2004–2014 vs 2014–2026 — total incidents
Hover a bar for details.
Periods are split at 26 May 2014. Incidents are counted once each, on the date of the exam or of the leak.
Headline numbers
Patterns
- State recruitment exams are the epicentre. ~78% of all incidents are state-conducted (public service commissions, subordinate-service boards, state police) — Rajasthan, UP, Bihar, MP, Gujarat and Uttarakhand recur most. National exams (NEET, SSC, CBSE) get the headlines, but state government-job exams are the bulk. Because most exams are run by state governments of many different parties, these counts should not be read as a verdict on any single government.
- Two mechanisms dominate: pre-exam question-paper leaks (WhatsApp / print-press) and
OMR / answer-key / marks manipulation, impersonation or remote-access solving. About a quarter
of rows are the second kind rather than a classic paper leak — the
notefield says which. - Response has hardened. Exams cancelled in response rose from ~16% of incidents in 2004–2014 to ~61% in 2014–2026, alongside the central Public Examinations (Prevention of Unfair Means) Act, 2024 and several state anti-cheating laws.
- Convictions are rare and slow. The recurring pattern is arrests and CBI/SIT/ED probes, with convictions — where they come — arriving years later.
Peak years
Incidents cluster hard in 2021–2024 (14, 15, 11 and 12 incidents) before tapering into 2025–26 — REET, UKSSSC, TSPSC, BPSC/TRE, UP Constable & RO/ARO, and NEET-UG 2024 — the window in which paper leaks became a front-line political and street-protest issue.
⚠ Big caveat — read before comparing periods
The raw counts for the two periods are shaped by a large reporting & digitisation bias, and three forces all push the same way — fewer early incidents recorded, not necessarily fewer occurring:
- Digitisation & the smartphone. Pre-2012 news is sparsely online, and WhatsApp — which both causes modern leaks and documents them virally — barely existed in the earlier period. More leaks are detected and reported now, independent of how many happen.
- Survivorship in the old record. Every 2004–2014 row is coded Confirmed, because only the biggest, arrest-backed scandals left a durable online trace; the later period also captures unproven Alleged / Denied / Suspected claims.
- English-language / national-outlet skew. Regional-language leaks are under-represented, more so in earlier years.
Read the qualitative patterns above as the trustworthy signal; weight rows by
their confidence and leak_status when aggregating. This is a record of
reported incidents, not a complete census.