I want to separate graduate unemployment from the quality of the jobs graduates actually get.
I am building a state-by-year view of young graduates in India using PLFS 2017-18 through 2023-24. The main question is whether changes in unemployment and access to regular salaried work tell the same story across states.
Why I am asking this
Graduate unemployment often gets reduced to a single headline rate. I think that misses a second question: among graduates who do find work, what kind of work are they entering? A state can improve on unemployment while still offering weak access to regular salaried employment, or the reverse.
Who is in the study
The target group is young adults aged 20 to 29. I use the PLFS education categories corresponding to graduate and postgraduate education, with a lower-education comparison group for context.
Unemployment is measured among people in the labour force. Regular salaried access is measured among the employed. Contract, paid-leave, and social-security outcomes belong inside regular salaried employment. Mixing those denominators would answer different questions with the same-looking percentage.
What I am building
The plan is a consistent seven-round state-by-year panel from 2017-18 to 2023-24 using annual survey weights, pooled weighted ratios, official-table checks, and uncertainty measures where the survey design permits a defensible calculation.
What the paper can contribute
The broad fact that graduate unemployment is high is not new. The narrower contribution is to track unemployment and job quality together across states, and to distinguish changes within a state over time from persistent differences between states.
Where the project stands
I have assembled the documentation, variable inventories, denominator rules, and validation targets. The current work is harmonizing the annual PLFS rounds and checking the reconstructed estimates against official tables before I report substantive results.
What the evidence will not establish
This is not currently a causal design. Even if unemployment and regular salaried access move together within states, that would not by itself identify what caused the change. I want the paper to be clear about that from the beginning.