter, lawmakers would not have bothered writing its removal into the Act. The fact that it was written into the Act at all tells us how much it constrained employers before.

What To Watch Next

None of this means the RBI’s narrower finding is wrong. Non-farm rural output probably has become less exposed to a bad monsoon, and that is worth knowing. But resilience measured in national output is not resilience measured in a labourer’s ability to refuse an unfair wage, and the second kind decides whether a family eats well after a weak season. Rural inflation has swung wildly through this period, from a sharp price spike in 2024 to outright food deflation the next year, and a fresh shock from sugar and cooking gas this year. The honest conclusion is that the aggregate story and the household story can both be true, and a reading that rests only on the first will miss the second.

Don't expect a clear verdict within one sowing season. India’s labour markets are too large and too varied by state for one legal change to show up cleanly in a single harvest cycle. States will declare their no-work windows at different times and lengths, employers will adjust at different speeds, and any shift in wages or migration will take several seasons, if not longer, to separate from ordinary noise. What should start now is the tracking, not the verdict: the Labour Bureau, state governments and independent researchers need to follow private wage rates in those windows, state by state, through this season and beyond, before anyone claims to have been proven right.