During a recent cricket match, one of the commentators, comparing the skills of two batters, said the difference between them was that between an IIT (Indian Institute of Technology) and an ITI (Industrial Training Institute); the unsaid point being the former is vastly superior to the latter. This seemingly innocuous comparison reveals a fundamental problem with India’s skill ecosystem – a bias towards elite technical institutions to the neglect of mass ones.The experience of successful industrialisers from England in the 19th century to China in the 21st, tell us that mass industrial technical training is essential. Such training is generally acquired via off-site technical institutes as well as on-site (on-the-job) in firms. In India, both these routes are in need of serious reform.A recent NITI Aayog working paper (Virmani 2026), notes that only 7.7% of Indian firms provide formal training to employees. A plausible explanation of such a low rate is that firms hope to hire workers already trained at someone else’s expense, reflecting a conflict between who provides the skills training and who stands to gain from it. In other words, skill training generates positive externalities, and like most such goods it tends to be under-supplied by the market. The policy response thereby, is to treat skilling as a public good, or else to engineer a market in which provision is subsidised and certified.While this problem is not unique to India, we do face a second fundamental challenge that very few economies encounter. In our context, what counts as a skill is often entangled with what counts as dignified work, and here the country is triply disadvantaged: by the near-universal hierarchy that elevates the prestige of mental over manual labour, by a caste order that maps that hierarchy onto birth and occupations, and by a colonial inheritance that split technical education into prestigious and subordinate tiers.India’s oldest skill training institutions, the ITIs, were designed to meet both these challenges. Established under the Craftsmen Training Scheme in 1950, ITIs were established to equip young trainees with practical skills aligned with the industry demands, and in turn meet the dual goals of reducing youth unemployment and enabling structural transformation. Being public institutions, ITIs were meant to meet the externality challenge by public provisioning of skill training at scale. They were also, by design, supposed to address the dignity challenge by replacing informally trained manual and technical workers into an accredited, formal workforce.We analyse administrative data from ITIs to gauge whether either of these challenges are mitigated, or if they continue to persist.Rapid growth led by private ITIsFor its first five decades, the ITI system remained mostly under public administration and grew slowly. This changed abruptly in the 2000s, when policy enthusiasm for harnessing the demographic dividend triggered an extraordinary surge, with more than a thousand new ITIs opening annually between 2010 and 2015. In two decades, the network more than tripled, from roughly 3,700 institutes in 2005 to over 14,500 today (Figure 1, left panel). In terms of expanding access to vocational training, this was a remarkable achievement.However, the expansion changed the character of the system since the growth was overwhelmingly private. Until the early 1980s, more than 80% of ITIs were publicly run as opposed to the boom years, when private institutes accounted for 70-85% of new establishments (Figure 1, right panel). Much of this growth was also restricted to rural areas, and while this expansion increased access faster than could be managed via public provisioning, it also pushed them further from industrial belts.Figure 1. The ITI boom during 2005-2015 (left) has been largely led by the private ITIs (right)Did the new ITIs fare as well as the old?Did the rapid expansion and privatisation affect the quality of instruction imparted in the ITIs for the better or the worse? The readily available metric is the rating of ITIs done by the Ministry of Skill Development and Entrepreneurship (MSDE). Each ITI gets a score on a 10-point composite scale that folds together admissions, pass rates, learning outcomes, trade diversity, and the inclusion of women and disadvantaged groups. Sorted by the period in which institutes were established, the pattern (Figure 2) reveals that ITIs set up between the 1950s and 1980s, predominantly public, continue to cluster in the highest quality bands. From the 2000-2004 cohort onwards, and sharply after 2005, the share of institutes scoring in the top band is drastically reduced, while those scoring below four increased steadily. Among the most recent cohort, only a small fraction achieves high grades.Figure 2. Newer ITIs fare much worse on quality parameters than older ITIsThe decline in quality as measured by the official index follows the privatisation of the system closely, and analysis disaggregated by public and private ITIs presented in SWI Report (2026) shows that the institutional quality of private ITIs is significantly worse than their public counterparts.Even in metrics other than the MSDE ratings, such as instructor availability and instructor qualifications, most of the ITIs do not perform well. Across large stretches of the northern states, more than three-fourth of sanctioned instructor posts lie vacant. Among the posts that are filled, qualification is rare: in nearly 72% of districts, fewer than one in four instructors hold the Craft Instructor Training Scheme (CITS) certification meant to qualify them to teach.To summarise so far, the growth in the ITI system shifted the public provisioning of skill training to private entities on the one hand, and on the other hand, it created institutions that persistently fail to meet a desirable standard in any metric whatsoever. Thus, it appears that the private solution to the under-provisioning of this vital public good has not been successful.Who enrols in ITIs, for which trade, and do they enrol by choice?Data on 96 lakh graduates between 2014 and 2022, show that the student demographics of ITIs are distinct from that of colleges and universities. The overwhelming majority enter right after Class 10, with very few coming from higher secondary or beyond. OBC (other backward classes), SC (scheduled castes), and ST (scheduled tribes) students are all more strongly represented in ITIs, while general-category students, who make up nearly 40% of higher education (and thus are over-represented with respect to population shares), constitute less than quarter of ITI trainees (Table 1).Table 1. Comparing enrolments by caste between higher education (AISHE) and vocational training (ITI) source:While it is tempting to read this as inclusion by design, the more accurate reading, perhaps, is exactly the opposite. ITIs absorb students from disadvantaged backgrounds not by institutional design, but because the higher-status pathways remain less accessible to such students. Dropout after Class 10 and 12 falls disproportionately on communities that have faced generations of barriers, funneling them toward vocational streams, while relatively better-off students continue into degree programmes and other higher educational pathways. The caste representation in ITIs, therefore, is a sign that these are viewed as sitting at the bottom of the educational hierarchy. Thus, it is safe to assume that for many of the trainees, enrolling in ITIs is a fallback option and not an aspirational choice.In terms of gender, men account for around 86% of trainees and concentrate in a narrow band of industrial trades such as electrician, fitter, welder, mechanic, etc., whereas female trainees, occupy an almost entirely separate set of trades such as sewing, dress-making, cosmetology, stenography, etc. (Table 2).Table 2. Top-10 trades by enrolment and gender. source:While the caste representation in trainee enrolment follows the larger social hierarchies, the gender segregation in trade choices reflect India’s labour market logic that codes labour into hard (masculine) and soft (feminine). Usage of industrial machinery and physical labour therefore is seen more as male while care work or industries such as textile are seen as female. The consequence is that vocational training, rather than loosening such segregation, reproduces it before women have even entered the labour market.Geography of skills versus geography of jobsFieldwork by Professors Srishti Yadav and Dontha Prashanth in Rewari (Haryana) and Karumanchi (Andhra Pradesh), reveals an interesting insight on how locating ITIs near an industrial belt matters for employment of the trainees. Field narratives emerging from Rewari – a town surrounded by industrial belts – suggest that ITI training does convert into employment opportunities related to the course completed by the trainees. By contrast, in Karumanchi, with no manufacturing belt within reach, similar training leads nowhere, as captured in the narratives of vocational trainees who report working on construction sites, painting buildings etc., for much lower rates of daily wages.These insights imply that the success of ITIs in providing a steady supply of trained workers to the manufacturing sector hinges upon the co-location of jobs and skills. This is not merely because of more job opportunities but because it enables trainees from nearby ITIs to actually visit workshops and factory floors and learn hands-on. However, once we plot the district-level number of ITIs and juxtapose it with the number of manufacturing units in the district, we do not find any pattern that indicates such strategic locationing of ITIs near industrial belts. The other way to examine this, is to look at the rural-urban gap. Currently, 75% of trainees graduate from a rurally located ITI, even though 65% of manufacturing establishments with hired workers are located in urban areas.Figure 3. No systematic relation between the number of manufacturing firms and the number of ITIs in a districT. source:The path aheadThe NITI Aayog working paper by Virmani (2026), calls for a National Job Skilling Policy and an annual skills-and-employment survey to close the gap between credential and competence. While such a survey is long-awaited and its implementation will certainly strengthen our understanding of skill-job mismatches, one must, in addition, address the two fundamental challenges we laid out at the outset.First, skill training, like education in general, have large spillover effects and hence are quasi-public goods. There is thus, an essential role to be played by the union, state, and local governments in supporting technical training institutes. But it is also true that governments may not be well-informed with regard to market conditions, for example, the way demand for skills will evolve in the future. Therefore, the private sector needs to play a crucial role in ensuring the curricula are updated and that the programmes are relevant enough to meet the industry requirements. As it stands, the ITI and the broader skilling ecosystem does not fare well on either of these counts.The second problem goes deeper than just a question of formal institutional reform, to a larger social and cultural reform. The “IIT versus ITI” comment reflects a deeply internalised social and cultural hierarchy of intellectual versus physical labour, and it is unlikely to be a rare sentiment even if it is not always expressed so explicitly. The idea, however, is not to wait for this culture to change. Rather, policies should focus on how to improve the social perception of vocational education. Early exposure to vocational training within the school curriculum, for example, could help reposition it as an aspirational choice rather than a fallback option. While we admit that no single measure may dissolve a hierarchy this old, the direction must be clear – that the road to Viksit Bharat must run through ITIs as much as through IITs.This article is republished from Ideas for India. Read the original hereThis post draws from the Chapter 5 and Chapter 7 (specifically, Section 7.5) of the SWI Report 2026, to which the same set of authors contributed significantly. The views expressed do not necessarily reflect that of the authors’ organisation or the I4I Editorial Board.Further ReadingVirmani, A (2026), ‘Education and skilling for employment: From credentials to learning outcomes’, NITI Aayog Working Paper, Government of India.“State of Working India 2026 – Youth in the labour market: Pathways from learning to earning” , Centre for Sustainable Employment,Azim Premji University.