GovernanceOpen access · Free to read
Creamy Layer Leakage & Quota Capture in Public Sector Employment
By Constitutional Policy Institute · Published 02 May 2026 · 10 min read
Audit of public recruitment showing 68% of category benefits accrue to the top 15% affluent tier within reserved groups.
Abstract
This whitepaper investigates category benefit distribution across Class-I civil services and public sector undertakings over the past three decades. Data reveals severe benefit concentration among second-generation affluent households, advocating for strict generational sunset clauses.
Key Findings
- ✓Top 15% affluent urban families capture nearly two-thirds of quota benefits.
- ✓First-generation underprivileged applicants face structural exclusion within categories.
- ✓Recommends strict 1-generation exclusion limits once a family reaches middle-class status.
Open-access research paper · Free to read, no paywall
Concentration, not eligibility, is the finding
The question this audit asks is deliberately narrow. Within a reserved category, how are the benefits distributed? Not between categories — within one. That framing removes the most common objection to work of this kind, because nothing here compares one community to another. It compares a community to itself.
Across Class-I civil service appointments and public sector undertaking recruitment over three decades, the answer is consistent: roughly two-thirds of category-reserved positions were filled by candidates from the most affluent fifteen per cent of households inside that category. These are, overwhelmingly, second- and third-generation beneficiaries — urban, salaried, English-medium educated, and often the children of earlier beneficiaries of the same provision.
Why this happens, mechanically
There is nothing sinister in it. A reserved seat still has to be won in a competitive examination against other members of the same category. Inside that contest, the same advantages apply as anywhere else: coaching, a stable home, an educated parent who understands the process, and the financial ability to attempt the examination three or four times. A first-generation aspirant from a landless rural family competes against these advantages with none of them.
The result is that the provision works precisely as an examination works, and the households that were already best placed capture it. The tag does not distinguish between a landless labourer's daughter and an IAS officer's son. Both carry the same certificate into the same restricted contest, and one of them has been preparing for it since Class Six.
This is the mechanism the phrase 'creamy layer' was invented to describe, and India already accepts the principle — it applies an income ceiling to some categories and not to others, and the ceiling is rarely revised or enforced against non-salaried wealth.
What a sunset clause would actually do
The recommendation is a generational limit: once a household has produced a beneficiary who reached a Class-I position or an equivalent income band, the household's next generation competes on the general merit list while retaining every anti-discrimination protection. Nothing is withdrawn from anyone who is still poor.
The predictable objection is that this punishes success. It does the opposite: it recognises that the provision succeeded for that family and releases the seat to a family for which it has not yet succeeded. A ladder that the first climber pulls up behind them is not a ladder, and a ladder that the same family occupies for four generations is not one either.
The data needed to implement this already exists in service records and tax filings. The obstacle has never been technical.
Analysis: Creamy Layer Leakage & Quota Capture in Public Sector Employment
“Creamy Layer Leakage & Quota Capture in Public Sector Employment” is filed under the governance track of the movement's open library because the question it tests is measurable, not rhetorical.
Audit of public recruitment showing 68% of category benefits accrue to the top 15% affluent tier within reserved groups. Published 02 May 2026 by Constitutional Policy Institute, the paper runs to roughly 10 min of reading and is released open access so that students, journalists and policy staff can cite it directly.
What each finding actually shows
Finding 1 · Governance
Top 15% affluent urban families capture nearly two-thirds of quota benefits. Read against the stated intent of the provision, the gap between design and delivery is the finding — not the individual beneficiaries, who are simply using the system as written.
Finding 2 · Governance
First-generation underprivileged applicants face structural exclusion within categories. Anyone contesting this is welcome to publish the counter-figure; the movement's standing offer is to print corrections with attribution rather than argue in the abstract.
Finding 3 · Governance
Recommends strict 1-generation exclusion limits once a family reaches middle-class status. The number matters less than its direction: it has not corrected itself over successive review cycles, which is exactly what a time-bound corrective was supposed to do.
Why it matters for reform
The reform conclusion follows from the evidence rather than from sentiment: replace inherited category tags with a verified need index so that the same rupee of support reaches the household that has none of the advantages the tag was meant to proxy.
Sources, method and limits
Figures in “Creamy Layer Leakage & Quota Capture in Public Sector Employment” are drawn from publicly available material current to 2026 — parliamentary answers, commission reports, university and PSU disclosures, RTI responses and court records — and are reproduced for public education under Article 19(1)(a). Where the record is incomplete, the paper marks the gap instead of estimating over it. This is civic research and commentary, not legal advice, and it advocates only lawful, non-violent, constitutional change. Reuse is permitted under CC BY 4.0 with credit to the Reservation Hatao Andolan; see the legal and transparency charter and the NEMI policy framework.
