The Degree and the Job

Pritiraj BrahmaFounding Partner, Pritiraj & Partners · 28 August 2026

Apprentices engaged under NAPS rose from 3.08 lakh in 2020-21 to 9.85 lakh in 2024-25 — real growth, set against 4.33 crore students in higher education. Two ministries run one statute, two evaluations report employment on incompatible definitions, and nobody publishes the number that matters: how many apprentices hold a job a year later.

ProvenancePublished 28 August 2026. Apprentice engagement figures are from a PIB release and Rajya Sabha reply of 11 March 2026 (Ministry of Skill Development and Entrepreneurship) and from figures reported in May 2025 covering the five years to 2024-25. The application series discussed in the analysis is from apprenticeship portal data cited in a draft scheme document as reported in June 2026, not from a primary government release; it is treated as unverified and is not relied on as a finding. Outcome rates are from third-party evaluations by AJNIFM (NAPS) and NILERD (NATS).

Apprentices engaged under the National Apprenticeship Promotion Scheme rose from 3.08 lakh in 2020-21 to 9.85 lakh in 2024-25; across NAPS and the National Apprenticeship Training Scheme together, 48.4 lakh apprentices were engaged in the five years to 2024-25. That is real growth, and it should be said first. Set against 4.33 crore students in higher education, it is also an order of magnitude short of being the bridge its architects describe — and the state does not publish the one number that would settle the argument either way: how many apprentices hold a job a year after they finish.

This series has moved from who supplies the seats, to who teaches in them, to who certifies them, to who leaves. The last question is the one every family actually asks: does the degree lead to work. India's most direct instrument for answering yes is the apprenticeship — a legal relationship, under the Apprentices Act of 1961, in which a young person learns inside a firm, under workplace discipline, for a stipend, with a nationally recognised certificate at the end. It is the one skilling instrument whose outcome is not self-reported by a training provider, because a firm has to accept the person into its premises.

Five findings anchor this analysis:

1. Engagement has grown substantially, and this should be said first. Data placed before the Rajya Sabha on 11 March 2026 records apprentices engaged under NAPS rising from 3.08 lakh in FY 2020-21 to 9.85 lakh in FY 2024-25 — more than a tripling. Women's participation rose over the same period from 0.56 lakh to 2.25 lakh, lifting the female share from around 18 per cent to nearly 23 per cent. Figures reported in May 2025 put cumulative engagement at 35.5 lakh under NAPS and 12.9 lakh under NATS since 2020-21: 48.4 lakh in five years. On its own terms the scheme has scaled.
2. The scale is an order of magnitude below the population it is meant to serve. AISHE 2021-22 records 4.33 crore students in higher education, 78.9 per cent of them at undergraduate level. Against that, the best year yet delivered 9.85 lakh NAPS apprenticeships and, on the Ministry of Education's figures, 2.53 lakh NATS apprenticeships in 2023-24. Even read generously, the structured workplace-learning channel reaches something in the order of one in fifteen of the people moving through the system each year, and most of those come from the ITI and diploma stream rather than the degree stream. Apprenticeship is not currently a bridge from the Indian degree to the Indian job. It is a parallel track running alongside it — which is a defensible design, but it is not the design ministers describe.
3. The stipend is the binding number, and the state's own council has said so. Under NAPS-2, in force since 2022-23, the government contributes 25 per cent of the prescribed minimum stipend, capped at ₹1,500 a month, paid by direct transfer to the apprentice; under NATS the contribution is 50 per cent. At its 38th meeting on 26 May 2025, the Central Apprenticeship Council recommended raising the stipend band from the existing ₹5,000-9,000 range to ₹6,800-12,300 and linking it to the consumer price index for half-yearly revision, expressly to reduce dropouts. A recommendation is not a notification. Fifteen months later, the operative cap on the central share remains ₹1,500.
4. Where outcomes have been evaluated, they are good — and incomparable. A third-party evaluation of NAPS 2.0 by the Arun Jaitley National Institute of Financial Management found around 72 per cent of apprentices surveyed in employment after training. A NILERD evaluation of NATS reported 74 per cent securing full-time employment. Two evaluations, two ministries, two survey designs, no published common definition of "employed", no published sample frame, and no linkage to payroll records that would allow either figure to be independently checked. Numbers this good deserve better verification than they have been given, for their own sake.
5. One statute, two ministries, two registers. NATS is run by the Ministry of Education for engineering graduates and diploma holders, with an approved outlay of ₹3,054 crore for 2021-22 to 2025-26 targeting around 9 lakh apprentices; it engaged 8.68 lakh apprentices over five years, including 2.53 lakh in 2023-24, and its NATS 2.0 portal was launched on 30 July 2024 with ₹100 crore released for stipends. NAPS is run by the Ministry of Skill Development and Entrepreneurship for ITI pass-outs and a wider set of candidates. Both operate under the Apprentices Act, 1961. A firm wishing to take apprentices, and a student wishing to become one, face two systems, two portals and two stipend rules for one legal relationship.

A number that needs explaining

One series in circulation does not fit the picture above, and it deserves to be handled carefully rather than either amplified or ignored. Applications generated under NAPS are reported to have fallen from 42.8 lakh in 2022 to nearly 30 lakh in 2023, 25 lakh in 2024 and a little over 4 lakh in 2025, while employers continued to list opportunities — some 77 lakh vacancies created on the portal across the period. These figures come from portal data cited in a draft scheme document and reported in June 2026, not from a primary government release.

We have deliberately not made that a finding. A single non-primary source cannot carry a claim of that magnitude, and there is an innocent explanation available: portals change registration flows, deduplicate accounts and shift from open applications to matched recommendations, any of which can cut a reported applications count by an order of magnitude without a single young person losing interest. Engagement rose in the same year the applications figure fell, which is more consistent with a counting change than with a collapse in demand.

The governance point survives either way. A tenfold movement in a flagship scheme's headline demand indicator is now in public circulation, and the department that owns the portal has published neither the series nor the definitional note that would confirm or dispose of it. That silence is what allows the figure to be used as an attack and dismissed as an artefact, with no way for a reader to adjudicate.

The counter-case

Three arguments deserve to be put honestly.

First, the scale criticism cuts both ways. No country runs workplace apprenticeship for the whole of its tertiary cohort; Germany's celebrated system covers a vocational track, not its universities. Judging NAPS against 4.33 crore students sets a bar no apprenticeship system anywhere has cleared, and a scheme that tripled in four years is performing, not failing.

Second, apprenticeships are not the only bridge, and treating them as the sole measure of employability is unfair to a system where campus placement, internships and the informal apprenticeship of family enterprise do enormous work.

Third, the comparison with Germany, Switzerland and Austria — routinely invoked — understates what those systems rest on: chambers of commerce with statutory functions, employer associations that co-own the curriculum, and a vocational route with social standing. India's law and portal exist; the institutional scaffolding around them does not, and no stipend revision creates it.

The counter-case is strong enough that a fair verdict on the apprenticeship mission is "promising, growing, and inadequately instrumented" rather than "failing". That verdict is still a governance finding. A programme cannot be steered on numbers whose meaning its own administrators have not published.

What we would do

1. Publish the application series officially, with its definitional note. Registrations, applications, engagements, completions and dropouts, by state, sector and scheme, monthly, with a stated note whenever the counting method changes. Two ministries can retain their programmes; one number has to be authoritative. Until that exists, the 2025 figure will be quoted by critics and dismissed by officials, and neither will be able to prove the other wrong.
2. Make the published outcome an absorption rate, linked to payroll. The measure that matters is the share of apprentices on a formal payroll six and twelve months after completion, established by linkage to EPFO records rather than by survey. India already has the administrative data to do this. A 72 per cent survey figure and a payroll-linked figure may well agree — in which case the scheme's case becomes unanswerable, which is reason enough to do it.
3. Notify the stipend revision the Council recommended, and index it. The Central Apprenticeship Council recommended the increase and CPI linkage in May 2025 to reduce dropout. Either the recommendation was right, in which case delay is costing cohorts, or it was wrong, in which case the reasons should be published. An indexed stipend also removes this decision from the political calendar permanently.
4. Make the apprenticeship credit-bearing inside the degree. Under the national credit framework, a completed apprenticeship should carry defined credits towards the degree, recorded in the academic bank of credits, so that a student does not choose between finishing a qualification and gaining workplace experience. This is the single change that would move apprenticeship from a parallel track to the bridge, and it requires regulation, not money.
5. Publish employer-wise engagement and absorption. Firms that take apprentices and hire them should be visible; firms that take stipend support and absorb nobody should be equally visible. The information is already collected. Publishing it turns a subsidy into a reputation, which is generally the cheapest enforcement instrument a state possesses.

This series began with 91 lakh additional students and ends with 9.85 lakh apprenticeships. Between those two numbers sits the whole of India's employability argument. The state has built the statute, the portals, the stipend and the schemes. What it has not built is the single published record that would let anyone — a minister, a firm, a family — say with confidence whether the degree is reaching the job. Politics is the promise. This is the delivery, and it is one register away from being visible.

Companion essay The Certificate Is Not the Job — Why India Skills Millions and Employs Few