PLI Scheme Reality Check: Rs 35,354 Crore Paid, 14 Lakh Jobs

Key takeaways

• PLI has disbursed Rs 35,354 crore against an approved outlay of about Rs 1.97 lakh crore. Six years in, roughly 18 per cent of the money has moved.
• That Rs 35,354 crore sits against Rs 2.4 lakh crore of realised investment, over Rs 15.2 lakh crore of cumulative exports and about 14.15 lakh jobs.
• Roughly Rs 2.5 lakh of public money per job. On jobs per rupee, this is the best-value industrial programme India runs.
• PLI pays on production, not capex. The cheque arrives after you build the plant and beat a sales threshold. The investor carries all the risk up front.
• The US tariff went 25 per cent, then 50, then 18 inside six months. That is the risk premium stopping the factory decisions that would claim the rest.

PLI gets written up as either a triumph or a waste. Both readings miss what the numbers say. The scheme worked extremely well where it worked, and has barely been used anywhere else. Those are the same fact, and the reason sits in the design.

How a PLI rupee actually gets paid

Read that as an investment memo. To reach five thousand crore of revenue you commit land, machinery, working capital and several years before one rupee arrives. That is a defensible design, and it is why nobody has been paid for a factory that never produced. It also means only firms already willing to place a very large bet ever claim it. Lower step two and the take-up changes. Raise the incentive rate and it does not.

Rs 35,354 crore is a bargain, and that is the problem

Rs 35,354 crore against 14.15 lakh jobs is roughly Rs 2.5 lakh of public money per job. The AI data centre projects announced this year work out near Rs 2.85 crore per job, which I have set out in the case for India as AI landlord. On this measure a factory is about 114 times better at turning money into employment. Which makes 18 per cent the strange number, not 35,354.

The risk environment, in two charts

Eighteen is far better than fifty, and the February deal is genuinely good news. But a factory is a ten-year decision. The question is not the rate today, it is your confidence in the rate in year four. The whipsaw is itself the cost, independent of where the number settles.

Treat that as a forecast, because it is one. But it is the number sitting inside the model of anyone underwriting an export plant this year. The money is not unavailable, it is unwilling, which is the same pattern driving the Indian funding drought. Capital did not leave. It repriced Indian risk.

Where it worked, and how shallow it still is

Electronics is the exception that explains the rest. It accounts for about Rs 6.2 lakh crore of the Rs 15.2 lakh crore of cumulative PLI exports, two fifths of the programme from one sector. Uttar Pradesh now makes close to 65 per cent of India’s mobile phones, behind Rs 28,000 crore of electronics investment and Rs 32,146 crore of approved semiconductor investment. That is the cluster effect working.

The depth is the problem. Displays, memory and printed circuit boards still arrive from outside, so 81 of every 100 rupees of factory-gate value leaves the country. We assemble at world scale and manufacture at a much shallower one. PLI 2.0 targets more than 55 per cent value addition, aligned with a Rs 40,000 crore components scheme. That is the right correction and it has not happened yet.

What I think happens next

• Total disbursement stays under half the outlay. Across the life of the 14 schemes, expect payouts below Rs 1 lakh crore, with the rest quietly re-appropriated.
• Thresholds get lowered before outlays get raised. The binding constraint is the qualifying bar, not the incentive rate.
• Concentration gets worse. Electronics takes a rising share of both disbursement and exports. Textiles, solar, drones and auto components do not converge.
• Value addition improves slowly. Moving 19 towards 30 is a components and chemicals problem, not an assembly one. It takes the rest of the decade.

What to do about it

If you manufacture, read PLI as a rebate on success, not a subsidy for starting. Build only if the plant works at zero incentive. Every business I have watched get hurt by an incentive scheme had built the model with the incentive inside it. If you supply, the opportunity is upstream in components, where value addition has to go, where the next scheme points, and where domestic competition is thin. And if you are only reading headlines, hold both numbers at once: Rs 35,354 crore for 14 lakh jobs is excellent, and Rs 35,354 crore out of Rs 1.97 lakh crore is a poor take-up rate. The distance between them is a policy problem, not an accounting one. Who actually underwrites that risk is the question I take up in who will fund India.

Read next: the AI data centre landlord case study for the Rs 2.85 crore per job comparison in full, and what happened to venture capital in the same year for the same risk repricing on the venture side.

Sources

• Ministry of Commerce and Industry, PLI progress to 31 March 2026
• PIB releases on PLI investment, production, exports and employment
• MeitY, large-scale electronics manufacturing PLI data to February 2026
• Global Trade Research Initiative, India-US trade projections
• Government of Uttar Pradesh electronics and semiconductor statements

Figures are as reported by the sources named above at the time of writing.