TL;DR The government has now paid for the reform that people like me asked for since 2018: a stipend subsidy large enough to pull young people into internships. On that measure it worked, and the joining rate rose to 9.5% of posted seats in Round III from 6.9% in Round I. But completion fell to 39.4% from 42.6%, so six in ten who join still leave. The binding constraint has moved from money to quality. The fix is to pay for completed internships backed by a filed learning plan, not for posted seats.
What happened
The Indian Express reported on 21 October that the Prime Minister's Internship Scheme (PMIS) drew sharply more young people after its revamp this year. The report uses data from the Ministry of Corporate Affairs, and it is the first hard look at Round III. The scheme opened as a pilot on 3 October 2024 with a promise of one crore internships in five years.
- Joining rate: 9.5% in the ongoing Round III, against 6.1% in Round II and 6.9% in Round I. Joiners are up 67% to 100% over the earlier rounds.
- Completion rate: 39.4% in Round II, down from 42.6% in Round I.
- What the March revamp changed: duration cut to 6 or 9 months from 12, monthly allowance raised to Rs 9,000 from Rs 5,000, age band widened to 18 to 25 from 21 to 24, master's holders and final year students made eligible, and MSMEs and sunrise sectors brought in beyond the top 500 companies.
- Who actually joined: 8,760 interns in Round I, 7,308 in Round II and 14,620 in Round III as on 6 August, according to a Civilsdaily summary of the ministry's numbers. That is about 30,700 in two years.
Mohit, a BBA student sent from one petrol pump to another by a public sector oil refiner to push its payment app, described his month there in one line: "There was no knowledge or learning, so I dropped out."
Why this matters
This matters because the stipend argument has now been tested on a national scale, and it held. Money moves young people. What money cannot do is put a trainer in the room.
For two years the standard criticism of PMIS was that the sum was too small and the door too narrow. I made that case in STIPEND SUBSIDY IS THE ANSWER in January 2018, arguing that the state should subsidise the stipend so that employers hire young people they would otherwise not risk, and in PM Internship Scheme: Expand to 50,000 Companies in February 2025, asking for the host base to widen beyond the top 500. Both were delivered in part. The stipend nearly doubled and MSMEs were allowed in. The result, in today's data, is that participation is no longer the crisis.
The crisis is what happens on day one at the host firm, and the three interns in the Indian Express report show it better than any percentage does. Padma, 23, an electronics and communication engineering graduate from Vijayawada, joined a top IT company in January 2025. She got four months of instructor-led training, two months of self-paced digital learning and mentoring, and six months of on-the-job training. She then sat the National Qualifier Test and had a full-time offer from the same company in March 2026. Sunil, 22, holds a diploma in computer hardware maintenance from Ranchi. A microfinance and rural banking company sent him to villages to recover loans. He left midway and is still unemployed. Mohit, from Lucknow, was sent pump to pump for a cashback promotion.
One scheme, one government, three outcomes. The variable that separated them was not the stipend. It was whether anyone had written down, and then delivered, a training plan.
| What changed | Round I (2024-25) | Round III (2026) |
|---|---|---|
| Monthly stipend | Rs 4,500 by DBT plus Rs 500 from company CSR | Rs 9,000 |
| Duration | 12 months | 6 or 9 months |
| Age band | 21 to 24 | 18 to 25 |
| Host firms | Top 500 companies | Plus MSMEs and sunrise sectors |
| Joining rate | 6.9% | 9.5% |
| Completion rate | 42.6% | 39.4% (Round II, latest available) |
The best case for the government deserves a hearing. Officials quoted in the report say the challenge is for young people to take the internship seriously and finish it, and the report also names two interns who left for other full-time jobs or personal reasons, not because the work was empty. A shorter stint of six or nine months, plus nearly double the money, may well lift completion in Round III, and we will not know until it closes in December. Round II's dip from 42.6% to 39.4% is still a dip, though, and it came before the revamp. The money went up after that slide began.
What this means for interns, host companies and the ministry
For the intern: the offer letter tells you very little
Ask, before you join, what the first four to eight weeks look like and who is teaching. A structured internship has a named trainer, a schedule and an assessment. Padma's did. If the answer is that you will "learn on the job" and the job is a field beat, you are being hired as sales or collection labour at a subsidised wage, and you should say so on the portal when you leave.
For host companies: the dropout rate is about to become a public number
Once joining is no longer scarce, the published dropout rate is the figure that will separate serious firms from firms using the scheme to staff a field beat. A manufacturer that converts interns into hires is advertising something a microfinance branch is not. Expect students and colleges to read those numbers before applying, and expect your CSR story to be judged by them.
For the ministry: pay for completion, not for posting
Three changes would move the binding constraint in one round. First, make a filed learning plan, role by role, a condition of onboarding, so an intern cannot be placed before somebody has written down what will be taught. Second, record a dropout reason code for every exit and publish it by host firm, flagging firms with high dropout. Third, pay the subsidy against completion and tie completion to a recruitment test, the route that demonstrably converted for Padma.
Then there is the arithmetic. One crore in five years is 20 lakh a year. About 30,700 have joined in two years, which leaves roughly 99.7 lakh in the remaining three, or about 33 lakh a year. Round III had 14,620 joiners as on 6 August. If it holds that April-to-August pace until it closes in December, it will end near 33,000, which is about 1% of the annual requirement. That is my arithmetic on the reported numbers, not the ministry's projection. It is also the least emotional way to say that the one crore target is unreachable on the current design, because a joining rate is a measure of demand for seats, not of anyone's skills.
What to do now
These are recommendations, most urgent first.
- This week: the Ministry of Corporate Affairs should state whether a learning plan is mandatory for every posted role. If it is not, make it so before Round IV opens.
- This month: add a dropout reason code to the exit flow on the PMIS portal, with a short fixed list of reasons, and report the aggregate by sector.
- When Round III closes in December: publish completion and placement separately for manufacturing, IT and services, and for roles that are essentially field sales or collections. The reported gap between those categories is the whole story.
- Longer term: shift the subsidy from posted seats to completed internships, with a bonus for firms that convert an intern to full-time work, and link completion to NQT-style recruitment tests.
What not to do: do not read a 9.5% joining rate as success, and do not scrap a scheme that has already proved the hard part, which is that young people will show up. The design is the problem, not the idea.
The bigger picture
The backdrop is grim and it is not improving. PLFS data for August 2026 put youth unemployment at 15.1%, three times the overall rate of 5%, and the monthly series has ranged between 15.1% and 16.2% through the first five months of this financial year. The National Career Service portal had 2.89 crore active jobseekers against 17.68 lakh active vacancies as on 15 July 2026, according to a PIB release of the labour minister's Lok Sabha reply. Against that queue, a subsidised internship that teaches nothing is not a stepping stone. It is a delay with a certificate attached.
I made a related point in When Public Money Buys Privilege, Make It Buy Skills Too, where I argued that when the state pays, it must be able to see what it bought, and warned that the likely failure mode was token internships that tick a compliance box. Two years of PMIS data now reads like that warning with names attached. The thing to watch is not the joining rate in Round IV. It is the completion rate when Round III closes in December, because that number decides whether this scheme becomes a labour subsidy with a training story or an actual skilling programme.
Sources
- Aanchal Magazine, "Two years later, PM internship revamp draws more youth, but quality gaps persist", The Indian Express, 21 October 2026. Link
- Jagriti Kumari, "पड़ताल: 5 साल में 1 करोड़ युवाओं को इंटर्नशिप देने वाली थी मोदी सरकार, 2 साल में कितनों को मिला?", Hindustan (Live Hindustan), 21 October 2026. Link
- "Two years later, PM internship revamp draws more youth, but quality gaps persist", Civilsdaily, 21 October 2026. Link
- "BT Big Story: No jobs, no security, the crisis facing India's educated youth", Business Today, 21 October 2026. Link
- Press Information Bureau, "National Career Service Portal Mobilizes Over 10.05 Crore Vacancies Since Inception", 5 August 2026. Link
No comments:
Post a Comment