There are problems that are indifferent to the level of economic advancement of a country. It was at once both a relief and a shock to learn that the problem of skills mismatch troubles not just India but advanced economies as well. However, this ‘common ground’ should not lull us into treating the problem as not deserving of immediate attention.
The writings on this matter make it clear that there are two parts to it: one is a matter of data and training; other is a matter of curriculum, where, unfortunately, a bureaucracy exists disguised as process, preventing any chance as of an agile response. What we have seen in India is that every attempt at adjusting curriculum is always trailing the pace of change, subject as it is to multiple levels of approval. ‘Reskilling’ should be seen as being no different from redefining and (re)developing curriculum. Calls for reskilling must explain the ‘how’.
This writing focuses on data, not curriculum.
Misreading the future
When we talk of skill mismatch, we refer to an environment where qualifications, skills, acquired experience are not in sync with what the industry and business demands. Of course, there is more than one kind of mismatch: Vertical mismatch, where people are either over or under qualified; horizontal where there is no match between the educational qualification and the job. Both may suffer from a skill gap, where people fall short of certain needed skills.
One obvious reason is the misreading of the skills required in the future based on a reading of how the future is shaping and hence what skills are likely to be needed. Once we accept that the world is dynamic it follows that we need to be equipped to respond to changing environments, assuming however that we have understood the changes and are able to progress to identifying and developing the requisite education and skills.
Employability
The Economic Survey of India 2025 announced a disturbing piece of data: only 8.25% of graduates have jobs matching their qualification and that over 50% of graduates and 44% of post-graduates are underemployed in low skill jobs. In October 2024, The Times of India (among others) reported that only 10 per cent of the 1.5 million engineering students were likely to be employed in 2024.
A possible explanation is overcrowding which is the result of thousands of students taking to engineering while, for many reasons, prospects of employment dwindled, indicating a mismatch between developing supply for an anticipated demand. A few weeks ago, Mumbai Mirror ran a news story about engineering students taking to doing memes as a means of livelihood, given the vast gap between the number of engineering students and employment opportunities. A March 2025 news report says that IIT placements fell from 90% to 80% from 2021 to 2024.
The India Employer Forum also highlighted the problem – “According to the India Employment Report 2024, about 80% of employers reported skill gaps in the IT, engineering and manufacturing sectors, which indicates the persistent issue of underemployment in India”.
An article by Anil Kapoor and Atul Tiwari in July 2025 argued that the problem is one of data gathering as well. According to the duo, “One of the primary reasons India faces this recurring challenge of incongruence between acquired and required skills is the lack of comprehensive, disaggregated data. Current surveys like the PLFS (periodic labour force surveys), though thorough in their approach and collecting valuable information on employment and education, do not fully account for the complexity of occupational demands, sectoral transitions, or evolving job roles within the services and technology sectors”.
They are championing for a “direct, detailed, and regular data on vocational training, skill acquisition, and labour market outcomes to address the mismatch. In coordination with skilling agencies, the Ministry of Statistics and Programme Implementation (MoSPI) must institutionalize a dedicated system for skill data collection. This should include data collection across different population groups, regions, and sectors to support evidence-based policymaking. Such a dataset would allow for forecasting emerging skill needs, identifying persistent vacancies, and curriculum updates that are better aligned with industry expectations”.
The role of data and training
In an article titled ‘When Jobs Change: Skills Mismatch and the Value of Training’, in the British Journal of Industrial Relations (22 June 2026), Lorcan Kelly, Paul Redmond and Luke Brosnan say this: “Using data on approximately 70 million online job vacancies, we estimate the extent to which occupational skill requirements changed between 2019 and 2023 across European Union (EU) countries plus Norway and the United Kingdom. STEM-related occupations saw the greatest degree of change, whereas lower skilled manual jobs saw the least. By linking job vacancies to survey data, we show that employees in fast-changing jobs are more likely to experience skills deficits. Training plays a role in mitigating these negative effects, and the type and intensity of training matters. Training seminars/workshops are most effective. Formal training courses and on-the-job training are less effective when used in isolation but can mitigate skills deficits when combined with other types of training”.
It is obvious that the incidence of mismatch varies across occupations, largely as a function of changes in the industry or business and the curriculum followed by educational institutions. In February 2026, the Times of India ran a story titled ‘Even Stanford and MIT graduates are struggling to find jobs’, which detailed a story that “For years, a CS degree from a top school was practically a job guarantee” but is now a completely different story.
In an essay titled ‘Mismatch in the 21st century: An overview’ in October 2026 in Science Direct, the seven authors observe that “According to the 2023 Survey of Adult Skills, the fraction of mismatched workers ranges between 33% and 36% in OECD countries, with notable differences across countries (OECD, 2024). Many firms report difficulties filling vacancies, while workers often hold qualifications or skills that are not fully used in their jobs. Meanwhile, the green transition, digitalization, artificial intelligence, aging, and work reorganization after COVID-19 are changing the demand for skills and the allocation of workers across firms, occupations, and locations”. OECD countries span North and South America, Europe and the Asia-Pacific.
The European Centre for the Development of Vocational Training in spring 2014 undertook the first European skills and jobs survey (ESJS), a large-scale primary data collection of about 49 000 adult employees in 28 EU Member States. The Foreword to the study states: Data repeated by several sources indicated that four in 10 EU employers said in 2013 that they have difficulty finding the right skills when recruiting. One of its observations was that educational qualifications are an imperfect signal of skills.
Getting the response right
How do we respond to this issue is going to be critical as it is an obstacle to growth. Data collection is step one – as disaggregated as is feasible to facilitate precise skill gaps. Else, we will stay at a level of generality which presents an incomplete picture, as Kapoor and Tiwary point out. This is where industry and government need to work together in a transparent manner without institutional egos becoming a hurdle. Supporting data is collection is training; of two kinds – one in data collection and the other is responding to the skill gaps.
