India’s skilling problem is less about the supply of courses than about the weakness of the signal a course sends. Training is plentiful. What employers lack is a reason to believe that a certificate means a person can do the job on Monday. Where an employer has agreed to hire the graduate, or helped write the curriculum, the signal is strong and training works. Where neither has happened, the certificate is mostly a receipt.
Why certificates stop meaning anything
A hiring manager in a factory or a hospital has a plain problem. The candidates’ skills cannot be observed before the salary is paid. A credential helps only if past holders of it turned out to be good. When the same certificate is issued by providers of very uneven quality, its average value falls, employers learn to discount it, and they fall back on other proxies: a degree, a reference, a known college, a probation period. Good trainees from good programmes then pay for the failures of everyone else.
India has built much of the machinery of skilling. The Skill India Mission brought short-term training together under schemes such as Pradhan Mantri Kaushal Vikas Yojana. Industrial Training Institutes, run mostly by state governments under national standards from the Directorate General of Training, teach trades such as fitter, electrician and welder. Sector Skill Councils, led by industry, write occupational standards and run assessments. None of this is useless. But a programme measured on enrolment or certification will deliver enrolment and certification, and the incentive to train a person can drift away from the incentive to place one.
What changes when an employer commits
Two things change. First, the employer takes on a risk the trainee cannot carry. A person who spends months on a course with no promise of work is betting on a signal they cannot check. When a firm commits to hire everyone who passes a defined test, the bet becomes a job offer with conditions. Second, the employer supplies the part of the curriculum no training provider can: the machine on the floor, the tolerances that matter, the plant’s quality system, its safety rules.
Apprenticeship is the oldest form of this, and India already has the law for it. The Apprentices Act, 1961 lets firms engage apprentices on a stipend, and the National Apprenticeship Promotion Scheme (NAPS) shares part of the stipend cost with the employer. The scheme exists because the cost to a small firm of taking on someone who cannot yet produce is real. Take-up has been uneven. Paperwork weighs on small firms, and many owners fear training a worker only to lose that worker to a competitor down the road.
A certificate is a promise from the provider. A job offer is a promise from someone who will pay if it turns out to be false.
National Education Policy 2020 pushes the same way from the school end. It calls for vocational exposure to be built into school and higher education instead of being kept as a separate, lower-status track. That is the right direction, but it hands schools a problem they cannot solve alone. A teacher cannot prepare a student for a trade unless some local employer will say what good work looks like.
What a good employer-linked model needs
The parts are mostly dull, and each one fixes a specific failure.
- A hiring commitment made before training starts, with a clear test of what passing means.
- Curriculum written or reviewed by the employer, with time on real equipment early, not at the end.
- Assessment by someone other than the trainer, and published figures on how many passed and how many were hired.
- A stipend during practice, so that families without savings can take part.
- Several employers per trade in a district or cluster, so the trainee is not tied to one firm and no firm bears the whole cost of training workers a rival may hire.
The last point matters most for manufacturing. Much of its skill is tacit: how a setter hears a machine going out of tolerance, how a welder reads a joint. It is learned beside someone who knows it. In clusters where many small firms make similar parts, an industry association or a cluster body can pool apprentices and share trainers and equipment. That spreads the cost, and a worker lost to a neighbour becomes a hire from a pool that every member draws on.
Our position is that public money should follow placement, not capacity. Paying a provider for seats filled produces seats filled. Paying for graduates who are hired and still in the job after a fixed period moves the provider’s attention to what employers want. Critics will say this rewards providers who pick easy candidates and leaves out those who need the most help. They are right to worry, and the fix is in the design: pay more for placing people from weaker starting points, and verify through payroll records instead of provider reports.
What we are watching is whether placement-linked funding can be audited cheaply. The first scheme that publishes, trade by trade and district by district, how many trainees were hired and kept will also give employers the signal they have been missing.